WORDS IN ASIAN CULTURAL CONTEXTS (REVISED VERSION)

Transcription

1

2 WORDS IN ASIAN CULTURAL CONTEXTS (REVISED VERSION)

3

4 WORDS IN ASIAN CULTURAL CONTEXTS (REVISED VERSION) Proceedings of the 4th Asialex Conference 1 3 June 2005 M Hotel, Singapore EDITED BY VINCENT B Y OOI, ANNE PAKIR, ISMAIL TALIB, LYNN TAN, PETER K W TAN AND YING YING TAN DEPARTMENT OF ENGLISH LANGUAGE AND LITERATURE, FACULTY OF ARTS AND SOCIAL SCIENCES & ASIA RESEARCH INSTITUTE, NATIONAL UNIVERSITY OF SINGAPORE

5 Jointly published in Singapore 2005 by Department of English Language and Literature, Faculty of Arts and Social Sciences, National University of Singapore; and Asia Research Institute, National University of Singapore 10 Kent Ridge Crescent, SINGAPORE , Republic of Singapore DELL FASS, ARI All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronically or mechanically, including photocopying, recording or any informational storage or retrieval system, without either prior permission in writing from the publisher or a licence permitting restricted copying. Revised version 2005 ISBN iv

6 CONTENTS Acknowledgements Plenary and featured speakers Foreword Introduction page ix x xi xii (Papers by alphabetical order of the surname of the first author except the last three) Hashan Al-Ajmi and Lafi Al-Harbi, The treatment of greeting formulas in a bilingual 1 dictionary of Gulf Arabic Fatemeh Alipanahi and Parvini Sani Hossein, EFL/ESL students and ESL/EFL 7 websites Nicolas Arriaga Agrelo, 猫匿 māo nì ; aquí hay gato encerrado : a cross-cultural 21 view on Chinese and Spanish idioms around animals Somsukla Banerjee, Achla M. Rai and Harish Karnick, Lexical organisation and access 22 of code-mixed collocations Paul Bruthiaux, Asian Englishes or English in Asia? Kachru s circles and the 27 sociolinguistics of othering Susan Butler, Words in Asian cultural contexts 28 Louisa Chan, Japanese influence on the lexicon of Hong Kong Chinese 36 Vincent Tao-Hsun Chang, The pragmatics of foreign words in discourse of Chinese 40 popular literature Yin Ling Cheung, A contrastive study of bare nouns and [classifier+noun] phrases in 41 Hokkien and Cantonese Huang Hoon Chng and Peter K W Tan, Changing patterns and attitudes: naming 45 practices among Malay Singaporeans Hochol Choe, Janggeun Oh and Donghyeok Lee, Alignment of the idiomatic 46 expressions in Korean lexical semantic net for connecting with EuroWordNet Gilles-Maurice de Schryver and David Joffe, One database, many dictionaries 54 varying co(n)text with the dictionary application Tshwanelex Jennifer Eagleton, Pressing names creating meaning in a fictional dictionary 60 Charles J. Fillmore, Verb-to-noun argument-sharing 66 Lily Wong Fillmore, Second-language learners and academic language: how can 67 teachers and dictionaries help? Gwyneth Fox, Macmillan English dictionaries new dictionaries for a new era 68 Gwyneth Fox, Metaphor in learners dictionaries 68 Tetsuro Fujii, Learner perspective for effective dictionary skill training and 69 lexicography Wengao Gong, Computer-mediated language: a new challenge for dictionary-making? 70 Rufus H. Gouws, The selection, presentation and treatment of cultural phrases in a 75 multicultural dictionary Jack Halpern, Lexicon-based orthographic disambiguation in CJK intelligent 75 information retrieval v

7 Reinhard Hartmann, Interlingual lexicography, with special reference to research 76 priorities (abstract; full paper on pp Jisheng He, Some interesting lexical features of Singapore Colloquial English 77 Shu-Kai Hsieh, A character-driven three-layers network model of meaning and 77 conceptual representation Da-Fu Huang, Dictionary use strategies by EFL learners in Taiwan: a re-examination 81 of the system of foreign language learning strategies Ai Inoue, Go and do, go to do, and go do a corpus-based phraseological analysis 86 Shin ichiro Ishikawa, Lexical features of the English composition by Japanese EFL 90 learners Yuka Ishikawa, Usage of cultural keywords seen in English dictionaries and corpora 91 Mitsuhiko Ito, Brythonic loanwords in English and the receptive knowledge 91 maintained by native speakers of English Gregory James, Designing a pedagogical dictionary of English academic vocabulary 96 for Chinese speakers Tse-Chun Jao and Shu-Chen Chiu, The changes in second language attrition lexicon 97 and syntax Girish Jha, The socio-psychological motivation for code-switching/mixing: the 103 Maithili-Hindi situation (abstract; full paper on pp ) Girish Jha, Sudhir Mishra, and R. Chandrashekhar, Information technology 104 applications for Sanskrit lexicography: the case of Amarakosha David Joffe and Gilles-Maurice de Schryver, Representing and describing words 108 flexibly with the dictionary application Tshwanelex Terry Joyce, Lexical association network maps for basic Japanese vocabulary 114 Beom-Mo Kang and Jae-Eun Cha, Multi-lingual(KJCE) lexical database using ILI 121 Takashi Kanazashi, Improving English-Japanese dictionaries, with particular reference 126 to systems of labelling Madoka Kawano, Effective ways of using native language in English classrooms: 131 suggestions for non-native teachers Ilan Kernerman, Dictionaries and translation, from art to science? 134 Adam Kilgarriff, Linking dictionary and corpus 135 Adam Kilgarriff, Chu-Ren Huang, Pavel Rychly, Simon Smith and David Tugwell, 142 Chinese word sketches Adam Kilgarriff, Michael Rundell and Elaine Ui Dhonnchadha, Corpus creation for 143 lexicography Gerry Knowles and Zuraidah Mohd Don, Grammatical class, tags and lemmas: a 149 corpus-based study of the Malay lexicon Yuri Komuro, Theoretical aspects of the compilation of the Macmillan Essential 149 Dictionary Workbook Montatip Krishmra, And the tiger cries, animal metaphorical expressions in the Thai 150 language Jacqueline Kam-Mei Lam, Sue Chang, Gregory James and Bronson So, Lay and lie 151 where should they go? A practical hands-on workshop on lemmatizing English words? Jacqueline Lam, Lan Li and Tom McArthur, A three-way dictionary: bringing 151 Cantonese, English and Putonghua together vi

8 Batia Laufer and Tamar Levitzky-Aviad, Towards a bilingualized electronic dictionary 168 for second language production Cynthia Lee, Teaching language use and communication skills: a bilingual approach 173 Keedong Lee, An analysis of polysemous verbs: a cognitive view 174 Saihong Li and Xiuying Wang, Survey of pragmatic information in bilingual English- 174 Chinese dictionaries Ee Ling Low and Siew Siew Lim, Do triphthongs exist in words spoken by young 180 educated Singaporeans? Calvin McMillin, Exit Charlie Chan, enter Allen Choice: an Asian American detective 181 for the 21st century Alexei Medvedev, Compiling a Russian-English lexicon: from terms to culture 186 Julia Miller, English learner s dictionaries: an undervalued resource 192 Andrew Moody and Azirah Hashim, Asian magazine advertising: database compilation 198 and avenues for future research Elizabeth Morales-Nuncio, Adjectives in other wor(l)ds: the uses and functions of 198 adjectives in sociocultural context Minoru Moriguchi, Students attitude toward dictionary use 203 Dave Moskovitz, Matapuna the open source dictionary writing system 204 Kyoko Motobayashi, Tomoko Takeda and Satoko Tokumaru, Strategies used by 210 international participants in discussions in an electronic bulletin board Quynh Phu Pham Na, Translations of topic-comment structures of Vietnamese into 214 English Nakane Sadayuki, Multiple word class entries in advanced learner s dictionaries of 221 English Rhoderick Nuncio, Language rationalization of technical words in Asian ICT 227 perspectives Shinya Ozawa and James Ronald, Electronic dictionaries in the classroom 233 Sang-Jun Park and Marc Duval, Words, meaning, speech acts and vocabulary in Far 239 East cultural contexts Tommaso Pellin, The introduction of the lexicon of grammar into China: typologies of 244 neologisms and historical background of their creation Richard Powell, Corpus planning and language shift in Malaysian law 249 Saravan Raja, The microstructure for a dialect dictionary of agricultural vocabulary in 250 Tamil Helen Riha, Roman letter words in Mandarin Chinese 255 James Ronald and Akira Tajino, A comparison of paper and electronic monolingual 255 dictionaries: location, comprehension and retention of secondary senses Kensei Sugayama, A lexical semantics approach to designing a more user-friendly EFL 261 dictionary Mats-Peter Sundström, Intermediate languages in bilingual lexicography for lesserused 267 languages: the pros and cons Monika Szirmai, The benefits of CD-Rom dictionaries in teaching 273 Ismail S Talib, Narrative and the multimodal contexts of lexicography 279 Loretta Chung-Wing Tam, Trick or chic: code-switching in colloquial Taiwanese 279 mandarin revisited Siew Imm Tan, The Malaysian English newspaper corpus: construction and function 284 vii

9 Ying Ying Tan, The Malayness of the Teochew lexicon 290 Eunice Tang, What words do Hong Kong students know? : a study of lexical input 290 from textbooks Yuanke (Yorke) Tao, Nominal, verbal and adjectival structures in Japanese and 296 Turkish Keiko Terashima and Minoru Moriguchi, The first collocation dictionary for the 303 Japanese language Kingkarn Thepkanjana and Satoshi Uehara, The verb of breaking and its semantic 308 variations in context: a contrastive study of Thai and Japanese Eric C. Thompson, Nation-states as a semantic domain 315 Yukio Tono, Exploring the potential of learner corpora for pedagogical lexicography 315 Yukio Tono, Miwa Kamura and Shinichi Yagi, Pocket e-dictionaries in Japan: new 321 developments Touno Tsuyako, The gap between the importance level indications in English-Japanese 322 (E-J) dictionaries and the vocabulary used in English textbooks Hugo T. Y. Tseng, Reverse indexing and customization future trends in bilingualized 327 dictionaries Anneke Tupan, The implicit meaning of written advertisements 332 Uemura Toshihiko, How good are graded English readers for ESL/EFL students? 338 Valluvan Sivagnanam, Genesis and function of words with astronomical and religious 344 implications in Indus Valley civilization Tamara S. Wagner, Occidentalism in Singapore s Anglophone fiction: Victorian 349 wor(l)ds lost in C. M. Woon s The Advocate s Devil Shengyu Wang, Multilingualism and the alienation effect in Kuo Pao Kun s theatre 354 Jonathan Webster and Ian C. Chow, Relational network notation and the intelligent 361 web Hai Xu, Word combinatory information represented in illustrative examples in English 367 learners dictionaries Xiaoyu Xu and Xue ai Zhao, Transform: from the product approach to the processoriented approach in online English class 374 Shigeru Yamada, Students preferences of web-based EFL dictionaries 379 Yihua Zhang, Lexical concerns about neologism in Chinese lexicography 380 R. R. K. Hartmann, Interlingual lexicography, with special reference to research 393 priorities Girish Nath Jha, Socio-psychological motivation for code-switching/mixing: the 401 Maithili-Hindi situation Miwa Nakamura, Misinterpretation analyses of data from 2004 pop song translating 406 competition comparing electronic handheld dictionaries and printed dictionaries users Jan-Olof Svantesson, Cultural information in a bilingual dictionary 411 viii

10 ACKNOWLEDGEMENTS We are grateful for all the support and help received from the following individuals and groups. Tan Tai Yong, Dean of the Faculty of Arts and Social Sciences Robbie Goh, Head of the Department of English Language and Literature Anthony Reid, Director of the Asia Research Institute The United States Embassy Macmillan Publishers Ilan Kernerman and K Dictionaries Pte Ltd Staff of the Faculty of Arts and Social Sciences, staff of the Department of English Language and Literature and Staff of the Asia Research Institute in the National University of Singapore All our student helpers, in particular Soe Marlar Lwin and Gong Wengao All our IT support staff, in particular Neo Sian Hu, John Riordan Yap and Kwok Wai Tat All our exhibitors ix

11 PLENARY SPEAKERS Charles Fillmore Department of Linguistics University of California Berkeley & International Computer Science Institute Berkeley Reinhard Hartmann School of English University of Exeter & University of Birmingham Yihua Zhang Guangdong University of Foreign Studies & Chinalex Bilingual Commission FEATURED SPEAKERS Susan Butler Macquarie Dictionary Lily Wong Fillmore University of California, Berkeley Gwyneth Fox Macmillan Dictionaries Gregory James Hong Kong University of Science and Technology Ismail Talib National University of Singapore Yukio Tono Meikai University x

12 FOREWORD ASIALEX 2005 Singapore with its theme of Words in Asian Cultural Contexts is coorganised by the Department of English Language and Literature, Faculty of Arts and Social Sciences and the Asia Research Institute, National University of Singapore (NUS). One of a hundred events happening during the year-long NUS Centennial Celebrations, this conference has attracted participation from almost every major Asian country. Words are the primary locus of study in several disciplines, and this conference with paper and workshop presentations from almost a hundred participants over a threeday period is of interest to lexicographers, lexicologists, educationists, linguists, philologists, cultural geographers, cultural theorists, writers and literary scholars, and those who have an interest in words in diverse Asian contexts. The examination of the mental lexicon and its relation to print and electronic dictionaries can be generally characterised as one that involves relating word knowledge to world knowledge. The study of words alone, however, may not suffice and often leads to the wider question of how individuals relate to their place in society, define and occupy cultural space and construct it through language. The theme of Words in Asian Cultural Contexts is thus followed through in four major strands: lexicography and lexicology; sociolinguistics and language pedagogy; literary, cultural and postcolonial studies; and information and communications technology. Papers that concern lexicographic principles and practices have been selected as well as those that refer to broadly-defined notions of Asia and words in diverse cultural settings. We hope that these perspectives will shed light on lexicographic, linguistic and cultural perspectives, as well as areas such as education, literature and the arts. I am deeply indebted to the Organising Committee comprising Peter Tan (Vice- Chair), Vincent Ooi (Treasurer), and Ismail Talib, Lynn Tan and Tan Ying Ying (Secretariat). Acknowledgements are also due to the head and staff of the Department of English Language and Literature, the dean and staff of the Faculty of Arts and Social Sciences, and the director and staff of the Asia Research Institute for their support and encouragement of this Fourth Biennial Conference of ASIALEX. I thank the international Executive Committee of ASIALEX whose invaluable support is much appreciated, as also the sponsors and supporters of this event. We look forward to the ASIALEX 2007 conference, the Fifth Biennial Conference, which will be held in India. Last but not least, I thank the plenary and featured speakers and the participants at this Conference, wishing all a fruitful and engaging pursuit. Anne Pakir Department of English Language and Literature National University of Singapore ASIALEX President, Chair, Organising Committee, ASIALEX 2005 xi

13 INTRODUCTION This volume of papers is appropriately entitled Words in Asian Cultural Contexts, in keeping with the objective of the Asian Association for Lexicography (ASIALEX) to promote the study of words in such contexts. Alongside counterparts from our sister associations EURALEX (Europe), AFRILEX (Africa), and AUSTRALEX (Australasia) ASIALEX scholars are invariably interested in the lexicon as the central repository of language, and of its forms, functions and representations. It is therefore not surprising to find the lexicon reified in such strands as the following: lexicology and lexicography; sociolinguistics and language pedagogy; literary, cultural, and postcolonial studies; and information and communications technology. While the accompanying programme leaflet to this volume of papers conveniently colour-codes these strands (blue for lexicology and lexicography; yellow- sociolinguistics and language pedagogy; green for literary, cultural and postcolonial studies; and purple for information and communications technology), a paper classified under a particular strand (and often proposed by the paper presenter as such) can arguably be said to belong to another. Such is the nature of multi-disciplinary knowledge, in which such perspectives are never mutually exclusive but often reinforce one another. The lexicology and lexicography strand understandably comprises the largest number of papers, with rich offerings on the theory and practice of the lexicon including innovative practices to suit the dictionary to different contexts. The sociolinguistics and pedagogy strand takes the wider study of the lexicon to the speech community, and stresses the importance of the dictionary to the language learner in her or his desire to be a full-fledged speaker of the community. The literary, cultural, and postcolonial studies strand sheds light on how words take on a larger significance in such contexts, including historical and sociological ones. The information and communications technology strand takes cognisance of the fact that we live in a world in which the computer is often indispensable, and discusses issues that relate to the computerisation of the lexicon. These strands collectively foster a deeper understanding of the nature of language and enrich the fabric of society. At the time of going to print, this volume reflects a combination of abstracts and full-length papers. Not all paper presenters given their busy schedules - had the time to write full-length papers before the conference proper. We trust that this volume complements their respective presentations at the conference, while serving as a very useful read on its own. The Editors xii

14 WORDS IN ASIAN CULTURAL CONTEXTS The treatment of greeting formulas in a bilingual dictionary of Gulf Arabic Hashan Al-Ajmi Kuwait University hashan98@yahoo.com Lafi Al-Harbi Kuwait University lalharbi@hotmail.com Abstract Although it is spoken by more than ten million people in an area stretching from southern Iraq to Oman on the Arabian Sea, Gulf Arabic is one the most under-researched varieties of Arabic. A few bilingual dictionaries with English as the target or source language have been published in recent years all of which try to provide more or less accurate versions of the linguistic reality of Gulf Arabic. This study examines the treatment of greetings in the Glossary of Gulf Arabic (1996), a bidirectional bilingual dictionary intended for English-speaking users. The aim is to highlight, from discoursal and translational perspectives, the problems which compilers of bilingual dictionaries of Arabic dialects encounter when trying to represent conversational phrases such as greetings. The analysis focuses on five categories of greetings commonly used by Gulf Arabs: meeting another person, asking about health and state of affairs, time of day, requesting and thanking, and closings and farewells. For the sake of brevity, only the first two categories will be discussed in this paper. Introduction The Gulf region has captured the world s attention for its economic and strategic importance. An interest in the region s language becomes inevitable; consequently, this interest has materialized in considerable awareness and urgent appeal for an adequate description of Gulf Arabic. Bilingual dictionaries (Gulf Arabic/English) emerged as a sign of the growing attention being paid to Gulf Arabic. Some bilingual dictionaries serve the needs for businessmen and representatives of commercial establishments who may come to the area on short business trips or to reside there for relatively longer periods to work for international organizations or in some local development programs in the Gulf states: Saudi Arabia, Kuwait, Qatar, Bahrain, the United Arab Emirates and Oman. Among the initial phrases that newcomers wish to be able to use effectively will be greeting phrases and forms of politeness, particularly in the Gulf culture which is very dissimilar to the English culture. In order for the newcomers to communicate effectively, particularly in business and diplomacy, they need to be able to achieve the desirable impression on their interlocutors. The literary meaning of greetings and phrases of politeness is highly contextualized, thus it may not be exactly identical to the meaning presented in the standard form of Arabic. Hence, foreign users of Gulf Arabic need to turn to a specialized dictionary, as a truly ethnographic book of reference for cultural expressions and information. However, dictionaries are generally deficient in this particular area of culture-specific vocabulary, such as greetings and phrases of politeness (e.g. Fatani 1998). Some leading lexicographers have emphasized that such vocabulary requires explanation (Landau 1989:8). 1

15 ASIALEX 2005 Dictionaries of Gulf Arabic About ten million people speak this Arabic dialect in the area along the Gulf coast including southern Iraq, Kuwait, Saudi Arabia s eastern province, Qatar, Bahrain, the UAE and Oman. Some scholars also include the dialect spoken in central Iraq although they distinguish between the Iraqi dialect and Gulf Arabic with its Bedouin origins (Abu Haider 1994). A number of bilingual works engaging Gulf Arabic have been produced in recent years. The list includes Clive Holes Colloquial Arabic of the Gulf and Saudi Arabia (1988), Hamdi A., Qafisheh s Glossary of Gulf Arabic (1996) and NTC s Gulf Arabic-English Dictionary (1999). This study will deal specifically with the Glossary of Gulf Arabic (1996) to see how it treated greetings and phrases of politeness in this Arabic dialect from a lexicographic and cultural perspective. The compiler, an Arabic speaker but not a native of the Gulf area, claims that the dictionary contains words and phrases commonly used by native speakers of Gulf Arabic in the UAE, Qatar and Bahrain. He indicates that the dictionary fills an important linguistic gap in modern linguistic studies on peninsular Arabic and that it will serve the needs of English-speaking learners of Arabic as well as Arabists and dialectologists. The dictionary consists of two parts, Gulf Arabic-English and English-Gulf Arabic. Our concern here is the English-Gulf Arabic part which is used productively by English speakers to find Arabic equivalents for common phrases in conversations. Therefore, the focus will be on the English-Gulf Arabic part of the Glossary of Gulf Arabic to examine the treatment of greetings and phrases of politeness. The study aims to find out how well the dictionary provides a) a comprehensive description of these expressions showing cultural and discoursal values of these expressions, b) accurate translations of these expressions, and c) enough assistance to locate these in the wordlist. Greetings from the Gulf and the wider Arab World Gulf Arabs use fixed greetings and responses in their daily encounters and conversations. From a westerner s point of view, these rituals may take a relatively long time as interlocutors may exchange the same greetings and use a variety of responses. Involved here are questions about health and state of affairs as a way of establishing social communication and friendship. Contrastive studies of greetings emphasize their importance in understanding the culture of another language and that both native speakers and learners of a language should learn how to greet appropriately (e.g. Schliecher 1997). According to Duranti (1997), greetings and farewells perform a number of functions in social communication: to establish, prolong or stop communication. They are very essential elements in a conversation with a Gulf Arab, who always appreciates the use of Arabic phrases of politeness by foreign visitors, especially in business and diplomatic exchanges. Some of these Arabic expressions have global religious dimensions and are found in the languages of the Islamic world e.g. as-salamu alaikum, alhamdu lil-lah, insha Allah, etc. But many greetings in Gulf Arabic are culture-specific and are not found even in other Arabic dialects, e.g guwwa may God give you strength. 2

16 WORDS IN ASIAN CULTURAL CONTEXTS Greeting exchanges have been studied systematically to determine their underlying principles of social interaction. Arabic greetings are rarely mentioned in such literature. A review of the few studies on the topic is necessary to highlight the essential features of Arabic greetings in general and Gulf greetings in particular. This is because we believe that a dictionary of this and other Arabic dialects should provide a realistic and informative account for these expressions for the benefit of the foreign user. One study involving Arabic greetings (Ferguson 1978) highlighted some cross-cultural contrasts between Syrian Arabic and American English. Ferguson examined the root-echo responses formulas that characterize Arabic politeness expressions. Many differences in detail were noticed e.g. in the good morning greeting. He found that the structure and use of greetings are so culture specific and tied to the cultural history. Greetings in the Mecca speech community have been analyzed and classified by Hassanain (1994) who provided several examples of Saudi social expressions used in a variety of situations ranging from hello and leave taking to condoling and speaking of a dead person. After a quick comparison, we found that the majority of expressions in his list are common in the speech of Gulf people (39 out of 57 expressions, i.e. 68.4%) while the rest were still intelligible to speakers of Gulf Arabic. The religious nature prevails in these Meccan greetings especially in the responses where the greetee does not thank the greeter, but rather thanking God for the state of his own health (p.72). A variety of Gulf Arabic has been investigated by Emery (2000) who focused on three categories of politeness formulas in Omani Arabic: greeting and parting routines, congratulating and condoling. His first observation was the considerable overlap with Classical and pan-arabic usage although Omani Arabic had its own linguistic routines and patterns. He found that older people tend to use more formulas than young ones and are perceived as linguistically conservative by the younger generation. He identified three stages of greetings: a summons salam, health enquiry (with repetition and variation on the same theme), and ritualized news enquiry and the reply must always convey optimism or reassurance. From a lexicographic perspective, the present study will deal with the most common greeting expressions in the Gulf arranged in separate categories according to their possible order and contexts: meeting another person and asking about health. For each set of formulas a table shows the most common greeting exchanges, their literal and dictionary translations, the part of the dictionary in which they are listed, their usual and dictionary responses and their dictionary explanations. Greetings used to initiate an encounter In addition to the usual assalam alaikum greeting there are certain expressions used by Gulf people for opening a conversational exchange. Sometimes these are used to draw attention such as marhaba and allah bilxeer. Like their Classical Arabic source, Gulf Arabic and other Arabic dialects have highly polysemous lexical items including politeness expressions. For example, it has been found that there are so many contexts in Egyptian Arabic exchanges where inshalla is used (Gregory & Wehba 1986). Therefore, it is important to revise the treatment of these greetings in terms of the number of senses described. Most of the greetings examined in this study are allocated monosemous entries in the dictionary, thus providing the user with one context of usage among others. The same applies to the responses which may vary. For example, it is 3

17 ASIALEX 2005 possible to use hala as a response in all of the situations for the four greetings in this group. Not all Gulf greetings are specific to the speech community of the region. In this set of greetings marhaba and hala are also common in other areas of the Arab world while guwwa and alla bilxeer stand out as strictly Gulf-specific. Of course, dialectologists and ordinary users would benefit from such a distinction in the entries concerned. Most Gulf greetings are weakened forms from Classical Arabic but many learners may not see the connection. Thus the dictionary should make this clear to the user by providing the original Classical word or phrase. For example, hala was originally ahlan people, a shortened form of the Classical Arabic greeting ahlan wa sahlan people and plain, i.e. you are among your people and on your land! Of this group of greetings the dictionary gives a few of the possible responses but the inflated responses are ignored as in the case of marhabteen two welcomes and maraahib several welcomes, the latter showing pan-arabic influence (Emery 2000). The dictionary does not treat these greetings consistently in terms of the part of dictionary where they are listed. The greeting marhaba is listed in both parts while guwwa is found in the Gulf Arabic-English part only. The former is given the English translation hello / hi! but without a literal meaning. This reliance on functional translation may conceal the exact meaning of the linguistic sign and may result in an entirely different English greeting. Guwwa is given the English version how are you? which functions, to some western users, as a request of information about health. The dictionary further confuses this user by giving I m fine as the translation of the response alla ygawwiik, literally may God give you strength (see table 1). This kind of entry may lead some learners to use the greeting clumsily e.g. in a situation where one is actually enquiring about health. It is therefore necessary to have a special label for these politeness phrases similar to idiom labels as they are both fixed units and usually convey non-literal meanings. Another point concerns the accuracy of transliterations which depart from the local norms of pronunciation. Some Gulf greetings like halo biik indicate interference with the compiler s own Arabic dialect and emphasize the need to consult with natives for a more detailed and accurate description. Table 1. Greetings used to initiate an encounter Expression: marhaba guwwa alla bil-xeer hala Literal translation A welcome Strength! By God do the good! People Dictionary Hello, Hi How are you? Good morning! equivalent Dictionary part GA-E + E-GA GA-E GA-E (under alla) Response marhaba a welcome / marhabteen two welcomes /maraahib alla ygawwiik may God strengthen you /hala people alla bilxeeraat By God do the good deeds /alla bilxeer By God do the good /hala people Hala people /ahl een two peoples welcomes /hala people / ahlan wa sahlan people and plain Dictionary response marhaba/ ahlan wa sahlan alla ygawwiik The response is the same Dictionary Explanation It is said at any time of the day; characteristic of Iraqi and Kuwaiti) 4

18 WORDS IN ASIAN CULTURAL CONTEXTS Asking about health and state of affairs This phase takes relatively longer time than other types of greeting and involves the use of a much wider variety of expressions and possible responses. However, the dictionary fails to provide the usual responses except in the case of the phrase shi luumak where the compiler gives a negative response ma 9indi 9luum I have no news instead of the usual formulaic response luum xeer information of goodness. The dictionary lists only five out of six greetings of this type with the first two in both parts while the rest are listed in either part of the dictionary (see table 2). Even the English adverb hopefully is the entry where one greeting is given as an equivalent of I hope you are well, fine making it hard for the user to locate such expressions in a productive mode. The translations of health greetings show a tendency to concentrate on the function of the phrase i.e. to enquire about health. This explains why three out of the five health greetings have been given the English phrase how are you? as translation. As a result, the user will not have access to the exact meaning of the Arabic expression and some cultural and etymological information will be missing. For example, shluunak is explained in the dictionary as being a contraction of shloon or shu loon how? (see table 2), but it literally means What is your color?. It is a greeting form that appeared in the 19th century when an infectious fever that caused one s face to turn yellow spread throughout the Gulf region and people used the question as a way of ensuring that the addressee was not affected by the disease. Table 2. Asking about health and state of affairs Expression: keef Haalak shloonak asaak tayyib/ibxeer shaxbaarik shi loomak keef essihha Literal translation Dictionary translation Dictionary part Response Dictionary response Dictionary Explanation How is your state? What is your color? How are you? E-GA+GA- E Bxeer in goodness / tayyib well / lhamdilla praise be to Allah / bahsan Haal in the best state How are you? E-GA+GA- E Bxeer in goodness / tayyib well / lhamdilla praise be to Allah Contraction of sh-loon or shu loon: how? Hope you are well/in goodness? I hope you are well, fine! What are your news? What is your information? How is the health? What news How are do you have? you? E-GA GA-E E-GA tayyib asaak tayyib well, hoping you are well / bxeer asaak bxeer in goodness, hoping you are in goodness axbaar xeer news of goodness / lhamdilla praise be to Allah luum xeer information of goodness Ma indi luum I have no news Bxeer in goodness / lhamdilla praise be to Allah 5

19 ASIALEX 2005 Conclusions and proposals: It is clear from the above analysis that the Glossary of Gulf Arabic and similar dictionaries must accord special attention to these expressions of politeness, which are usually sought by people when first learning another language or dialect. These words and phrases mostly do not have referential meanings and the learner is required to know the appropriate responses. We have found that there exists a great deal of inconsistency in the provision of responses in the Glossary of Gulf Arabic. Only 35.7% of the examined greetings are followed by their responses in the dictionary. Also, explanations that indicate the functions and situations are rare. As for the translation of greetings, we noticed a tendency to provide functional translations, i.e. giving the closest English cultural equivalents despite the differences in communicative functions. It is also necessary to provide the literal translation for each greeting so users can recognize the exact meaning of the Arabic phrase. In some cases there is need to refer to the etymology of some greetings, e.g. eshloonik, which literally means what is your color?. Locating Gulf greetings in the dictionary poses great difficulty for the English-speaking user as the dictionary does not follow the root-based system consistently. Words and phrases are entered in many cases according to the English alphabetical order in the Gulf Arabic-English part. In addition, the organization of entries in the English-Gulf Arabic part was not helpful either. For example, the greeting asaak tayyib is listed under the entry word hopefully although tayyib well is the main part of the greeting. It seems that the differences between the two languages have forced the compiler to list most of these expressions in the Gulf Arabic-English section (71.4%). Another problem is related to the microstructure of the dictionary where we can find two synonymous greetings sabaah elxeer and sabbahk allah bilxeer in the same entry but located five lines away from each other. The dictionary entries that we examined show a bias in the dictionary in favor of greeting forms often used by the elderly, which emphasizes the importance of data collection of material for such a dictionary i.e. a lexicographer should make sure that different age groups are covered. Surely, the dictionary user will benefit from some sort of indication at the entry as to whether the greeting in question is likely to be used for an elderly or young person. It may be more beneficial to provide a list of these greetings in their usual sequence and adjacency pairs along with their possible responses in a special section in such a dictionary with cross-references to the relevant entries in both parts. This will improve accessibility of the dictionary especially for a user interested in further information such as example sentences, word roots or etymology. Finally, the treatment of greetings in dictionaries of Gulf Arabic requires comprehensiveness, accurate translation and improved accessibility. This can be achieved by reliance on a team of editors who are natives of the Gulf and include specialists in linguistics, translation and lexicography. References A. Dictionaries Holes, C. (1988), Colloquial Arabic of the Gulf and Saudi Arabia (Brunner-Routledge). Qafisheh, H. (1996), Glossary of Gulf Arabic (Beirut: Librairie du Liban Publishers). Qafisheh, H. et al. (1999), NTC s Gulf Arabic-English Dictionary Ntc Publishing Group. 6

20 WORDS IN ASIAN CULTURAL CONTEXTS B. Other references Abu-Haidar, F. (1994), Iraq: language situation, in: R. E. Asher (ed.), The Encyclopedia of Language and Linguistics 4:1774 (Oxford: Pergamon Press). Emery, P. (2000), Greeting, congratulating and commiserating in Omani Arabic, Language, Culture and Curriculum 13(2): Fatani, A. (1998), The treatment of culture-specific vocabulary in dictionaries for translating from English to Arabic: a critical and empirical exploration, PhD thesis (Exeter: University of Exeter). Ferguson, C. (1978), The structure and use of politeness formulas, Language in Society 5: Gregory, S. and Wehba, K. (1986), The contexts of inshallah in Alexandria Egypt, Anthropological Linguistics 28(1): Hassanain, K. (1994), Saudi mode of greeting rituals: their implications for teaching and learning English, International Review of Applied Linguistics 32(1): Landau, S. (2000), Dictionaries. the art and craft of lexicography (Second edition) (New York: S. Scribner & Cambridge: Cambridge University Press). Schleicher, A. (1997), Using greetings to teach cultural understanding, Modern Language Journal 81(3): Zughoul, M. and Abdul Fattah, H. (2002), Temporal expression in English and Arabic, Languages and Linguistics 1(4): EFL/ESL Students and ESL/EFL Websites Fatemeh Alipanahi Islamic Azad University of Zanjan; Zanjan University; IASBS, Iran alipanahi@hotmail.com Abstract There are different EFL websites for language teachers (Opalka, 2002). Reasons for using the websites in ESL learning include their authentic language; leading to increased motivation, interaction and creativity; and encouraging learner centeredness, the discovery method of learning and learner independence. Nevertheless, teachers contribution remain important for coordinating group planning, focusing learners attention on linguistic aspects of computer-mediated texts, in helping students in developing individual learning strategies, and creating an appropriate atmosphere for language learning. To fully exploit these opportunities, the teacher must learn to become a guide on the side rather than a sage on the stage (Warschauer, 1997.) This study investigates the potential role of Hyperextend ESL/EFL websites in EFL Learning and evaluates a work in which 24 low-advanced students in an EFL class were introduced to some hyperextend EFL websites and instructed to use them in their reading classes as well as self-study. Collected data showed that in spite of some difficulties students had a positive attitude to using the chosen hyperextend websites in their EFL learning. The students found that learning and reading English through ESL/EFL websites was not only interesting but attractive and creative also. 7

21 ASIALEX 2005 Introduction Technology, especially the Internet has become indispensable in the contemporary world as the powerful means for communication and education. Learners interest to learning languages has been enforced by the availability of the Internet, which provides easy access to every possible kind of information and an effective tool to facilitate learning. The increasing emphasis and promotion of autonomous learning of foreign languages and the widening role of online technologies have lately become the major features in language teaching. The question today is how to harness electronic technology and guide our students in their study and use. There are different EFL websites for language teachers (Opalka, 2002). Authentic language, increased motivation, interaction, creativity, student centeredness, discovery method of learning and working independently can be the reasons for using the websites in ESL learning. However, we should be aware of the problems such as accuracy, appropriateness, and appeal (easy to use, interesting to read) and reliability of the information on the web when applying websites in the classrooms (Opalka, 2002). Nevertheless, teachers contribution remain important for coordinating group planning, focusing learners attention on linguistic aspects of computer mediated texts, helping students in developing individual learning strategies, and creating an appropriate atmosphere for language learning. To fully exploit these opportunities, the teacher must learn to become a guide on the side rather than a sage on the stage (Warschauer, 1997). What is Hypertext? V. Bush originally elaborated the concept of hypertext. He foresaw the need for a major effort to mechanize the scientific literature system. He felt the need for a system that would help people find information more easily than was possible by using print technology. It was much later and particularly with the development of the microcomputer that hypertext came into relatively common use. Conklin s hypertext survey appeared in 1987 and described no fewer than 8 systems. Essentially a hypertext system is a computer-based approach to information management in which data are stored in networks of nodes connected by links. (Locatis, Letouneau, and Banvartd 1989). Harvey (1991) characterizes hypertext as an integrated tripartite application: a data base method providing a novel way of accessing data, a representation schema, that is a kind of semantic network: and an interface modality featuring link icons which can be embedded anywhere in the material by the user. Scrimshaw (1993) sees it as an open-ended set of topics connected by variable links. For Nelson (1987) hypertext is non-linear text. Whatever the differences between systems, they all allow information to be accessed non-sequentially through machine supported links and in terms reminding of Bush s, are said to mimic the associative properties of the mind (Fiderio 1988, Jonassen 1991). Advantages of hypertext When using hypertext the students have a more active role than is possible with traditional textbooks. They are required to make decisions about the information they are accessing through the computer. Students using hypertext must be mentally active. Consequently, learning becomes more student-centered. 8

22 WORDS IN ASIAN CULTURAL CONTEXTS Hypertext learning system introduces an exploratory or discovery method of learning into classroom. Students become active learners. Simultaneously, students not only become more mentally active, they also interact with new ways of presenting information through computer-based technologies. Reasons for using the web in the FL classroom By now, it is evident that the web is here to stay and that it will influence society. It is impossible to predict exactly what effects the web will have in the future, but based on the experiences so far, it would be reasonable to conclude that it will have great significance. General reasons The world is becoming an information world in which Information Communication Technology (ICT) plays an important role. In the information world, there is a need for understanding of the technology. New technologies should be assimilated into the curriculum and into teaching methods. Thus, universities will have to prepare their students for the information world and make effective use of ICT to provide better education (Van Assche, 1998). Van Assche (1998) points out several benefits of using the web: Universities will be able to cope better with the challenges of preparing the students for the information society. By ICT education will change from being told to exploration with a much deeper understanding. Tolerance and respect for other cultures will be formed in universities. Finally, the use of the web will link the abstraction of teaching material with real life in the outside world. When examining the FL teaching goals, one discovers that the use of the web can be useful in attaining these objectives. Students will understand foreign cultures by having access to vast resources of authentic material about them. The ultimate goal of turning students into life-long learners by using the language for personal enjoyment and enrichment becomes much more realistic when students continue to use the web. The web can play an important role in EEL learning. It is all about communicating and offers possibilities that would never be feasible in a traditional classroom setting. Furthermore, the use of the web supports the shift from the teacher-centered classroom to one in which the student is in the center. Using computers and talking live to others motivate students. Reading Comprehension Reading is defined by scholars like Allen & Corder (1979) (cited in Grabe, 1997, page 7) as the activity of reconstructing a reasonable, spoken message from written symbols to a form of language to which the person can already attach meaning. Goodman (1988), on the other hand, defines reading as follows: Reading is a receptive language process. It is a psycholinguistic process in that it starts with a linguistic surface representation encoded by a writer and ends with meaning, which the reader constructs (P. 12) 9

23 ASIALEX 2005 Widdowson (1985) believes that reading should be regarded as interaction between writer and reader, mediated through the text. It has been confirmed by Carrell, Devine, and Eskey (1988) that reading is no longer a process of decoding but rather an integration of top-down processes that utilize background knowledge and schema, as well as bottom-up processes that are primarily text or data-driven (cited in Richards, 1990, p. 87). Reading requires sufficient knowledge of language. Purpose of the study The degree to which the Internet is useful in Language learning especially in reading comprehension depends on how well the found materials match the students needs and their ability level. Despite of the availability of resources, they may not be useful unless the learners know how to use them to facilitate language learning and reading comprehension. In this regard, the writer of the paper has done this research to introduce students to use selected hyperextend EFL websites for their language learning. Therefore, the main purpose of this study is to evaluate these questions: Do the teacher introduced websites increase learner autonomy in EFL? Are teacher used guidelines considered to be useful in enhancing EFL learning through hyperextend web sites? What problems and difficulties did students have in using selected hyperextend web sites? Are the students willing to use ESL websites for other classes during their four years of education? Methodology Subjects 24 low advanced level students majoring in English in University were selected. They already had 5 years of formal education in English language at high school. The students were freshmen at the time of the research and were taking four obligatory English courses: Lab, Writing, Reading, and Structure. This project was implemented in their reading classes. Procedure The students were given two 90-minute sessions instruction per week at the English language department. The students worked in pairs on the computers at the learning sessions. The training session aimed to introduce students to the computer facilities and some websites that the learners used to practice their English and reading skills in class and on their own. Before the teaching session started, the students were asked to fill out a questionnaire about their computer access and usage experience, frequency of the Internet use, and their familiarity with EFL websites. Then, they were introduced to the following hyperextend EFL sites. Dictionaries: Beard, R. A Web of On-line Dictionaries URL Merriam Webster On-line URL 10

24 WORDS IN ASIAN CULTURAL CONTEXTS URL Sterling, D. Dave s ESL Café URL Pen Pal Connection URL Interesting Things for ESL Students Activities for EFL Students Interactive on line activities, Interesting things for EFL students Basic grammar, plenty of free materials Activities for students Longman online English courses website for learning English via news,. service/learningenglish After the training session, the students were instructed to read the passages and then were asked to do a series of homework activities presented through the websites. The activities included different exercises on reading comprehension and guessing the meaning of the vocabulary from context or checking the meanings of the new words by mouse clicking on the hyperextend words, which were linked to the correct definitions. At the end of the ten-week instruction, the students were asked to complete a questionnaire concerning their insights into using the web sites. Likert scale based on 4 points (from 1 for strongly disagree to 4 strongly agree ) was used. It included the following sections: The probability of creating autonomous EFL learning through web sites Insights into the provided teaching guidelines. Difficulties students had during this study; and The probability of using the websites for other classes during their four years of education in the university. Discussions and conclusions Computer ownership and locations for using the web Figure 1 indicates the computer ownership. It can be seen that 20% of the students stated having home computers. Figure 2 indicates the most frequent locations where the learners access the web. 80% of the students reported using university facilities. 18% indicated using coffee nets. Only 10% reported getting online at home and 4% indicated using library facilities. It can be seen from the data that the same students use different access to the web, which explains why the total percentage exceeds 100%. 11

25 ASIALEX 2005 Figure 1. Owning home computers 1 2 Figure 2. Access to the web Number of the students, % Series1 University Coffee net Home Library Time spent on line per week The 3rd figure indicates how frequently learners get online. 28% reported getting online everyday. 20% got online 4 times a week, 30%, and 22% got online 3 and twice a week respectively. Figure 3. Time spent on line per week the % Number of students, Series1 everyday 4 times 3 times twice 12

26 WORDS IN ASIAN CULTURAL CONTEXTS EFL web experience 80% indicated not having any experience of ESL web sites before (Figure 4). Regardless of their online experience most of the students were unfamiliar with EFL web sites and those familiar ones did not know how to use them. Figure 4. EFL web experience 1 2 Insights into reading 80% reported EFL readings in introduced sites were fun. 75% liked to read in English. 10% liked to read in Persian. 9% considered on line reading to be a waste of time. 11% reported on line reading to be boring. 60% reported time assigned f or reading classes was very short. 90% indicated the need for taking more web oriented reading courses. 85% stated that doing reading exercises online is interesting and useful (mean=2.75). 12% stated they could do without EFL reading web sites. 88% reported introduced web reading courses were less time consuming. 90% enjoyed reading comprehension through the web sites (mean=3.1). 92% were willing to use the sites more for online reading comprehension courses (mean=3.2) (Figure 5 (a &b)). Figure 5 (a). Insight into reading on line nomber of stidents, % Series1 Figure 5 (b) Mean We enjoyed reading comprehension through the web sites 3.1 Doing reading exercises online is interesting and useful 2.75 We are willing to use the sites and take more online reading comprehension courses

27 ASIALEX 2005 Students overall insight into using hyperextend websites to learn EFL Concerning the data collected from questionnaire, 89% of the students asked to be given more opportunities to learn English using those sites ( mean=2.89). 90% thought online learning should be included in English classes. 81% believed EFL can be more effective and interesting using web sites (mean=2.75). As table 6 indicates, learning English through hyperextend sites is considered to be not only necessary for EFL but also effective. In addition, the analysis of the results obtained from the questionnaire indicates the following: fully 98% of the students believed that the web helped them to improve their EFL ability and skills. 70% found that it was not difficult to learn to use the web for EFL reading. 90% of the student agreed that materials become more comprehensive by using web (mean=2.99). 89% of the students agreed that the web creates active learning and effective contact between students and teacher. 92% of the students agreed that the Internet helps them learn the subject matter on a deeper level, makes class fun, encourages monitoring, helps them think creatively and critically, and also helps them practice independent thinking. All the data above support the useful ness of the web in the EFL reading comprehension classroom and indicates that the application of the web in EFL reading classes not only helps EFL students but improves the students attitudes toward reading and leads to improved proficiency as well. Figure 6. Students insight into using the hyperextend websites to learn EFL Figure 6 We need more opportunities to learn English using web sites. Using the hyperextend sites is not only an effective way to learn English but also interesting Mean Materials become more comprehensive by using web 2.99 Students insight into creating autonomous learning through web sites Generally speaking, most of the students agreed on developing learner autonomy through the web sites. 87% of the students indicated that they felt not only comfortable but also empowered to take responsibility for their own learning when they used the web in EFL classes. 69% reported that web sites empowered them to be initiative and creative, and to work independently. (Figure 7). Figure 7 We felt more empowered to take responsibility for their own learning using the web in their EFL learning. Mean 2.89 We felt more comfortable using the web in their EFL classes 2.89 Web sites empowered us to be initiative, creative, and work independently

28 WORDS IN ASIAN CULTURAL CONTEXTS Students insight into used guidelines Collected data (figure 8) reveals 76% students felt that instructions were necessary for online navigation (mean=2.75), 89% considered the recommendation on useful sites to be necessary (mean = 2.99) and 90% needed an introduction to the contents of the chosen sites (M = 3.12). The results indicated that the students lacked experience using EFL resources on the web. In addition, most of them found the web to be (mean = 2.97) effective in doing the homework assignment and their learning of English through ESL Figure 8 The teacher s instructions for online navigation were necessary Mean 2.75 The teacher s recommendation of websites was necessary 2.99 The teacher s introduction to the content of the websites was necessary The web is effective in doing the reading comprehension assignment and EFL learning The probability of using the web for other classes during their four years of education in the university. 90 % of the students indicated that they will use websites during their four years of education in EFL (mean =3.12). 88% stated they will even look for more ESL websites themselves for self-learning (mean=2.9). These findings suggested that students attitude towards future use of the web is very positive. They found this supplemental mode of learning interesting and 85% wanted more recommended sites (mean=2.63). ( Figure 9) Figure 9 M ean I will use web sites for EFL learning during my education here 3.12 I need more recommended sites for learning EFL and doing my homework 2.98 I will look for more ESL websites myself 2.63 Difficulties and problems encountered The students encounter some difficulties and problems during this study. The main problem was a lack of availability of computer networks, cost of using and getting connected to it. They had to get appointment to use the Internet. (mean= 3.12).The other problem was spending too much time visiting the websites (M = 3). Concerning technical problems, downloading sounds and pictures were serious problems (M =2.90). It was surprising to find out that not having enough access to computer equipment was still a problem (M = 2.80). 90% of the students attempted to complete the assigned homework together using university facilities so that they could 15

29 ASIALEX 2005 get help from each other (mean=3.12). 19% of the students also expressed not being able to understand the instructions on the websites (M = 2.11). This led the researcher to conclude that more guidance should be provided. (Figure 10) Figure 10 Mean It is really expensive to use the Internet outside 312 I had to spend too much time on the web 3.06 It was difficult to download sounds 2.90 I did not have enough access to Internet 2.80 I did not understand instructions on the websites 2.11 We completed the assigned homework together using university facilities 3.12 Conclusions This study evaluated a project on familiarizing students with ESL Web resources. The results of this study indicated that students consider the Internet a useful tool to supplement the class instruction. However, students needed instruction regarding where the s ites are and how to use them. This finding echoes Paulsen s (2000) remarks about the importance of guiding the students in the use of Internet resources. The guidelines used by the teacher were considered to be necessary. Satisfaction with the used guidelines and the contents of the websites outweighed the encountered difficult ies, such as cost, accessing and using the web. Spending too much time, money and getting connected were perceived as the main problems for the students in using the web. Analysis of the data also indicated that the students are more likely to use ESL websites for their further language learning. The results showed that students do not currently access ESL websites on their own due to a lack of availability of the computer networks and the cost. Concerning EFL learning through the web 90% of the students agreed that materials become more comprehensive by using the web. 89% of the students agreed that the web creates active learning and effective contact between students and teacher. 92% of the students agreed that the Internet helps them learn the subject matter on a deeper level, makes class fun, encourages monitoring, helps them think creatively and critically, and also helps them practice independent thinking and autonomy. The students reported to do the following when working with the Internet: to think about what they saw on the web page. to learn technological skills. to learn research skills. to pursue and learn a subject at a deeper level because of the hands-on approach used. to have fun in class. to monitor their own work. to exercise critical thinking. to exercise independent thinking. 16

30 WORDS IN ASIAN CULTURAL CONTEXTS to work cooperatively. to be less dependent on the teacher. All the above data indicated that overall insights into using the teacher selected web sites for EFL were positive. Finally since target language communication and cultures are well within reach through these new technologies, this will make a tremendous impact on the ability of students to communicate directly with native speakers (Leloup and Ponterio, 1998). According to Leloup and Ponterio (1996), there are a number of aspects that are inherent to the web, which make it suitable for teaching languages. First of all, the World Wide Web is a veritable treasure trove of authentic materials for the FL teacher (Leloup and Ponterio, 1996). Teachers always look for authentic materials, for them the web is a real boon with up-to-date material. Lastly, the students of this age are becoming more visually oriented and students who navigate the Web are reading. References Assche, F. van (n.d./1998). The web for schools project. Retrieved in 1999 from the World Wide Web: Barn et, M.A. (1989). More than meets the eye: Foreign language reading theory in practice. Englewood Cliffs, New Jersey: Prentice Hall. Berge, Z., & Collins, M. (1995). Computer-mediated communication and the online classroom in distance learning. Cresskill, NJ: Hampton Press. Boswood, T. (Ed.). (1997). New Ways of Using Computers in Language Teaching. Alexandria, VA: Teachers of English to Speakers of Other Languages, Inc. British Council (n.d./1996). The Internet and English Language Teaching. Retrieved in 1999 from the World Wide Web: Brown, G., & Yule, G. (1983). Teaching the spoken language. Cambridge: Cambridge University Press. Bush, V. (1991). As we may think. In J. M. Nyce & P. Kahn (Eds.), From memex to hypertext (pp ). Academic Press, Inc. Carrell, P. L., Devine, J., & Eskey, D. E. (1988). Interactive approaches to second language reading. Cambridge: Cambridge University Press. Carrell, P. L. (1988). Text Boundedness & Schema Inference. In P. L. Carrell, J. Devine, & D. E. Eskey.(Eds.), Interactive approaches to second language reading. (pp ). Cambridge. Cambridge University Press. Chastain, K. (1998). Development of second language skills: Theory and practice. Harcourt Brace Jovanovich Publishers. Cohen, Andrew D. (1986). Mentalist measures in reading strategy research: Some recent findings. The Hebrew University of Jerusalem. Depoe, M. (2001). Guide to Online Resources. CALL-EJ Online, 2 (2). Available: Egbert. J. (2001). Active Learning through Computer-enhanced Activities. Teaching English with Technology. 1 (3). Available: Farhadi, H., Jafarpour, A., & Birjandi, P. (1994). Language skills testing: From theory to practice. Tehran, Iran: SAMT Publishers. Fowler, R. M. (1994, July). How the secondary orality of the electronic age can awaken us to the primary orality of antiquity or what hypertext can teach us about the Bible. Interpersonal Computing and Technology: An Electronic Journal for the 21st Century, pp Retrieved in 1999 from Archived as fowler IPCTV2N3 on LISTSERV@CUNYVM.CUNY.EDU Felix, U. (1999). Exploiting the Web for language teaching: selected approaches. ReCALL, 11 (1), Frizler, K. [a.k.a. Frizzy] (1995, December 6). The Internet as an educational tool in ESOL writing instruction. Unpublished Master s thesis, San Francisco State University. Retrieved from the World 17

31 ASIALEX 2005 Wide Web in 1999: Godwin-Jones, R. (1998). Language Learning and the Web. Retrieved from the World Wide Web in 1999: Grabe, W. (1997). Reading reseach and its implications for reading assessment. LTRC Papers, Paper1. Papers.com Grabe, W. (1991). Current developments in second language reading research. TESOL Quarterly, 25 (3), Harris, V. (1996). Developing pupil autonomy. In E. Hawkins (Ed.), 30 years of language teaching. London: Centre for Information on Language Teaching and Research. Haworth, W. (1995, July 6). World Language Pages. Retrieved in 1999 from the World Wide Web: Heim, M. (1991). The metaphysics of virtual reality. New York: Oxford University Press. Kern, R. (1995). Restructuring classroom interaction with networked computers: Effects on quantity and quality of language production. Modern Language Journal, 79 (4), Kilian, C. (n.d./1994). How an online course works [ document]. Toronto Globe and Mail, November issue. Available from: Kitao, K. & Kitao, K. (2000). On-Line Resources and Journals: ELT, Linguistics, and Communication. Available: Landow, G. P. (1992). Hypertext: The convergence of contemporary critical theory and Technology. Baltimore: The Johns Hopkins University Press. Lee, J. & Schallert, D. L. (1997). The relative contribution of L2 language proficiency & L1 reading ability to L2 reading performance. TESOL Quarterly, 31 (4), LeLoup, J. W., & Ponterio, R. (n.d./1998). Internet Technologies for Authentic Language Learning Experiences. Retrieved in 1999 from the World Wide Web: Li, R.C. & Hart, R.S. (winter, 1996). What Can the World Wide Web Offer ESL Teachers? TESOL Journal, Lyma n H. (1998). The promise and problems of English on-line: A primer for high school teachers. English Journal, 87 (1), Meloni, C. (1998). The Internet in the classroom: A valuable tool and resource for ESL/EFL Teachers. ESL Magazine. Available:// Mörth, T. (n.d./1998). Eriksdale school: A fully IT-integrated school. Retrieved in 1999 from the World Wide Web: Neu, J., & Scarcella, R. (1991). Word processing in the ESL writing classroom: A survey of student attitudes. In P.Dunkel (ed.), Computer Assisted Language Learning and Testing: Research Issues and Practice. pp Newbury House: Harper Collins, New York. Nyce, J. M. & Kahn, P. (1991). From memex to hypertext: Vannevar Bush and the mind s machine. Boston: Academic Press. Osuna, M. M. & Meskill, C. (1998). Using the World Wide Web to integrate Spanish language and culture: A pilot study. Language Learning & Technology. 1 (2), Paulsen, P. (2001). New Era Trends and Technologies in Foreign Language Learning: An Annotated Bibliography. Interactive Multimedia Electronic Journal of Computer-Enhanced Learning. Available: Raimes, A. (1983). Techniques in Teaching Writing. Oxford, Great Britain: Oxford University Press. River, W. M. (1981). Teaching foreign language skills. Chicago: University of Chicago Press. Roger, C. (n.d./1998). Comenius likes the Web. Retrieved in 1999 from the World Wide Web: Santoro, G. M. (1994, April). The Internet: An overview. Communication Education, 43 (2) Schulz, R. (1983). Literature and readability bridging the gap in foreign language reading. Modern Language Journal, 65, Sin ghal, M. (1997). The Internet and Foreign Language Education: Benefits and Challenges. The Internet TESL Journal, 3 (6). Available: Sperlin g, D. (1997). The Internet Guide for English Language Teachers. Upper Saddle River, NJ: Prentice Hall Regents. [-13-] Sus sex, R & White, P. (1996). Electronic networking. Annual Review of Applied Linguistics, 16, Tudor, I. (1996). Learner-centredness as language education. Cambridge, Great Britain: Cambridge University Press. Warschauer, M., & Healey, D. (1998). Computers and language learning: An overview. Language 18

32 WORDS IN ASIAN CULTURAL CONTEXTS Teaching, 31, Warschauer, M. (Ed.). (1995). Virtual Connections: Online activities and projects for networking language learners. University of Hawaii: Second Language Teaching & Curriculum Center. Warschauer, M. (2000). The changing global economy and the future of English teaching. TESOL Quarterly, 34(3), Warschauer, M., Schetzer, H., & Meloni, C. (2000). Internet for English Teaching. Alexandria, VA: Teachers of English to Speakers of Other Languages, Inc. Widdowson, H. G. (1985). Reading & Communication. In Alderson, J. Charles, & Urquart, A. H. (Eds.), Reading in a Foreign Language. London: Longman. Willis, D. (n.d./1997). A suggested framework for the educational use of Internet technologies & the development of students learning abilities to critically evaluate materials. South Bristol Learning Network Ltd Retrieved in 1999 from the World Wide Web: Web for Schools (n.d./1998). Internet in Schools: Questions and Answers. Retrieved in 1999 from the World Wide Web: Appendixes I. Questionnaire about Reading Please write the number about how you feel. 1 : Strongly Agree 2: Agree 3: Disagree 4: Disagree strongly 1. Reading in English is fun. 2. I like to read in English. 3. I like to read in Persian. 4. Reading in English is a waste of time. 5. Reading in English is boring. 6. Reading in Persian is boring. 7. Time assigned for reading classes is very short. 8. Reading in English enables students to do better in other classes. 9. Students should take more reading courses in English. 10. Reading exercises in English is interesting, useful, and fun. 11. The Reading course has been one of the worst courses. 12. It seems I can do without reading courses. 13. Other English courses are better than reading ones. 14. Reading courses are less time consuming than other courses. 15. Readings should be related to everyday life values 16. Reading in English is worth spending time and it is enjoyable. 17. Most Reading sessions, in English, are dull. 18. Most Reading sessions, in Persian, are dull. 19. I enjoy reading comprehension better than listening. 20. I like to take more reading comprehension courses. II. Questionnaire on the use of the web for EFL Learning Please write the number about how you feel. 1: Strongly Agree 2: Agree 3: Disagree 4: Disagree strongly 1. The web helps me to learn better I spend more time working on EFL when I use the web than other forms of reading. I could think of more ideas for my EFL when I used the web I like using the Internet better than other ways for EFL Usually, I like to read in English through the web I think I am a good reader in English using the web 7. Using the web has helped me to become better in English. 8. I feel I ve learned more about English from this class than I have from other English classes I ve taken in which the web was not used. 9. I plan to continue using the web after this class is finished. 10. The feeling in the class is friendly. 11. Using the web makes me less worried about EFL because I am independently involved in activity. 12. I think I can read longer passages and articles using the web. 19

33 ASIALEX I don t like it when I can t understand what to do when I am trying to read my passages on the web. 14. I can easily make changes when I use the web. 15. I feel I can get more individual attention from the teacher in the web e class than I do in other non-internet classes. 16. I pay more attention to organization when I use the web. 17. The students in this class help each other. 18. When I use the web, I pay more attention to the grammar also. 19. I found it was not difficult to learn to use the web. 20. I was worried that I might break the computer. 21. I was worried that it would take me longer to learn to use the web than it would other students. 22. I think using the web in EFL class is interesting. 23. I would like to take another English course if I could use the web. 24. I get better scores on tests using the web. 25. It was difficult to learn how to use the web. 26. I can change learning speed more easily and more often when I use tee web 27. I feel that I learn better when I get individual attention from the teacher. 28. I think it makes effective contact between students and teacher. 29. It helps cooperation among the students. 30. It creates active learning. 31. It teaches practical technology skills to students. 32. It provides best practices for classroom activities and lessons. 33. It helps me think critically as well as work actively and independently. 34. It helps me improve my reading ability and skill. 35. It helps me to understand that there is a difference between just finding something on the net and finding something good. 36. Materials become more comprehensive by using web. 37. It helps to learn the subject at a deeper level because of the hands-on approach. 38. It helps to make the class fun, and encourages mentorship. 39. It helps to think creatively. 40. It helps to exercise independent thinking III. Questionnaire concerning autonomy 1. I think it is important to read widely on my academic course work. 2. I think it is important for us to learn about the purposes behind given activities. 3. I think the web empowers us to be responsible for our own learning 4. I always take the initiatives when learning about something through the net. 5. I do not enjoy studying using the web. 6. I wish I were given some opportunities to learn English using the web sits. 7. I believe that English classes can be conducted more effectively in a multimedia environment. 8. I think on line learning should be included in English classes. 9. I think teachers should give us help and opportunity to decide on where and how to learn using the web. 10. I like the teacher who introduces EFL web site which empowers us 11. Learning at our pace using the web will create a student centered classroom which increases learner autonomy. 12. I think teachers should give us a lot of opportunities to select the web units we like to learn. 13. I fell more comfortable using the web. 14. Students should be encouraged to challenge their teachers. 15. Web sites empowered us to be initiative, creative, and work independently 16. I like the opportunities to self-correct minor mistakes in my work. 17. I need a lot of guidance in my web learning. (Navigation) 18. I believe it is important to have web-learning sessions. 19. I think the chosen web site empowered my language learning. 20. I think it is important for teachers to give us feedback on our web work. IV. Questionnaire concerning the access and use of the computer Please select the appropriate choice. 1. Do you have access to computer network at home? 20

34 WORDS IN ASIAN CULTURAL CONTEXTS a. Yes b. No 2. Average number of times you get on line every week. a. everyday b. 4 times c. 3 times d. twice 3. Where do you usually get on line a. home b. coffee net c. university d. library 4. Have you ever visited EFL/ESL web sites before this course? a. Yes b. No V. Questionnaire concerning given guidelines Please write the number about how you feel. 1: Strongly Agree 2: Agree 3: Disagree 4: Disagree strongly 1. Before using EFL/ESL web sites, I needed the teacher to provide the selected web sites. 2. Before using the EFL/ESL websites, I needed the teacher to indicate the content of the websites. 3. Before using the websites, I needed guidelines about how to use the websites 4. Using selected hyperextend EFL/ESL web sites made my access more easy and convenient. 5. The explanations provided by the teacher helped me use EFL/ESL web sites for EFL later on. 6. The hyperextend homework in the introduced websites helped me to improve my reading ability. 7. The selected websites were appropriate to my level of reading skill 8. The introduced websi tes were closely related to my lessons in the course. 9. The introduced websites complemented my course lessons. 10. EFL Learning through EFL/ESL web sites is an efficient and effective method. 11. EFL Learning through EF L/ESL web sites is interesting. 12. I need more EFL websites to be introduced. 13. I will use the websites to direct my own learning. 14. When using the websites I often understood the guidelines. 15. Getting on line and downloading pictures or sound were always the problem. 16. I usually had difficulty finding computer network to get on line. 17. More computer networks should available for me to use. 18. I have to pay a lot use the network out side of the university. 19. I spent a lot of line getting on line. 20. I spent a lot of time on line. VI. Questionnaire concerning the difficulties they encountered while using the web It is really expensive to use the Internet outside I had to spend too much time on the web It was difficult to download sounds I did not have enough access to Internet I did not understand instructions on the websites It took a long time to get on line We completed the assigned homework together using university facilities 猫匿 māo nì ; aquí hay g ato encerrado : a cross-c ultural view on Chinese and Spanish idioms around animals Ni colas Arriaga Agrelo City University of Hong Kong narriaga@cityu.edu.hk Abstract Idioms are sayings or expressions whose meaning is not predictable from the usual implications of the constituent elements, or from the general grammatical rules of a 21

35 ASIALEX 2005 language. Chinese and Spanish languages are typologically and culturally quite distant, therefore, metaphorical phrases, similes, familiar sayings, and proverbs in both languages are expected to differ due to their different cultural conceptions and beliefs. This paper aims to present a contrastive analysis of some Chinese and Spanish idioms containing a reference to animals as key word in order to get at some insights about their degree of correspondence in the way they shape figurative meaning and underlying conceptual strategies. Some implications which these findings raise can be useful to second-language teaching and learning, as well as for translation purposes. Lexical organisation and access of code-mixed collocations Somsukla Banerjee Indian Institute of Technology, Kanpur India somsukla@iitk.ac.in Achla M. Raina Indian Institute of Technology, Kanpur India achla@iitk.ac.in Harish Karnick Indian Institute of Technology, Kanpur India hk@iitk.ac.in Abstract Code-mixing is a common phenomenon in bilingual speech and has been studied extensively from a sociolinguistic, and in recent years, a structural perspective. Not enough attention has, however, been paid to lexical restructuring in code-mixing. We study lexical restructuring constraints on code-mixing in code-mixed varieties. We consider a class of code-mixed collocations from Hindi and English, often referred to as idiom chunks, and examine the constraints governing code-mixing in these lexical composites. We propose a distinction between fused and disjointed collocations based on an empirical study that elicited acceptability judgments on English-Hindi code-mixed collocations,. We argue that the way bilingual lexical representations are organised and accessed has a bearing on the acceptability status of code-mixed collocations. Introduction Code-mixing and code-switching are well known features of the bilingual speech repertoire (Grosjean, 97). It has been proposed that often two languages are mixed in a manner that a third, new code emerges in which elements from the two languages are incorporated in a structurally definable pattern (Maschler, 98). Studies on code-mixing have focused on the sociolinguistic and grammatical constraints that govern code-mixing (MacSwan, 97, Agnihotri, 98). However, very little attention has been paid to the lexical restructuring that is at work in code-mixed varieties. Lexical semantic constraints governing a mixed code have not been looked into either. In this paper we explore the nature of the lexical semantic constraints at work in code-mixing. Hindi-English code-mixed lexical composites are examined to look into the issues of their organization and access. The paper is divided into three parts. We undertake a brief discussion on the 22

36 WORDS IN ASIAN CULTURAL CONTEXTS nature of idiom chunks and the motivation for selecting this category of lexical composites for the present study. In the second part of the paper, we report the empirical study that was conducted on Hindi-English bilinguals eliciting acceptability judgments on English-Hindi code-mixed collocations based on idiom chunks. In the concluding part of the paper, we propose a distinction between fused and disjoint collocations and argue that the acceptability status of code-mixed collocations has a bearing on how bilingual lexical representations are organised and accessed. Idioms and code-mixing There are various kinds of composite lexical items in natural language like compound words, lexical doublets and reduplicatives, among many others but possibly none of these categories show as many interesting properties as what is commonly called idiom chunks. Various studies have proposed that idioms are represented and listed in the lexicon as whole entities. But there are problems with this view, as all idioms do not behave in a uniform manner. Certain idioms do allow for constituent modifications while some idioms do not. For example it is possible to passivize the idiom spill the beans as beans were spilled but not the idiom kick the bucket as the bucket was kicked. From these examples it is obvious that certain idioms allow displacement of their constituents while others do not. This asymmetry exists in case of code-mixed idiomatic collocations as well. While it is possible to code-mix a section of these collocations, there are items in this category which strongly resist code-mixing. For example an idiomatic collocation in English like break one s heart can exist in a code-mixed form. It is possible to replace one of its constituents in Hindi and obtain the Hindi English code-mixed forms heart todaa and dil break kiyaa where dil and todaa are Hindi equivalents of the English lexical items heart and break respectively. However this kind of replacement by items from another language is not possible for all idiomatic collocations. In case of an English idiom like break grounds it is not possible to replace either of it s constituents as the code-mixed forms ground todaa and zameen break kiyaa fail to retain the idea conveyed by this item where zameen is the Hindi equivalent of ground. The above examples show that code-mixed composite structures differ in terms of how fused they can be. The code-mixing behavior of lexical composites is likely to give clues about their representation and access in the lexicon and the semantic constraints that underlie code-mixing. In code-mixing there is one language which provides the basic sentential structure fo r a code-mixed utterance and lexical items or phrases from another language are brought into it. In our study the source language is the one which provides the basic framework for the sentence structure and the target language is the one from which items are b rought into the source language. In the present study, Hindi was the source language o f all the items in the task. Sentences in Hindi were constructed which contained idioms from both Hindi and English having the constituent structure V NP (NP V in Hindi). In each sentence the idiom was used in a code-mixed form, i.e., an original constituent of the idiom was replaced in the other language. For example, for an idiom in English, either the verb or the NP was replaced by a Hindi equivalent. This will be clear from examples 1-2 (see below). The empirical study For the purpose of our empirical study we constructed code-mixed sentences in 23

37 ASIALEX 2005 Hindi which contained code-mixed idioms from Hindi and English. 30 Hindi English bilinguals were asked to give acceptability judgements on sentences which contained code-mixed idioms. All the idiom chunks used for the purpose of this study had the constituent structure of [V NP]. Each idiom chunk appeared twice in the task set, once with the [V] replaced by a close translation equivalent in the target language and the other time with the [NP] replaced by a close translation equivalent in the target language. For example an idiom like break a record which has the constituent structure V NP was used twice in the task set with one of its constituents replaced by a Hindi equivalent. Sentences 1 and 2 exemplify the use of code-mixed idioms that were used in our task set: 1. unhone pichle records tode they- nom previous records break past-pl They broke previous the records. 2. unhone pichle kirtimaan break kiye they-nom previous record break do-past They broke the previous records. Sentences (1) and (2) above contain code-mixed versions of the English idiom break the record. In (1) the verb break has been replaced by its Hindi equivalent tod and in sentence 2 the English noun record has been replaced by its Hindi equivalent kirtimaan. We elicited acceptability judgements on 96 sentences which contained 24 idioms each in Hindi and English. The participants, consisting of students and faculty members of the Indian Institute of Technology, Kanpur, were asked to rate the code-mixed sentences on a three point acceptability scale, where 1 was acceptable, 2 moderately acceptable and 3 was to be rated as unacceptable. The code-mixed idiom chunk was highlighted in the sentence and the participants were instructed to rate the sentences on the basis of the semantic content and usage pattern of the code-mixed idiom. In the following section we shall look at the results of our empirical study. Results In this paper we restrict ourselves to the discussion of the English code-mixed idioms tha t were used in our empirical study. Participants rated some of the idiom chunks as acceptable in a code-mixed form while others were rated unacceptable. Some idioms received very high acceptability ratings while others were rated as unacceptable by many subjects. In some of the code-mixed idioms replacing the nouns was acceptable whereas replacing verbs obtained very low acceptability ratings. The following table shows the acceptability ratings of the idioms that generated the highest acceptability ratings for both kinds of replacements. Table 1: Idioms with high acceptability ratings in noun and verb replacements Idioms Acceptability (%) Acceptability (%) (noun replaced) (verb replaced) Play one s cards Read one s mind Miss a chance Jump to conclusions

38 WORDS IN ASIAN CULTURAL CONTEXTS In the English idiom play one s cards, the verb play was replaced by the Hindi equivalent khel while retaining the rest of the idiom in English and a second code-mixed version was obtained by substituting the noun card with its Hindi equivalent pattaa. The two combinations obtained are shown in sentence unhone apne patte acche se play kiye they -cm their card-pl well cm play do-past-agr They played their cards well. Similar sentences were constructed for all th e other items in the task that was undertaken to elicit judgement data from the bilingual participants. However, certain sentences containing code-mixed idioms obtained extremely low ratings in both the possible code-mixed combinations. These idioms and the ratings obtained for them are shown in table 2. Table 3. unhone apne cards acche se khele they-cm their-reflex card-pl well cm play-past-agr They played their cards well. 2: Idioms that obtained low acceptability ratings in noun and verb replacements Idioms Acceptability (%) with Acceptability (%) with replaced noun replaced verb Break the ice Face the music Cool one s heels Throw cold water One interesting observation from the study was the fact that for some idioms one of the constituents could be replaced by an equivalent lexical item or phrase from another language while the other constituent was irreplaceable. For certain idioms replacement of nouns (by a Hindi translation equivalent) received high acceptability ratings but verb replacements received very low acceptability values. Table 3 shows a list of such items. Table 3: Idioms that obtained high acceptability for noun replacement and low acceptability with verb replacements Idioms Acceptability (%) (noun replaced) Acceptability (%) (verb replaced) Cross one s fingers 70 0 Stretch one s legs Pick a quarrel Similar results were found for the code-mixed Hindi idioms that were a part of our empirical study. From the results of our study, it is clear that all idioms do not behave uniformly as far as code-mixing is concerned. Organization and access of lexical composites In this study we have examined composite lexical units having a constituent structure of V NP (NP V in Hindi). As it is well known, verbs generally denote events. 25

39 ASIALEX 2005 In case of the idiom cool one s heels the event is encoded in the entire lexical composite whereas in case of an idiom like read one s mind the event is encoded in the verb. To the extent that the verb encodes the event structure the composite lexical unit will permit its constituents to be code-mixed whereas lexical composites which are less compositional in nature resist code-mixing. But perhaps this explanation is by and large ad hoc in nature as it only explains why certain verbs in composite lexical units resist code-mixing. On the basis of the results of our empirical study, we propose a distinction between fused and disjointed lexical composites. The fused items exist as a composite entity at the level of the Conceptual Structure, assuming bilingual lexical representation has the levels of Conceptual structure and Lexical Structure. Here Conceptual structure consists of arguments and events and encodes information about the number and kind of arguments that an object or event is licensed to take apart from features pertaining to the constitution of an object and its function. The lexical structure contains information about the lexicalization pattern of objects and events and other information belonging to the domain of morphology and syntax. Therefore an item like cool one s heels refers to the concept of taking rest. On the contrary, the items of the disjointed kind do not exist as a fused entity at the level of Conceptual structure. In case of a composite item like break the heart, the constituents exist as separate entities at the level of conceptual structure. In this case, the separate concepts break and heart are fused together to form the composite lexical unit in question. Therefore in case of fused entities, the mappings between the conceptual structure and the lexicon are one to one whereas in case of disjointed entities the mapping between the conceptual structure and the lexicon are not unitary as more than one concepts merge to give rise to such lexical composites. We propose that fused composite language users represent and access fused items as a fixed entity at the level of Conceptual structure and therefore these entities resist code-mixing whereas the disjointed variety that allow code-mixing are represented in terms of separate constituent units. For certain idioms, the nouns were found to be replaceable by the subjects but they refused to accept the sentence as a legitimate code-mixed utterance if the verb was replaced. It is well known that any verb encodes a range of senses in natural language and it is possible that all these senses do not find cross linguistic equivalents in the other language. The polysemy of nouns is more or less equivalent across languages. Therefore in case of an idiom like pick a quarrel in English the verb pick has the sense of pro voke (sense number 4, Wordnet 2.0) whereas this particular sense is unavailable in Hindi for uthaa which is the equivalent of pick in Hindi. Therefore, although the noun quarrel is replaceable by an equivalent lexical item in Hindi, the verb is not replaceable because of the unavailability of this particular sense in Hindi. This can be looked upon as a case of a lexical gap in the language and it is interesting to note how a lexical gap can impede code-mixing although it is often cited as one of the reasons that facilitate code-mixing. Conclusion In the present paper, code-mixed lexical composites were studied to examine how composite lexical units are represented in the conceptual structure and how they are accessed by the bilingual user. Based on results of empirical studies we have proposed a distinction between fused and disjointed lexical composites. We have also tried to account for why certain composite lexical items are relatively more acceptable 26

40 WORDS IN ASIAN CULTURAL CONTEXTS in terms of user judgements when they are used in mixed utterances. The distinction between fused and disjointed composite lexical items can possibly be extended to other categories of lexical composites. References Agnihotri, R.K. (1998), Social Psychological Perspectives on Second Language Learning, Sage Publications, New Delhi. Belazi, Hedi M., Edward J. Rubin and Alm eida J. Toribio (1994), Code switching and x-bar theory: The functional head constraint, Linguistic Inquiry, 25(2): Grosjean, F. (1997), Processing mixed Language: Issues, findings and models In A.de Groot and J. Kroll, Editors, Tutorials in Bilingualism: Psycholinguistic Perspectives, Lawrence Erlbaum Associates, New Jersey. Mahootian, Shahrzad and Beatrice Santorini (1995), Codeswitching and the syntactic status of adnominal adjectives, Lingua, 95:1 27. Pandit, Ira (1990), Grammaticality in code switching, In Rodolfo Jacobson, editor, Codeswitching as a Worldwide Phenomenon. Peter Lang, New York. MacSwan. Jeffrey (1997), A Minimalist Approach to Intrasentential Code Switching: Spanish-Nahuatl bilingualism in Central Mexico, PhD thesis, University of California, Los Angeles. Maschler, Yael (1998) On the transition from code-switching to a mixed code, In Auer, editor, Code- Switching in Conversation: language, interaction and identity. Routledge, London. Sankoff, David and Shana Poplack (1981), A formal grammar for code-switching, Papers in Linguistics, 14(1): Deepak Sharma, K. Vikram, Manav R. Mital, Amitabha Mukerjee, and Achla M Raina (2002), Saarthaka: A generalized HPSG parser for English and Hindi, In Recent Advances in Natural Language Processing - Proceedings ICON-2002, Vikas Publishing House. Abstract Asian Englishes or English in Asia? Kachru s circles and the sociolinguistics of othering Paul Bruthiaux Hong Kong Institute of Education, Hong Kong paulb@ied.edu.hk Most discussions of English worldwide have adopted the Kachru model postulating Inner, Outer, and Expanding contexts for English. I have previously suggested that the model had outlived its usefulness because it is over-reliant on colonial history and misses dynamic sociolinguistic realities. Though the motivation for the model was originally subversive, it no w perpetuates divisions by othering where it should unify. In response, I propose a continuum model of English worldwide that goes beyond simplistic demarcations based solely on historical accounts of national border formation. The aim is to assess manifestations of English across a range of contexts in terms of their degree of localization by representing aspects of local multilingualism, the extent to which English is used privately and publicly as a lingua franca, its role as 27

41 ASIALEX 2005 medium of instruction, and, crucially, the extent to which it is transmitted naturalistically across generations. I show that the distribution of these characteristics in SE Asian and other countries matches the Kachru model only weakly. I argue that the extent to which these characteristics are met determines where on the continuum each manifestation of English falls, ranging from deep integration into the sociolinguistic makeup of each setting or operating mainly as an add-on of relevance only to a local elite. Though the task is a challenging one, sociolinguists will need to incorporate these complexities into their models if they are to dismantle artificial barriers between Englishes in favor of an inclusive view of English across its multiple and multi-faceted manifestations. Words in Asian cultural contexts Susan Butler Macquarie University Susan.Butler@macmillan.com.au More than a decade ago I spent 4 months in Singapore teaching at National University of Singapore (NUS) and became interested in varieties of English in Asia, which was a natural progression from my experience of working on a dictionary of Australian English. I put together a small corpus (4 million words for each variety was my modest aim) and, on the basis of that, began nominating items of high frequency in Asian varieties of English. The schema for classifying the kinds of words I was after was one I borrowed from the introduction of E. E. Morris Austral English where he classified the types of words that seemed to emerge as key items in the Australian lexicon. Classification of the Words: The Dictionary thus includes the following classes of Words, Phrases and Usages; viz. (1) Old English names of Natural Objects Birds, Fishes, Animals, Trees, Plants, etc. (2) English names of objects applied in Australia to others quite different (3) Aboriginal Australian and Maori words which have been incorporated unchanged in the language, and which still denote the original object (4) Aboriginal Australian and Maori words which have been similarly adopted, and which have also had their original meaning extended and applied to other things (5) Anglicised corruptions of such words (6) Fanciful, picturesque, or humorous names given to new Australasian Natural Objects (7) Words and phrases of quite new coinage, or arising from quite new objects or orders of things (8) Scientific names arising exclusively from Australasian necessities, chiefly to denote or describe new Natural Orders, Genera, or Species confined or chiefly appertaining to Australia 28

42 WORDS IN ASIAN CULTURAL CONTEXTS (9) Slang I sorted my selection on a similar basis, beginning with the basics of plants, animals and features of the environment, and moving through social, political and cultural items to slang. Today I have the opportunity to check that selection against a much bigger corpus of English in Asia, the Accent corpus of 200 million words covering the Englishes of Hong Kong, China and Japan, Singapore, Malaysia and the Philippines, and Ind ia, compiled by Dr Kingsley Bolton while he was at Hong Kong University. This is a database of newspaper texts whereas mine was largely literary and it is clear that there are lexical items which feature in one genre and not in the other, but even so its size gives the dictionary editor a great deal of confidence. So I took the opportunity to do a small retrospective on some of my wins and losses from previous corpus reckonings. It is clear that I was wrong about belukar as the accepted term in Singapore and Malaysian English (SME) for secondary jungle probably an item from one of those writers who emphasises the local connection by scattering words of the other language, in this case Malay, throughout the English text. But on the other hand secondary jungle which I dismissed at the time as a rather unwieldy English gloss, seems to be the standar d term in SME. It is not transparent. Even now I am not too clear on what it means and what significance it has to the locals but the simple definition is jungle that has regrown where the original jungle has been cleared. As an Australian I would make a parallel with old-growth forest and new-growth forest, the new growth being lighter, flimsier and less diverse than the old. One of the pleasures of working with a corpus is the unexpected discoveries you make. Looking at jungle in Indian English (IE), I noticed the collocation jungle raj occurring with noticeable frequency. This is a wonderful example of how language communities make the leap from the physical to the metaphorical and coin new lexical items as a result. From the Accent corpus: South China Morning Post 30/03/1993 Developed and operated by Singapore Zoo, the 40-hectare park will be next to the zoo in hilly secondary jung le. Malaysia New Straits Times 16/03/2002 Several farmers who were seen starting a fire in Johan Setia managed to escape into the se condary jungle nearby. Times of India 08/07/2001 If a four-time chief minister, an elderly person, or Central ministers are treated in this brutal fashion, what will happen to the common man? Are we living in a jungle raj or a democracy? asked Maran. Again I seem to have miscalculated on the frequency of jaga kerata boys (car attendants) in ME or perhaps this is the kind of item that doesn t feature on a newspaper corpus. On the other hand, outstation is strong in IE and SME. In Philippine English (PE) dancing pavilion turns out not to be a lexical item but dirty kitchen (the kitchen where the actual work is done) is high frequency. On bayport I seem to have been very wrong a corpus accident rather than a common feature. 29

43 ASIALEX 2005 It seems that pakpai (a car illegally operating as a taxi) is obsolete in Hong Kong English (HKE), just as it has made its way into the New Shorter Oxford Dictionary. On the other hand, jai alai (a ball game) is very strong as is, so I discovered, the collocation jai alai fronton (jai alai arena) in PE. Seefoos and mafoos are still part of the HK horseracing scene. From the Accent corpus: Philippine Star 21/10/2000 Complying with a presidential directive, Interior and Local Government Secretary Alfredo Lim padlocked yesterday the jai alai fronton in Malate, Manila. I still get the sense that wooden house in Malaysia and Thailand is not just a house made of wood but refers to some particular style of house, but I would have to query the locals on that. And while rice barn seems simple enough I would have to guess that it is more like the Australian wheat silo than a traditional barn stacked with hay. The shophouse in SME and the sala in (PE) are here to stay but the pat kwa must be called something else in HKE because I can find no trace of it. Laksa, mooncake and lechon are regular items, but field chicken for a frog was I think a fleeting joke. I liked it beca use it reminded me of similar jokes in Australian English (AE), but it is not a permanent feature of HKE. The term community centre is deceptively simple we have them in Australia too, but it would need research to find out whether the functions of a community centre are the same in the different countries. My sense is that in Singapore it is closer to our City Council. Kongsi is out but clan association and clansmen are in. Mui tsai is an historical item, but a valid one. The same goes for bride-price. Bomoh and red packet are fine. In PE stampita was, I thought, the equivalent of holy card, a card with a saint s image on it, but I now find that you can get stampitas of film stars as well so a more general definition is appropriate. The term Asian values rates high in Hong Kong and Singapore, less so in Malaysia, and doesn t get a mention anywhere else. Actually it would probably rate just as high in AE where we debated the meaning of the term through the 90s. Utang na loob (debt of gratitude) still rates high in PE and gweilo (Caucasian) is still a common item in HKE, but not in SME. From the Accent corpus: Philippine Inquirer 23/07/2001 The F ilipino is also naturally used to reciprocal giving. If he sees that you are sharing with him whatever time, talent and treasure you have, he will make ganti by doing the same (utang na loob). The Way We Were: excerpts from a partial journal, Ai ya (but not aiyoh) is still a feature of SME but walang gawa and dai yat ho were probably snippets of code-switching ra ther than real items of PE or HKE. Many items are confirmed as being special to a particular variety of English in Asia and the basic pattern from the physical environment to the social structures of the commu nity to the linguistic markers of difference are still there. In each variety some words attain special significance and are usually the focus 30

44 WORDS IN ASIAN CULTURAL CONTEXTS for clusters of lexical items that rely on that basic ter m. In AE, the best example is bush, compounds of which stretch for pages in the dictionary. Similarly we can see such clusters developing in the Englishes of Asia. I chose bamboo and rice bowl to illustrate the point. Using the Accent corpus, we can see the wealth of collocation with these items in English in Asia: Collocations for bamboo Accent corpus arrow bamboo bamboo basket bamboo chair bamboo curtain bamboo fan bamboo flute bamboo forest bamboo mat bamboo paper bamboo raft bamboo salt bamboo shoots bamboo slice bamboo slip bamboo thicket bamboo toothpaste bamboo tube bamboo weirs bamboo yacht silk and bamboo genre Collocations for rice bowl Accent corpus South China Morning Post 20/04/1993 The smashing of the iron rice-bowl is putting millions out of work even as it improves the income of those still at work. South China Morning Post 03/08/1993 Two generations of military chairmen, deeply understand one principle: battle power also com es from the rice bowl, commented the newspaper. South China Mor ning Post 26/10/1993 Many people arg ue he s a single-issue chief executive, a man who concentrates solely on China and whose total preoccupation with political and constitutional issues tend to make him overlook many other daily rice- bowl issues that worry the bulk of six million Hong Kongers. South China Morning Post 17/07/1996 The TCDC controversy itself shows the degree to which the fate of the island s economy is entwined with domestic politics at both the more refined level of policy and the level of pork-barrel or rice-bowl benefits. Different diction aries make different selections from these collocations depending on the community of users to who they are directed and the significance that they therefore place on these terms as can be seen in the following Macquarie Dictionary has a minimal selection whereas the Grolier Dictionary, destined for an Asian market was much more generous. 31

45 ASIALEX 2005 Collocations for dragon Macquarie Concise Dictionary 3rd Edition 1998 Regional (Macquarie Project) dragon dragon dragon dragon boat dragon boat dragon boat race dragon cup dragon dance dragon dance dragon dancer dragon dancer dragonfly dragonfly dragon form naga dragon gate dragon gate dragon king dragon lady dragon lady dragon lines dragon lines dragon mask dragon pot dragon pot dragon shrimp dragon shrimp Dragon Throne fire dragon flood dragon komodo dragon komodo dragon komodo dragon rain dragon Collocations for temple Grolier International Dictionary 2000 Macquarie Concise Dictionary 3rd Edition Regional (Macquarie Project) Grolier International Dictionary temple 1 temple 1 temple 1 temple 2 temple 2 temple 2 temple 3 temple 3 temple altar temple bell temple cell temple ceremony temple compound temple compound temple deity temple elder temple festival temple god temple god temple goddess temple medium temple medium temple oracle tem ple priest Buddhist temple Chinese temple go to temple Taoist temple 32

46 WORDS IN ASIAN CULTURAL CONTEXTS These are obvious items of lexical significance but there are other choices we make as communities which are not so obvious. What curious set of historical circumstances led Australians to regard fair dinkum and mate as markers of AE and Singaporeans to regard kiasu and lah as markers of Singaporean English (SE)? The process by which the particular lexicon is developed for each variety is roughly the same. If we go back to Morris we can see that the categories he lists can be viewed as operating according to the processes of (a) normal expansion of meaning, (b) preservation of item s which have become infrequent in other varieties, (c) coinage and (d) borrowing. Dr Bautista provided examples of these in The Lexicon of Philippine English. Expansion of Meaning deep hard-to-understand deep Tagalog stick cigarette top notch successful student bath shower (also SME) Preservation folk barrio folk solon local magistrate Coinage bed spacer rallyist promdi a person who shares a room a political demonstrator provincial (from the province) Borrowing lechon utang na loob balikbayan bakya crowd sari-sari store roast suckling-pig debt of gratitude a Filipino living abroad common crowd convenience store Sometimes, by a curious accident of history we have ended up with the same words but done slightly different things with them. For example in AE, we have borrowed bush and bushranger from American English (AmE). Bush, as we can see above, has become a productive item in AE but has little significance in AmE. Bushranger in AE developed a slightly different meaning and again a much greater cultural significance. Early Australian quotations suggest that the word was first used in Australia with reference to men who lived in and were familiar with the bush and that the use of the word to describe a highwayman or robber was a later development. Similarly, thanks to British colonialism, IE, SME and AE have all ended up with the terms outstation and resthouse but compare the following: outstation Australian English Australian History an outlying station Obsolete a station or outpost of settlement well away from closely settled areas 33

47 ASIALEX 2005 (on a station) a smaller and outlying station or stock-handling depot with accommodation: *The Pines is an out-station of Arcoona where the old bush hospitality reigns. Ion L. Idriess, any remote post: a diplomatic outstation; a military outstation. a small community of Aboriginal people living away from the more populated areas, often associated with one of the larger settlements from which its community has decentralised; country camp: *Got to take a movie camera-man out to one of the mission out-stations: he s wanting further shots of the natives. Coralie Rees, Indian, Singaporean and Malaysian English absent from your usual place of work: an outstation agent; and outstation visitor to a place remote from your usual place of work: to go outstation resthouse Australian English a house or building where travellers can rest and spend the night Singaporean and Malaysian English a place to stay for government officials travelling outstation Philippine English a vacation house The different Englishes show the same effects of adaptation to particular environments, societies and cultures. The difference often lies in the background languages which are feeding items into English and in the layering of dialects within the variety. So for example let us consider this excellent summing up of the state of affairs in SE: Straits Times mid-april 1995 Ah Hock: Eh, what is this man, I heard they are making a Singlish dictionary? Beng: No, lah, who told you, it s not only Singlish. I think if there is a dictionary, it will be on Singapore English. Ah Hock: Aiyah, what s the difference? Beng: You see ah, I have this friend who is a linguistics lecherer. He told me Singapore English is what they call a variety of English, just like Australian English. Singlish is only one sub-variety. Ah Hock: Wah lau, chim man. Beng: No lah. Very simple. You see ah, Singlish has words like kiasu and phrases like blur like sotong. But the grammar also very special, everything not the same as Queen s English one. But Singapore English describes the whole way English is used here. So users include those who speak proper English but also use words with special meaning here, such as void deck in a block of Housing Board flats and hall for living room. These people speak Singapore English 34

48 WORDS IN ASIAN CULTURAL CONTEXTS but also throw in many Singlish words and phrases. Something like rojak, lah. Ah Hock: Orr, so Singapore English is bigger and Singlish is just one small part, lah. Beng: Something like that. But some people don t like Singlish, leh. Ah Hock: Why? Like that got problem, meh? Beng: They are worried that if you put Singlish words into a dictionary, people will think it s OK and then they will anyhow hantam. So each community makes its choices in terms of the words that best reflect the experience of that society. They share these choices with other members of the community and they pass them on to their children. This perfectly straightforward situation is confused by the value judgments that come with it. As individuals we adjust our language choices, usually in adolescence, to reflect our attitudes to life, the style that we wish to adopt. In AE, I could have gone for a broad accent with a lot of nasalisation (although that accent is distinctly unfashionable these days and is becoming an endangered species) and I could have peppered my talk with Australian colloquialisms. Being a woman, I have adopted a general/educated accent (it is a feature of AE that speakers of different dialects can be predominantly one style but include some features of another) and I would use the occasional colloquialism to show that I am an Aussie and that I don t take myself too seriously. This is a delicate matter of judgement to find the right balance between the individual, the local and the global English. As Chris Wallace-Crabbe puts it, speaking of his experience as a speaker and writer of Australian English: Falling into Language by C. Wallace-Crabbe Oxford University Press, 1990, page 44 I can assert my national and cultural identity on the smallest scale by the sounds of my accent, not quite Crocodile Dundee s, but adequate to the purpose: on another scale I could do so by means of a liberal scattering of specifically Australian terms. But if I take this scattering too far I can only invite your assumption that my aim is comic. Suppose I were to say in a lecture that a bloke would be a drongo or at least a bit of a dag if he didn t cotton on to the kind of Toorak-and-ruin lingo that the gubs and ockers keep smarming us with, listeners would know that social conventions were being contravened, the linguistic contract broken, and I could only be joking. The solemnity of genre would be mocked. My choices as an individual are within the range that the whole community values and that value judgment is based on life experience but also on the opinion voiced by the educated section of the community whose business it is to decide such issues teachers, publishers, journalists, writers. This collective value judgment filters through the whole community by osmosis but it can be enormously powerful. If that value judgment is respectful of the nexus between language and culture, then individuals make their choices within a framework that is productive. Their English has authenticity and is creative. It is both in tune with the community which has produced it but has something to offer the wider world. Englishes in Asia are beginning to achieve this recognition and to make their presence felt as legitimate vehicles of Asian culture. 35

49 ASIALEX 2005 Japanese influence on the lexicon of Hong Kong Chinese Louisa S. W. Chan The University of Hong Kong Abstract Constant cultural and commercial contact between Hong Kong and the outside world has brought about vast amounts of lexical borrowing from foreign languages especially English into Hong Kong Chinese (HKC). In recent years, a particularly noticeable trend has been the borrowing of lexical items from Japanese into HKC due to the influence of Japanese culture among the younger generations of Hong Kong. Although research to date on the issue has discussed Japanese loanwords in HKC in the past century, little has been reported on the most recent and ongoing trend of importing Japanese words into the entertainment and advertisement language in Hong Kong. This paper seeks to examine the motivations for such borrowings and fill the gaps in the list of Japanese loan words in HKC reported in the literature. Explanations will be given on how these newly borrowed items were adapted into the local lexicon. The list of items was mainly collected from the media, the entertainment and advertising industries in Hong Kong in the past few years. It was found that the borrowing was induced by the image and prestige that Hong Kong people and the advertising industry associate with the Japanese language. Moreover, the items were mainly borrowed orthographically and semantically but with the Japanese phonological forms entirely replaced by Cantonese sounds. The phenomenon described here illustrates how borrowing invokes very creative processes of adaptation and integration of elements from the source language into the recipient language. Introduction A number of studies have looked into the influence of English on Hong Kong Chinese (HKC) since contact with the West started in the early 16th century (e.g. Chan & Kwok, 1982; Lau, 1999) and a few studies (e.g. Chishima, 2003) have examined the Japanese loanwords in HKC borrowed for the past 100 years or so. However, the past few years saw a new trend of lexical borrowing from Japanese into the entertainment and advertisement language in Hong Kong. Thus, this paper aims to examine the reasons behind this recent phenomenon and fill the gaps in the list of Japanese loan words in HKC reported in the literature. The discussion is preceded by an overview of what lexical borrowing is and the motivations behind such a linguistic phenomenon. The term borrowing refers to the incorporation of foreign features into a group s native language by speakers of that language and it involves recipient language agentivity, which distinguishes borrowing from imposition in which the changes are induced by the source language speakers (Winford, 2003). Normally, contact-induced borrowing begins with lexical items like nouns, adjectives and verbs, and substantial lexical borrowing can invoke the borrowing of other structural features, e.g. phonological and morphological ones. He also points out that borrowing can occur even without real contact with speakers of the source language. Such borrowing can be brought about by marginal or distant contact like travel or exposure to the source 36

50 WORDS IN ASIAN CULTURAL CONTEXTS language through the mass media, and in these situations, there need not be a significant proportion of bilingual speakers in the recipient community. Additionally, he explains that contact-induced lexical borrowing is not restricted to direct importation of words but also includes adaptation and even creation of words. Several types of lexical borrowing have been identified by linguists. For example, Hock and Joseph (1996) categorize lexical borrowing into (1) lexical adoption, (2) semantic loan, (3) loan translation, and (4) loan blends. Winford (2003) added a new type called creations, which are made up of elements from the recipient language and/or the source language with no counterparts in the source language. Chan and Kwok (1982) also point out that phonetic loan is another common method adopted in lexical borrowing. The two most discussed factors leading to lexical borrowing are need and prestige (Hock & Joseph, 1996; Winford, 2003). The former is the need for vocabulary to refer to novel or foreign objects or concepts, which mainly include cultural, technical or religious artifacts or concepts. The second factor, prestige, is the wish of the speakers of the recipient language to associate with the culture of the source language. Lexical borrowing from Japanese in Hong Kong Chinese All the newly borrowed words from Japanese (with the exception of the letter の ) documented here are made up of Kanji ( 漢字 ), Chinese characters in Japanese since Chinese was a source of borrowing for Japanese centuries ago. However, the Kanji characters are d ifferent from their counterparts in Standard Written Chinese (SWC) orthographically, phonologically, and even semantically. Recently, HKC has incorporated the orthographic and semantic form s of some Kanji characters but not their phonological for ms since the p roportion of Hon g Kong people w ho can speak Japanese is still limited though increasing. As a result, very interestingly, these loanwords are written in Japane se Kanji, but pronounced entirely with Cantonese pronunciations and some of their meanings have also been changed. The following is a list of examples collected from the entertainment and advertisement language that appears in shops, commodities, popular newspapers and magazines. A comparison of the borrowed Kanji expressions and their counterparts in HKC in t erms of phonology and orthography is presented in the following table. Explanations are given below the table. Table 1: A comparison of the Kanji loanwords and their count erparts in HKC Japanese Kanji Japanese Cantonese Original Meaning** (now borrowed in pronunciation pronunciation used counterpart(s) in HKC) in Hong Kong HKC 1. 大割引 oowaribiki daai6 got3 yan5 大減價 [daai6 big sale gaam2 ga3] or 傾 銷 [king1 siu1] 2. 放題 hoodai fong3 tai4 蒲飛 [pou4 fei1] buffet (unlimited or 任食 [yam6 offering) sik6] 3. 美肌 bihada mei5 gei1 美容 [mei5 skin care yung4] or 護膚 [wu6 fu1] 4. 瘦身 sooshin sau3 san1 減肥 [gaam2 to lose weight, to keep fit fei4] 5. 魔力 maryoku mo1 lik6 魅力 [mei6 lik6] charm ( magical power) 37

51 ASIALEX 素材 sozai sou3 choi4 材料 [choi4 liu2] material 7. 一番 ichiban yat faan1 優質 [yau1 jat1] of top quality (most) 8. 新登場 shintoojoo san1 dang1 新上市 [san1 new on market cheung4 seung5 si5] 9. 新発売 shinhatsubai san1 faat3 hok3 新上市 [san1 new on market seung5 si5] 10. 人気 ninki yan4 hei3 流行 [lau4 hang4] popular 11. 年中無休 nenj uumukyuu nin4 jung1 mou4 nil round the clock yau1 12. 素人 shirooto sou3 yan4 nil ordinary women acting in adult movies (ordinary people) 13. 漢方 kanpoo hon3 fong1 中藥 [jung1 Chinese medicine yeuk6] 14. 無印 mujirushi mou4 yan3 環保 [waan4 environmentally friendly bou2] (not branded) 15. 花火 hanabi fa1 fo2 煙花 [yin1 fa1] fireworks 16. 喫茶店 kissaten hek3 cha4 dim3 茶館 [cha4 gun2] tea-drinking shop or 茶坊 [cha4 fong1] 17. 攻略 kooryaku gung1 leuk6 策略 [chaak3 tactics leuk6] 18. 暴走族 boosoozoku bou6 jau2 juk6 nil a gang of young motorcyclers in Japan 19. 中古 chuuko jung1 gu2 二手 [yi6 sau2] second-hand 20. 部屋 heya bou6 nguk1 房間 [fong4 room gaan1] 21. 財布 saifu choi4 bou3 銀包 [ngan4 wallet baau1] 22. 駅 eki yik6, chek3 站 [jaam6] station **Note: Where there have been changes to the meaning of the borrowed item, the word indicated in brackets in the last column is the original meaning in Japanese. 1. 大割引 is absent in SWC and exists only in Japanese. It means big sale. Now, the Japanese expression is widely used in shops and department stores in Hong Kong. 2. 放題 is absent in SWC. It means unlimited offering and is now widely used in restaurants to mean 3. buffet not only in Hong Kong, but also in Shenzhen and Macau. Now, there is also a new form of business in Hong Kong called 漫畫放題 [maan6 wa2 fong3 tai4], where the customers pay a fixed hourly rate for reading as many comic books as they like. 美肌 was seldom used in HKC in the past though the two characters do exist in SWC, but now the loan expression is widely used by the cosmetics industry in Hong Kong. 4. The original HKC expression should be 減肥 [gaam2 fei4] meaning to lose weight or to keep fit, but now the cosmetics industry prefers to use the loan expression 瘦身. 5. The original term should be 魅力 [mei6 lik6], which means charm. Now, 魔力, which means magical power is widely used by the cosmetics industry in Hong Kong to mean charm. 6. 素材 is now widely used in photography and computer graphical design to refer to material in the backdrop of pictures. 7. 一番, which means most in Japanese, can now be widely seen in shops in Hong Kong, especially in Shum Shui Po selling Japanese electronic products meaning good quality or the best. 8. 新登場 can often be found in shops and magazines. 9. Like 新登場, 新発売 is also widely used in shops and magazines to mean new on market but more interestingly, the Kanji characters 発売 do not exist in SWC and therefore Hong Kong people 38

52 WORDS IN ASIAN CULTURAL CONTEXTS pronounce the characters as [faat3 hok3], which is the pronunciation of the orthographically closest characters 發殼. 10. 人気 is now widely used but written as 人氣 ( 氣 is the complex form which is more widely used in Hong Kong than the simplified form 気 ) in the entertainment media, e.g. 人氣歌手 [yan4 hei3 go1 sau2] and the fashion business, e.g. 人氣手袋 [yan4 hei3 sau2 doi6] meaning popular singer and popular handbag respectively. 11. 年中無休 does not exist in SWC. It means round the clock and can now be found in some convenience stores in Hong Kong. 12. 素人 does not exist in SWC. It means ordinary people in Japanese. Now, it has been borrowed in Hong Kong to refer to some ordinary women (not celebrities) acting in adult movies, and is particularly commonly used among the young male generation. This is an example showing semantic change in the borrowed item. 13. 漢方 is widely used to refer to Chinese medicines in the health care business, e.g. herbal tea and herbal medicine, to indicate the product contains Chinese medicinal ingredients. 14. 無印 does not exist in SWC. It means not branded in Japanese. Now, the expression has been borrowed into HKC with its meaning extended to environmentally friendly because of a chain store called 無印良品 [mou4 yan3 leung4 ban2], which sells environmentally friendly products in Hong Kong. This is another example of semantically changed loan expressions. 15. Now, 花火 is generally understood by the young generation as fireworks because it was used as the title of a popular song in Hong Kong. 16. Some tea-drinking shops are called 喫茶館 although they are not selling Japanese but Chinese tea. 17. Due to the popularity of Japanese computer games in Hong Kong, 攻略 has replaced 策略 completely to mean tactics in the games and is used widely by the computer games players. 18. This can be the only example in the list used for naming novel items. Recently, there have been groups of youngsters called 暴走族 gathering for motorcycle races in Tokyo and now the term 暴走 [bou6 jau2] ( run fast ) is combined with the native character 鞋 [haai4] ( shoes ), to refer to an innovative merchandise, boots fitted with wheels. 19. Now 中古, meaning second-hand, is used in both Hong Kong and Taiwan. 20. 部屋 is now widely used to replace 房間 [fong4 gaan1] in popular newspapers and magazines. 21. 財布 is now widely used to replace 銀包 [ngan4 baau1] in popular magazines. 22. 駅 meaning station in Japanese is now used in the name of a shopping mall in Hong Kong. This character should be pronounced as [yik6] in Cantonese but since it resembles another Chinese character 尺 [chek3] orthographically, some Hong Kong people also pronounce it as [chek3]. Apart from the above, one particularly widely borrowed Japanese letter is の [no] which means of and very interestingly Hong Kong people only take its orthographic form but pronounce it as [ji1], which is in fact the pronunciation of the Chinese character 之, which also means of. Recently, there has been an extensive use of の in place of 之 in the commercial sector. For instance, a chain store in Hong Kong is called 優の良品 [yau1 ji1 leung4 ban2] (meaning products of good quality ). This の is commonly used not only in Hong Kong but also in Shenzhen, Zhuhai and Taiwan probably to project a Japanese image for the business companies. Most of the loan items listed above do have their counterparts in HKC, so the use of the Japanese words is not triggered by the need to name novel or foreign ideas. Also, most of their original counterparts in HKC are still in use, so the loan expressions have not entirely replaced the original ones and they are not commonly used in Hong Kong people s actual speech. They are mainly borrowed for commercial purposes to name commodities and businesses with an aim to associate the products or services with Japa nese culture and style. They are used as catchy phrases in the media and advertisements as a token of fashion. When Hong Kong people need to refer to this language in the media and the advertisements, they pronounce the expressions with Cantonese pronunciations. 39

53 ASIALEX 2005 Conclusion To conclude, this paper has attempted to account for the latest trend of lexical borrowing from Japanese into the entertainment and advertisement language in Hong Kong and provided a list of Japanese loanwords with their orthographical forms and most of their semantic forms borrowed, but entirely pronounced with the phonological forms of Cantonese. This illustrates that lexical borrowing involves creative adaptation and integration of elements from the source language into the recipient language. References Chan, Mimi and Helen, Kwok (1982), A study of lexical borrowing from English in Hong Kong Chinese (Hong Kong: University of Hong Kong). Chishima, Eiichi (2003), A study on Japanese loanwords in Hong Kong Cantonese, paper presented at the 9th International Conference on Cantonese and Other Yue Dialects, Macau, December. Hock, Hans Henrich and Brian D. Joseph (1996), Language history, language change, and language relationship: An introduction to historical and comparative linguistics (Berlin: Mouton de Gruyter). Lau, Martin (1999), Lexical borrowing in Hong Kong: A study of the Englishization of Chinese and the nativization of English, unpublished master s thesis (University of Hong Kong: Hong Kong). Winford, Donald (2003), An introduction to contact linguistics (Malden, MA: Blackwell Publishing). The pragmatics of foreign words in discourse of Chinese popular literature Abstract Vincent Tao-hsun Chang Southern Taiwan University of Technology vthchang@ms32.hinet.net This paper aims to propose a sociopragmatic study along with a critical analysis of the English/foreign words in a series of Chinese column articles (written by Wang Wen-hua in China Times ), where he exploited various forms of English expressions and foreign el ements. The names of persons (artist/writer/celebrity), place, literature, brand names, technical terms, acronyms an d short phrases, from both Western and Eastern/(Japanese) cultures, are widely employed within the articles to attract the attention of the audience, to perform diverse communicative functions thereafter, and to convey the significant ideologies, viz. intellectualism, elitism, friends rapport, feminine awakening and glocalization. These reflect the current lifestyles of petits bourgeois and the linguistic hegemony of English and foreign languages within the symbolic domain of popular culture, inviting the audience to recognize the prominent intercultural values and furthermore to construct the identity of cultural pluralism. The speaker-audience relationship in communication is reexamined by looking into the articles released from 1999 to The sociocultural aspect of language use, on the other hand, is further explored to see the inseparable relationship between language and social function. This functional linguistic study reveals that the popular literature could well serve to embrace competing forces and witness the symbolic power reified within language. The dialogic relations between form and function in literary discourse reflect the interaction 40

54 WORDS IN ASIAN CULTURAL CONTEXTS and dynamism of communicator and audience, thus keeping the dialectical relationship between sociocultural structures and social practice/discourse (Fairclough 1995). A Contrastive study of bare nouns and [classifier + noun] phrases in Hokkien and Cantonese Paulina Yin-ling Cheung Purdue University pcheung@purdue.edu Abstract This paper addresses the similarities and differences between Hokkien and Cantonese in the distribution and the various interpretations of the indefinite, definite and generic bare nouns and [classifier+noun] phrases. I propose the source of the differences lies in the language-specific limitation in Hokkien. More concretely, in interpreting the definiteness of bare nouns and [classifier+noun] phrases, while Hokkien, like Mandarin, prohibits the use of overt classifier without the presence of a Numeral, Cantonese does not. Introduction: The question and the hypothesis In this paper, I attempt to discuss the similarities and the differences in these two languages in terms of the distribution and various interpretations of bare nouns and [Cl+N] phrases. Why definite NPs in Hokkien can only be interpreted as bare nouns but not [Cl+N] phrases? Why definite NPs in Cantonese are [Cl+N] phrases but not bare nouns? Why bare nouns in Hokkien and Cantonese NPs can only be interpreted as generic, but never as [Cl+N] phrases? Specifically, I will argue for the points as follows: a. Hokkien [Cl+N] phrases cannot be interpreted as definite; b. Hokkien bare nouns and [Cl+N] phrases can both be interpreted as indefinite; c. Bare nouns in Hokkien and Cantonese must be embedded in a projection which carries out a deictic function. Brief background on Hokkien and Cantonese It has been observed that Hokkien and Cantonese are well-known and prominent varieties of Chinese other than Mandarin. Hokkien is a language spoken in southern Fujian. Min, Min Nan, and Fujianese are the alternate names for Hokkien. Hokkien is widely spoken in Brunei, Canada, Hong Kong, Indonesia (Java and Bali), Malaysia, North America, Philippines, Singapore, Taiwan, and Thailand. Cantonese is spoken in southern China, namely Guangdong and Guangxi, Hong Kong and Macau. Cantonese as spoken in Hong Kong is called Gwongdung Wa. In Singapore and Malaysia, Cantonese is known as Gwongfu Wa. Due to the emigration from Hong Kong and from Guangdong, varieties of Cantonese can also be found in Australia, Canada, and North America (Matthews and Yip, 1993). 41

55 ASIALEX 2005 Notion of bare nouns, classifiers and massifiers in Cantonese and Hokkien It is believed that in Cantonese and Hokkien, a bare noun comes with a classifier which heads its own projection. A bare noun is a classifier phrase (ClP) (Tang, 1990; Cheng & Sybesma, 1999) I follow Longobardi (1994) and Cheng & Sybesma (1999) in assuming that indefinite bare nouns, which are confined to lexically governed positions, are considered as a ClP plus an empty Cl 0. Longobardi analyzed the bare nominals in Italian while Cheng and Sybesma focused on Mandarin and Cantonese counterparts. The empty Cl 0 is like an empty determiner (D) in a determiner phrase (DP) in DP extended bar theory suggested by Abney ( 1987). Similar to D, Classifiers have an individualizing and singularizing function. (Cheng & Sybesma, 1999, p.518) D has a role to refer and NP to describe. D acts as an intermediate between the description of NP and the application in the description of objects and items in reality. Another analogy is that VP has the function to describe and deictic T to refer. I suggest that Cl 0 in Hokkien and Cantonese carries out the function of D 0. There are no other Ds in Hokkien and Cantonese since there are no articles in these languages. Some entities describe and other carry out the deictic function are typical characteristics of Universal Grammar. Classifiers have built-in semantic partitioning (Cheng & Sybesma, 1999, p. 515) while massifiers create the units of measure. The difference of count/mass nouns in Hokkien and Cantonese does not occur at the noun level, but at the classifier, due to an absence of singular or plural distinction in Hokkien and Cantonese (Doetjes, 1997). Differences between Cantonese and Hokkien in the interpretation of different kinds of nouns There are two crucial differences between Cantonese and Hokkien as regards the dist ribution and interpretation of bare nouns. First, in Hokkien, definite NPs are bare nouns rather than [Cl+N] phrases. Second, Cantonese definite is only manifested as [Cl+N] phrases instead of bare nouns. Syntactically, Hokkien bare nouns can be interpreted postverbally as indefinite (1a), definite (1b), or generic (1c). They can only be understood preverbally as definite (2a) and (2b), and as generic (2c). [Hokkien] (1) a. Wuhui key mui chia. Wuhui go buy book Wuhui went to buy a book/books. b. Wuhui lim-le tang. Wuhui drink-finish-le soup Wuhui finished the soup. c. Wa ai gau. I like dog. I like dogs. [Hokkien] (2) a. Gau ma guo mei lu. dog want cross road The dog wants to cross the road. * A dog wants to cross the road. 42

56 WORDS IN ASIAN CULTURAL CONTEXTS b. Gau ga li dza gwai. Dog today very obedient. The dog/dogs was/were very obedient today. c. Gau ai dzek mah. Dog love eat meat Dogs love to eat meat. Instead of bare nouns, Cantonese applies [ Cl+N] phrases (Leung, 1980), as depicted in (3b), (4b), (4c). [Cl+N] phrases cannot be interpreted as plural as in (3aii, 3b, 4b, 4c). [Cantonese] (3) ai. Wufei heoi maai syu. OR aii. Wufei heoi maai bun syu. Wufei go buy book Wufei go buy CL book Wufei went to buy a book/books. Wufei went to buy a book. b. Wufei jam-jyun *(wun) tong la. Wufei drink-finish CL 3 soup SFP 4 Wufei finished drinking the soup. c. Ngo zungi gau. I like dog I like dogs. [Cantonese] (4) a. *Gau soeng gwo maalou. dog want cross road A dog wants to cross the road. b. Zek gau gamjat dakbit tengwaa. CL dog today special obedient The dog today is specially obedient today. c. Ngo zungji tong zek gau waan. I like with CL dog play I like to play with the dog. d. Gau zungi sek juk. Dog like eat meat. Dogs love to eat meat. Bare nouns in Hokkien are interpreted as definite. Cheng and Sybesma (1999), in their analysis of Mandarin definite NPs, claim that definite bare nouns are classifier phrases (ClPs). I follow the same reasoning. If definite bare nouns in Hokkien are ClPs, it can be deduced that after N-to-Cl movement, bare nouns are assigned a definite interpretion. This means when the Cl position is filled, definite bare nouns are considered to be freed from the lexically governed position. It is assumed that an indefinite [Cl+N] is a NumeralP which has an empty Numeral head and must be lexically governed. [Cl+N] phrases that are NumeralPs having an empty Numeral head will yield an indefinite reading; [Cl+N] phrases that are ClPs will yield a definite reading as suggested by Cheng & Sybesma (1999, p.529). Further, I follow Chierchia (1998) and assume that, in the case of Hokkien which does not have a definite article, the covert iota operator ɩ will be applied. The ɩ operator functions like a definite article. Chierchia (1998) proposes that ɩ will be 43

57 ASIALEX 2005 unavailable if a language has a definite article. Though Hokkien does not have definite articles, the language has a classifier which is, theoretically, equivalent to a definite article. However, practically, since definite [Cl+N] is impossible in Hokkien due to the fact that classifiers in Hokkien must be accompanied by a Numeral the ɩ operator can still be used. Definite bare nouns in Hokkien can be interpreted as singular, for example, Gau ma guo mei lu (The dog wants to cross the road) or plural, such as, Gau ga li dze guai (The dogs were very obedient today). In short, [Cl+N] phrases in Hokkien cannot be interpreted as definite because overt classifiers in Hokkien should always be accompanied by a Numeral. In Cantonese, the application of a classifier does not necessarily imply the presence of a Numeral. Similarities between Cantonese and Hokkien In Hokkien and Cantonese, indefinite [Cl+N] phrases have only an empty Numeral head; indefinite bare nouns have an empty Numeral head and an empty Cl head. A Numeral is needed for the indefinite interpretation because the numeral can undo the definiteness of NPs (Cheng & Sybesma, 1999, p.528). The larger numeral phrase projection in (6) can explain why the noun phrases can be interpreted as indefinite. Also, in the case of Hokkien, classifiers must always be accompanied by a Numeral. Cantonese and Hokkien exhibit the same pattern in the interpretation of generic bare nouns. A subdivision of kind (contrary to an ordinary individual or an object) denoting noun phrases is, at semantic level, equivalent to proper names (Krifka et al, 1995). If I assume Longobardi s (1994) and Ritter s (1989) sayings that generics in English, like proper names, involve N-to-D is correct, given generics in Cantonese and Hokkien resemble proper names, bare nouns with a kind interpretation in these two languages will also go through N-to-Cl movement. This means that N fills the Cl position. As a result, the bare nouns are not limited to lexically governed positions under the generic interpretation because they are restricted in indefinite interpretation. Conclusion The most crucial factor that accounts for the difference between Hokkien and Cantonese in the analysis lies in the fact that overt classifiers in Hokkien must always be accompanied by an overt or covert Numeral. Put differently, in Hokkien, overt classifiers can occur only with a NumeralP, the head can either be overt or covert. Hokkien, unlike Cantonese, is banned from filling the Cl 0 by inserting a classifier. Therefore, Hokkien must make use of the ɩ operator, and N-to-Cl movement. This serves (a) to account for [Cl+N] phrase be indefinite in Hokkien, but definite or indefinite in Cantonese; and (b) to make clear why bare nouns can be indefinite in Cantonese, but definite and indefinite in Hokkien. Cantonese only undergoes N-to-Cl movement with generics for Cantonese does not require the use of overt classifier together with a Numeral. In my analysis, at least at syntactic level, Hokkien functions very much the same as Mandarin in terms of the distribution and the various understanding of bare nouns and [Cl+N] phrases (cf. Cheng & Sybesma, 1999). This study answers the third hypothesis I made: In order to point out the distribution of indefinite bare nouns and indefinite [Cl+N] phrases, NP must be embedded in a projection that carries out the deictic discourse function, that is, to 44

58 WORDS IN ASIAN CULTURAL CONTEXTS describe and to refer to the objects/items in the reality. Indefinite bare nouns and [Cl+N] phrases involve an empty head in the projection above NP, so they are restricted in a position that is lexically governed. On the contrary, definite counterparts do not involve an empty head; hence they are not restricted to lexically governed position. Last but not least, two important structures should always be taken into account in explicating indefinite and definite bare nouns and [Cl+N] phrases. First, the structure in (6), which contains a NumeralP and a ClP, represents the indefinite NPs. Second, the structure in (7) accounts for definite NPs in an overt classifier as well as the covert ɩ operator leading to N-to-Cl movement. References Abney, S. (1987). The English noun phrase in its sentential aspect. Unpublished Ph.D Dissertation, MIT. Cheng, L. & Sybesma, R. (1999). Bare and Not-So-Bare Nouns and the Structure of NP. Linguistic Inquiry, 30(4), Chierchia, G. (1998). Reference to kinds across languages. Natural Language Semantics, 6, Chinese, Min Nan: A Language of China. [On-line]. Available: Doetjes, J. (1997). Quantifiers and selection: On the distribution of quantifying expressions in French, Dutch and E nglish. Unpublished Ph.D dissertation. Leiden University, The Netherlands. Krifka, M., Pelletier, F., Carlson, G., Meulen, A., Link, G., Chierchia, G. (1995). Genericity: An introduction. In The Generic Book. Edited by Carlson, G & Pelletier, F. Chicago: Chicago University Press. Leung, C. S. (1980). Comparative grammar of Mandarin and Cantonese: markers of definiteness. Language Learning and Research, 1, Longobardi, G. (1994). Reference and Proper Names. Linguistic Inquiry, 25, Matthews, S. & Yip, V. (1994). Cantonese: A Comprehensive Grammar. London: Routledge. Min linguistics. [On-line]. Available: Ritter, E. (1989). A head-movement approach to construct-state noun phrases. Linguistics, 26, Tang, J. (1990). Chinese phrase structure and the extended X -theory. Unpublished Ph.D dissertation. Cornell University, New York. Yip, P. C. & Rimmington, D. (2004). Chinese: A Comprehensive Grammar. London: Routledge. Changing patterns and attitudes: naming practices among Malay Singaporeans Abstract Huang Hoon Chng National University of Singapore ellchh@nus.edu.sg Peter K W Tan National University of Singapore petertan@nus.edu.sg In research on words, the category of proper names has exercised the minds of linguistic philosophers, with proper names being said to emphasise reference (over sense) and connotation (over denotation). Partly because of these characteristics, personal names 45

59 ASIALEX 2005 play an important part in the linguistic culture of many communities. Evidence of hybridised communities can often be found in hybridised personal names. For example, the majority Chinese community in Singapore has been seen to have undergone some language shift towards English and the Englishisation (hybridisation) of personal names of this community is clearly evident through the use of English given names. However, the local Malay community is generally perceived as having been able to hold on to its traditional values. However, well known Malay actors in Singapore include Sharon Ismail and Aaron Aziz: Sharon and Aaron (rather than its cognate Harun) are English names. We therefore investigate if it is in fact correct to say that the community is culturally conservative through their attitude towards personal naming. Our data consist of name lists as well as questionnaires and interviews to tease out their attitudes towards traditional Malay names, Englishised names as well as Arabicised names. Englishised or Arabicised names can be seen through the actual name choices, orthographic manipulation or the use of particular pet forms. Respondents are categorised according to age and education level. Through this we will comment on the changing patterns in and attitudes to personal naming in the Malay community in Singapore. Alignment of the idiomatic expression in Korean Lexical Semantic Net for connecting with EuroWordNet Hochol Choe Korea University Janggeun Oh Korea University Donghyeok Lee Korea University Abstract In an effort to globalize and set up the information networks for the Korean studies, the work to build the Korean Lexical Semantic Net (KLSN) is being done vigorously in Korea. To connect the Korean Lexical Semantic Net with such a multilingual database as EuroWordNet is not an easy task, however. Semantic classification of word senses is quite affected by its cultural situation, being classified as too detailed or too general. Actually, Linking KLSN with EuroWordNet may cause some considerations such as mismatch and deficiency, which quite cannot be connected by ILI (Inter-Lingual-Index) used in EuroWordNet. In this case, the Korean idiomatic expressions, collocations and metaphors should be at issues to build KLSN. But, these kinds of expressions (idiomatic expression, collocation and metaphor) are usually reflected by cultural situation. Therefore, throughout this paper, we are going to concern the pragmatic gap that means being same concept also in other languages but expressed with different ways. Particularly, we are going to focus on determining the correspondent meaning in target language against the Korean idiomatic expressions. For instance, it is quite difficult to find Korean word sense as substitute for the English idiom take the veil. Even though, same lexical expression is translated differently such as Eating a Tofu means Bantering someone in Chinese 46

60 WORDS IN ASIAN CULTURAL CONTEXTS and Make someone clean in Korean. Therefore, the purpose of this paper shall be to treat the alignment of the idiomatic expression in KLSN for connecting with EuroWordNet, which usually shows up mismatch and deficiency in ILI-Linking. 1. Introduction This paper examines how to arrange Idiomatic Expression (IE) into Korean Lexical Semantic Network (KLSN) and how to connect Korean IE in KLSN to EuroWordNet. 2. Problems The IE is a multiword construction that is a semantic unit whose meaning cannot be deduced from the meanings of its constituents, and has a non-productive syntactic structure. Therefore, one cannot take the meaning of the IE literally, but must take it figuratively. As stated, due to its strong dependency on the culture where the target language is used, it is hard to construct Synset in the same language and to correspond directly in the different languages. For example, Korean IE, gimch iguk-butŏ masi-da (to drink the Korean style cabbage soup) does not have any words to be Synset (Synonym Set) in Korean, but also it is hard to find out the corresponding expression in any other languages. However, the Korean-language-based machine translation should make the corresponding expression about g imch iguk-butŏ masi-da in the target language. If you want to find a correspo nd expression through EuroWordNet, IE gimch iguk-butŏ masi-da should be connected with EuroWordNet appropriately. 3. Alignment of the idiomatic expression in KLSN KLSN defines IE as a phrase, consisted with more than two words, with a relatively prefabricated form, and the words in the phrase make a totally different meaning in forming the phase. KLSN characterized IE in a more strict sense when compared to WordNet, which defines a phrase as IE when at least one word in the phrase has a figurative meaning. Therefore, KLSN employs the whole IE as a representative word rather than a case of polysemy. For example, if miyŏkguk-ŭl mŏk-da (to eat seaweed soup), in the Korean IE, is treated as a polysemy of to eat, it can find the word network for the word to eat, but it fails to find the one for the other word components of the IE, seaweed soup, which consequently will make the IE meaningless. The same problem occurs when it treats seaweed soup as a case of polysemy. The IE in KLSN is a main headword as a whole and a presented meaning of IE is based on Synset. For example, Korean IE miyŏkguk-ŭl mŏk-da (to eat seaweed soup), which has semantic information as follows: lemma= miyŏkguk-ŭl mŏk-da The verb phrase miyŏkguk-ŭl mŏk-da (to have seaweed-soup) has three senses. 1. miyŏkguk-ŭl mŏk-da (to have seaweed soup), nakbangha-da (to fail an exam), bulhapgyŏkha-da (flunked from the contest) ( to fail an exam, He failed this exam again) 2. miyŏkguk-ŭl mŏk-da (to eat seaweed soup), naksŏnha-da (to fail in an election ) ( to fail in 47

61 ASIALEX 2005 an election, He didn t get elected from election ) 3. miyŏkguk-ŭl mŏk-da (to have seaweed soup), t oetchamat-da (to get rejected, to get the frozen mitten). ( to be denied, He proposed to her with courage, but his love got rejected from her ) And each sense of miyŏkguk-ŭl mŏk-da has another wordnet which is also different from the sense of other words. The following figure shows that the first sense of miyŏkguk-ŭl mŏk-da has semantic relations with other senses of word. Figure 1 Semantic relation of miyŏkguk-ŭl mŏk-da in KLSN 4. Connecting KLSN with EuroWordNet ILI (Inter-Lingual-Index) is used to connect Korean language WordNet to EuroWordNet which is a multi-lingual wordnet. The purpose of ILI, which provides the link of meaning of vocabulary between each separate language, is to connect with not only the internal relation of language but also mutual semantic correlations between separate languages. Since IE includes rich cultural meanings of the target language, it can be helpful to understand the meaning of words which cannot be represented with a word(s) in a certain language system. Due to this reason, the absence of the meaning and no correspondence between words often occur when one language system is corresponded with the other. But The EuroWordNet, which is trying to link to KLSN, has given no specific solution to solve this problem. In this section, we will design a way to connect Korean IE mŏri-rŭl ŏn-da and gimch iguk-butŏ masi-da to EuroWordNet. mŏri-rŭl ŏn-da mŏri-rŭl ŏn-da in Korean is IE which means to get married to a man. The semantic information and wordnet of mŏri-rŭl ŏn-da made by KLSN are as follows: lemma= mŏri-rŭl ŏn-da The verb phrase mŏri-rŭl ŏn-da has one sense. 1. mŏri-rŭl ŏn-da, sijibga-da... (To get married to a man, At last, she tied the knot when she 48

62 WORDS IN ASIAN CULTURAL CONTEXTS turned 30 years old. ) Figure 2 Semantic relation of mŏri-rŭl ŏn-da in KLSN But there is no sense number responding to the sense of mŏri-rŭl ŏn-da in the ILI-record of EuroWordNet, which means recorded senses in the ILI-record of EuroWordNet are based on the WordNet of English. If that was the case, there are two possible solutions to find the closest sense number of mŏri-rŭl ŏn-da in ILI-record. The first possible solution would be to find a corresponding sense for a hypernym of mŏri-rŭl ŏn-da in ILI-record. A second possible solution would be to find a corresponding sense for a coordinate of mŏri-rŭl ŏn-da in ILI-record. Fortunately, there are corresponding senses for gyŏlhonha-da (marry) as hypernym of mŏri-rŭl ŏn-da and corresponding senses for janggaga-da (wife) as coordinate of mŏri-rŭl ŏn-da in ILI-record. Looking at the English Wordnet, {marry, get married, wed, conjoin} is corresponding to senses of gyŏlhonha-da and {wife} is corresponding to those of janggaga-da. In order to indicate the closest sense number to sense of mŏri-rŭl ŏn-da in ILI-record, either one of senses, which are gyŏlhonha-da and janggaga-da, need to be selected. If there are hypernym and coordinates as candidates, when searching the 49

63 ASIALEX 2005 closest sense number, hypernym should be selected first. This is because to select a hypernym can make less mistranslation than selecting a coordinate. [Hypernym choice condition]: The sense of IE has no corresponding sense directly in ILI-record and if there are both a hypernym and a coordinate, search the first sense number corresponding to a sense of hypernym in ILI-record. a. Susan-ŭn olhae gaŭl-e mŏri-rŭl ŏn-ŭl gŏ-ya. Susan will tie the knot in this spring. b. Susan-ŭn olhae gaŭl-e gyŏlhon-ha-l gŏ-ya. Susan will get married in this spring c. #Susan-ŭn olhae gaŭl-e janggaga-l gŏ-ya. Susan will get married to woman in this spring. mŏri-rŭl ŏn-da is [+MARRIED, +WOMAN] and a coordinate word janggaga-da is [+MARRIED, +MAN]. Susan which is [+WOMAN] and janggaga-da can t be together in the culture area where people can t accept marrying a person who is of the same sex. gimch iguk-butŏ masi-da In Korea, the sentence gimch iguk-butŏ masi-da is IE which has the meaning to feel pleased by guess in advance. The semantic information and word network made by KLSN are as follows: Lemma=gimch iguk-butŏ masi-da. The verb phrase gimch iguk-butŏ masi-da has one sense. 1. gimch iguk-butŏ masi-da ( to feel pleased by guess in advance ; Nobody thinks to give you rice-cake ; don t count your chickens before they re hatched ) Korean IE, {gimch iguk-butŏ masi-da1} is a gloss of to count one s chickens before they re hatched and it can be added onto the list of ILI record but it s hard to find a word included in Synset in Korean because this gloss has a very particular meaning. Therefore, there is no corresponding sense to the direct meaning of {gimch iguk-butŏ masi-da1} in the EuroWordNet link system. So we have the only one way which is to try to connect indirectly to ILI-record through the hypernym relation in Korean. Figure 3 Semantic relation of gimch iguk-butŏ masi-da in KLSN 50

64 WORDS IN ASIAN CULTURAL CONTEXTS A hypernym, {yedanha-da1}, and Synset {gippŏha-da1, jŭlgŏwŏha-da1} in KLSN can link with {gimch iguk-butŏ masi-da1} and there is causation between these Synsets. However, the problem is that {gimch iguk-butŏ masi-da1} should be connected to through both the meaning of {yedanha-da1} and Synset {gippŏha-da1, jŭlgŏwŏha-da1} in order to connect {gimch iguk-butŏ masi-da1} to ILI in EuroWordNet. According to this, there would be different corresponding expressions. Figure 4 ILI-Linking of gimch iguk-butŏ masi-da in KLSN According to the above figure, in order to connect the meaning of {gimch iguk-butŏ masi-da1} to EuroWordNet through a hypernym which is { yedanha-da1}, the meaning of {gimch iguk-butŏ masi-da1} is connected with the sense 1 of predetermine in ILI_ record and {yuduan1} in Chinese. On the other hand, in order to connect the meaning of {gimch iguk-butŏ masi-da1} through {gippŏha-da1, jŭlgŏwŏha-da1} as a hypernym to EuroWordNet, the meaning of {gimch iguk-butŏ masi-da1} is connected with sense 2 of happy in ILI_record and {xihuan1} in Chinese. Figure 5 Semantic relation of gimch iguk-butŏ masi-da in KLSN 51

65 ASIALEX 2005 But if the corresponding expression is chosen as the meaning of {gimch iguk-butŏ masi-da1} by connecting to EuroWordNet, there is one which can meet their meaning, but the other cannot. To solve this problem, we suggest three possibilities: 1) There could be supersense and meaning-net including both hypernyms which are different each other. 2) Only one of hypernyms could be chosen. 3) New semantic relations with using both semantic relation of hypernym are showed. In the case of the first possibility, there are two difficulties. The first difficulty is that there is no supersense including both two senses of hypernyms in general. The second difficulty is that if there are supersenses, which consist of more than two levels of hypernym sense, it s hard to make the semantic relation, because the meanings are too universal. In the case of second possibility, we don t have any standard to choose one between two senses of hypernyms objectively. We can only decide it by the subjective judgement of the native speaker or by the frequency of the semantic use of each sense. In such cases, sense { gimch iguk-butŏ masi-da1} and ILI-record {predetermine1} or sense { gimch iguk-butŏ masi-da1} and {happy2} can be connected as a new semantic relation which is Eq_near_hypernym. In the case of the third possibility, the issue is spreading the ILI_record of EuroWordNet. We suggest EQ_Diathesis for it. There is no change of core meaning of word but if there are Synset which can be clarified their meaning by pattern of verb(intransitive/ transitive) or Causative relation or Causation etc., after making one group optionally through EQ_Diathesis and then connect them with Synset of separate language. The Following Table shows an example that ILI connection of {gimch iguk-butŏ masi-da1} consisted of the EQ_Diathesis relation.(ili numbers are given optionally just for an example so they are different from real numbers of ILI.) ChWN {yuduan} ILI-reference {xihuan} ILI-reference ILI-record (WN 2.0) predetermine sense sense happy sense sense sense KLSN {gimch iguk-bu tŏ masi-da1} ILI-reference Figure 6 ILI relation of gimch iguk-butŏ masi-da1 52

66 WORDS IN ASIAN CULTURAL CONTEXTS [Plural hypernyms choice condition 1] The sense of IE has no directly corresponding sense in ILI-record and if there are more than two hypernyms, the relation of ea_near_hypernym is given after searching for sense numbers which come under the sense of hypernym in ILI-record. [Plural hypernyms choice condition 2] The sense of IE has no directly corresponding sense in ILI-record and if there are more than two hypernyms after registering the new sense in ILI-record and if pattern of verb (intransitive/ transitive) or Causative relation or Causation are showed as semantic relations between hypernyms, it searches for sense numbers for IE with a relation of diathesis. 5. Conclusion This study focused on problems of connecting with EWN, which is a Multi-lingual Database using ILI and the semantic relation formulation of IE in KLSN. Since IE includes rich cultural meanings of the target language, it can cover the meaning which cannot be represented with a word(s) in a certain language system. For that reason, the absence of the meaning and the lack of correspondence between words often occur when one language system is corresponded with the other. In this paper, we attempt to solve the problem of IE by applying various semantic relations of EuroWordNet which happened especially when KLSN connects to EWN. References Copestake A. ed. (1995). Acquisition of Lexical Translation Relations from MRDs. Journal of Machine Translation, Vol. 9, issue 3. Fellbaum, C. (1998). Towards a representation of idioms in WordNet. In: Proceedings of the COLING/ACL Workshop on Usage of WordNet in Natural Language Processing Systems, Montreal. Fellbaum, C. ed. (1998). WordNet: An Electronic Lexical Database, Cambridge, The MIT Press. Hong, Chai-song (1993). Description lexico-syntaxique des phrases figées à meokta (manger). Language Research Lee, Chungmin (1993). Frozen Expressions and Semantic Representation. Language Research Lee, Donghyeok (2005). Computational Processing of Grammatical Idiomatic Expression. Korean Linguistics 26. Oh, Janggeun (2004). A study on ILI_Model for Multi-Lingual Information Retrieval. The Journal of Linguistic Science 27. Vossen, P.C. ed. (1998). EuroWordNet: A Multilingual Database with Lexical Semantic Networks, Kluwer Academic Pub., Dordrecht. Vossen, P. (1998). Introduction to EuroWordNet. Computers and the Humanities, Special Issue on EuroWordNet. 53

67 ASIALEX 2005 One database, many dictionaries varying co(n)text with the dictionary application TshwaneLex Gilles-Maurice de Schryver Ghent University & TshwaneDJe HLT David Joffe TshwaneDJe HLT Abstract This paper provides background information for a software demonstration of TshwaneLex, during which the actual use of the application is illustrated in real time. The focus of the demonstration is on two main aspects, together with a related aspect in each case, of particular interest to the ASIALEX 2005 conference. These are full Unicode support and customisable sorting on the one hand, and advanced DTD (Document Type Definition) aspects and Linked View mode on the other. Together, they provide the backbone for the claim that a single TshwaneLex database successfully provides for multiple dictionaries. The dictionary compilation software TshwaneLex TshwaneDJe HLT has been producing the dictionary compilation software TshwaneLex since In addition to most of the eleven South African National Lexicography Units who are currently using TshwaneLex both for the compilation itself of their (monolingual) dictionaries, as well as for the presentation of their results on the Web this software is now also used at a number of widely respected dictionary publishing houses such as Oxford University Press, Macmillan and Van Dale Lexicografie. Government-sponsored research centres, such as the Royal National Academy of Medicine in Spain, have also begun to build their latest reference databases around TshwaneLex. Copies of TshwaneLex have furthermore also been acquired by a variety of dictionary teams worldwide, who are compiling dictionaries for amongst others Lingála, Cilubà and Kiswahili (all spoken in Africa), Welsh, Irish and Estonian (all lesser-known European languages), Bai and Chinese (both spoken in China), Motu (an Austronesian language used in Papua New-Guinea), and Inezeño Chumash (a Native-American language from the US). Each of those languages needs its own script, and each of those projects needs its own dictionary grammar, both of which TshwaneLex provides for. A general introduction to TshwaneLex, with a focus on a selection of lexicographic underpinnings, may be found in Joffe & De Schryver (2004), while an example of an online application that revolves around TshwaneLex has been described in De Schryver & Joffe (2004). As pointed out in those publications, TshwaneLex contains numerous unique and highly developed lexicographic features. For example: An advanced cross-reference system not only shows related (incoming and outgoing) cross-references of the current lemma, but also automatically updates target homonym and sense numbers when these change. 54

68 WORDS IN ASIAN CULTURAL CONTEXTS A filter function not only allows the user to work with a subset of lemmas in the dictionary based on specified criteria, a dictionary text search function further enables complex search queries on that filtered section using Unicode regular expressions. A compare/merge feature visually displays differences between database versions, and allows changes to be selectively merged into the main database. In addition to paper, dictionaries can be published on the Web with the online dictionary module, which features a sophisticated query logging system. The localisable user interface allows users to browse the dictionary in their own language, and their preferred language may further be used to dynamically customise the language of the meta-language within returned articles. This feature is also extended to the electronic dictionary module. Full Unicode support Unicode, the international character set standard, is supported throughout TshwaneLex, and on all levels in the dictionary database. This allows not only the ability to enter data from virtually any language, but also even the simultaneous utilisation of both Asian and Latin characters in any attribute field in the database. For languages such as Chinese, Japanese or Korean, or say Arabic or Hebrew, data can be entered directly into TshwaneLex using any of the Input Method Editors (IMEs) available in Windows 2000 or XP. See Figure 1, which shows a screenshot of an elementary bilingual English-Chinese dictionary. Completely customisable sorting The default sorting method supported by TshwaneLex is a configurable four-pass table-based sorting system based on the ISO standard. The four different passes are used for various characteristics that may take precedence over one another, viz. the so-called base alphabet, diacritics, uppercase/lowercase differences, and so-called ignorable characters (typically non-alphabetic characters such as spaces and punctuation marks). This is shown in Figure 2, where the sorting tables have been configured for the Estonian alphabet. TshwaneLex automatically takes care of the sorting of lemmas, thus freeing the lexicographer from having to do so. However, many different methods of sorting exist, and often many even for the same language, thus the question arises as to how to support any possible sorting method that may be desired. To solve this, TshwaneLex includes an extendibility mechanism whereby users can create plug-ins to add support for new sorting methods. As a result, any sorting system (e.g. by radical/stroke count or by pinyin Romanised form for Chinese) may be used. 55

69 ASIALEX 2005 Figure 1. Unicode support for a bilingual English-Chinese dictionary in TshwaneLex Generating multiple dictionaries from a single database Elsewhere in this volume (cf. Joffe & De Schryver 2005) the main aspects of the customisable and multilayered DTD editor dialog are presented. Not only can the dictionary grammar for any project be flexibly configured and then kept under control with the built-in DTD, given that all elements and attributes are also linked to a comprehensive style system for generating the output (and preview), one single database can efficiently hold several dictionaries. Broadly speaking, this is achieved by doing two things: Firstly, by making use of multiple element categories to which the various data attributes are assigned by the lexicographer depending on which dictionary or dictionaries they should appear in, and secondly by defining a different set of styles for each view of the database, i.e. for each dictionary. Certain element categories are made visible or invisible in each style, which thus effectively functions as a kind of mask that filters and reveals only the portions of data to be shown for the current dictionary. Additionally, this also allows a different look to be defined for each dictionary. These features are illustrated in Figures 3 and 4, which respectively show the desktop and pocket editions of a bidirectional French-Dutch dictionary ( 2005 Van Dale Lexicografie). One hotkey allows the lexicographer to switch between the two views, and thus also the two dictionaries. The extent of co-text and context for the production of any particular dictionary may thus easily be decided on at the output stage. 56

70 WORDS IN ASIAN CULTURAL CONTEXTS Figure 2. Configuring table-based sorting for Estonian in TshwaneLex This feature may also be tied in with customising the language of the meta-language, as described earlier, potentially being used to customise aspects of the dictionary output further according to the language of the target user of the dictionary. For example, in a bidirectional Japanese-English dictionary, the information in some fields may inherently be primarily only useful to either a Japanese or English mother-tongue speaker. Lexicographers sometimes have to make editorial decisions and compromises based on assumptions about the language of the target market; by customising the output from a single database this need not be the case. In an electronic dictionary, one could take still other factors into account, such as the level of the user, presenting different views of the dictionary to beginner or advanced language learners. Linked View mode for bilingual dictionary editing Several innovative functions assist in bilingual dictionary compilation, such as side-by-side editing, automated reversal and Linked View mode. When in side-by-side editing mode, the screen is split in two down the middle, and the lexicographer can work on either side of a bilingual dictionary by simply moving between the windows. When in Linked View mode, as in the case of Figures 3 and 4, related articles in the reverse side of a bilingual dictionary are automatically displayed. For instance, from the left-hand side of Figure 3 one sees that the Dutch words bagagedepot, statiegeld, lege fles, instructie, kwartierarrest and (het) nablijven have been used as translation equivalents for the French word consigne. 57

71 ASIALEX 2005 Figure 3. Desktop-edition view of a bidirectional French-Dutch dictionary in TshwaneLex, from the same database as the pocket edition When in Linked View mode, TshwaneLex automatically shows all and only those articles that have these translation equivalents as lemma signs, in this case bagagedepot, statiegeld, instructie and nablijven as may be seen from the right-hand-side of Figure 3. The Linked View mode feature thus allows the lexicographer to attempt to honour the reversibility principle, that is, the condition whereby all lexical items presented as lemma signs or translation equivalents in the X-Y section of a dictionary are respectively translation equivalents and lemma signs in the Y-X section of the dictionary (cf. e.g. Tomaszczyk 1988: 290; Gouws 1989: 162; Gouws 1996: 80). The reversibility principle has always been a crucial but hitherto little-looked into requirement in lexicography, now easily made accessible in TshwaneLex. 58

72 WORDS IN ASIAN CULTURAL CONTEXTS Figure 4. Pocket-edition view of a bidirectional French-Dutch dictionary in TshwaneLex, from the same database as the desktop edition References De Schryver, Gilles-Maurice and David Joffe (2004), On how electronic dictionaries are really used, in Geoffrey Williams and Sandra Vessier (eds), pp Gouws, Rufus H. (1989), Leksikografie (Pretoria: Academica). Gouws, Rufus H. (1996), Idioms and collocations in bilingual dictionaries and their Afrikaans translation equivalents, Lexicographica: International Annual for Lexicography 12: Joffe, David and Gilles-Maurice de Schryver (2004), TshwaneLex A state-of-the-art dictionary compilation program, in Geoffrey Williams and Sandra Vessier (eds), pp Joffe, David and Gilles-Maurice de Schryver (2005), Representing and describing words flexibly with the dictionary application TshwaneLex, in ASIALEX 2005 proceedings. Tomaszczyk, Jerzy (1988), The bilingual dictionary under review, in Mary Snell-Hornby (ed.), ZüriLEX 86 proceedings, papers read at the EURALEX international congress, University of Zürich, 9-14 September 1986 (Tübingen: A. Francke Verlag), pp TshwaneDJe HLT ( ), Online info, (accessed: 31 March 2005). TshwaneLex ( ). Online info, (accessed: 31 March 2005). Williams, Geoffrey and Sandra Vessier, editors (2004), Proceedings of the eleventh EURALEX international congress, EURALEX 2004, Lorient, France, July 6-10, 2004 (Lorient: Faculté des Lettres et des Sciences Humaines, Université de Bretagne Sud). 59

73 ASIALEX 2005 Pressing names 1 : creating meaning in a fictional dictionary Jennifer Eagleton Macquarie University jenny@asian-emphasis.com Abstract This paper will show how the prescriptiveness of language is turned on its head in the fictional work, Dictionary of Maqiao by contemporary Chinese writer Han Shaogong. Using the English translation of this novel, which tells of his dislocation to the countryside during the Cultural Revolution, this paper will look at how it stylistically and grammatically subverts the dictionary form yet plays on our pragmatic reading of dictionaries as authoritative. It is hoped that this paper will help open the door to a general discussion on dictionary definitions and the difficulty of pressing names. Words have lives of their own. They proliferate densely, endlessly transform, gather and scatter for short bursts, drift along without mooring, shift and intermingle, sicken and live on, have personalities and emotions, flourish, decline and even die out. (Han, 2003: 319) 2 Han attempts to show in his work that words and their meanings do not consist of fixed, controllable entities, and that words are frequently manipulated, misrepresented, and re-invented. As the novel makes clear, words can often take non-envisioned connotations. This is ironic when one considers that he has chosen the dictionary form, which by definition is an authoritative static work of reference, to create a work of fiction, which is defined in The American Heritage Dictionary of the English Language as an imaginative creation or a pretence that does not represent actuality but has been invented. In this way he is opposing from the outset the futility of word defining. Words here really do drift without mooring : the text smoothly ebbs and flows, the words being the main protagonists of the story rather than humans. With this in mind, the concept of semantic drift arises, or more accurately, semantic direction (Butt, 1983; Hasan, 1995) that is, meaning that is highlighted by the foregrounded patterns converging toward the same direction which in Dictionary of Maqiao lies in the sense that there is no fixed meaning. The structural elements of the dictionary form used in this work of fiction greatly inform our reading of it both in a pragmatic and formal sense. Secondly, I will illustrate this by referring to two short extracts that highlight the particular lexico- semantic devices that create the ebb and flow of the text. Since the text stresses the constant transformation of words and their role as protagonists, the experiential metafunction of systemic functional grammar will be used to look at how words change their meaning in this work. Background to the Novel The author, Han Shaogong, was a youth at the time of the Cultural Revolution and was sent down to the countryside as an Educated Youth (zhiqing) to the town of Maqiao (an actual town) in Hunan Province, where he had to struggle with a new life and new ways of expression. During this time, the Maoist regime made strenuous 60

74 WORDS IN ASIAN CULTURAL CONTEXTS efforts to regulate language through manuals dictating correct forms of rhetoric, grammar and characterisation so that it could control literary production (McDougall, 1980). In other words, they developed their own lexicon. After the Cultural Revolution, Han and other writers struggled to free themselves from the linguistic confines of previous years. This literature was called root-seeking literature (to use a dictionary analogy, an etymology of who they were.). You could almost say that Han, by creating his own dictionary, is showing a form of resistance to the rigid control of words that was a feature of his youth. One Man s Dictionary: Subversion of the Dictionary Form A dictionary is a reference book containing an alphabetical list of words, with information given for each word, usually including meaning, pronunciation, collocations, etymology and situational use. Thus a word is a highly complex linguistic phenomenon and a dictionary is supposed to alert us to the problems regarding many aspects of its use. In the Dictionary of Maqiao, the cover suggests a reference work: it has a preface, and an index and it has entries of various lengths; there is a content page suggesting to the reader that it can be dipped into at random. However it really needs to be read in a linear fashion as the entries start to assume knowledge of dialect words and of characters already introduced. In this dictionary, not only words, but concepts, people and animals are defined. The given definitions range from brief vignettes to lengthy sequences and are almost like short stories. Unlike most dictionaries that are supposed to be neutral in language with no obvious narrator or spatial dimension, Han as author inserts himself in the text and talks about time past and present. In this way it subverts its authoritativeness and objectivity to the reader. The most critical part of the front matter is the Editorial Note, in which the purported editor (really the author Han Shaogong) states that the Dictionary of Maqiao is a unique category of its own as it is an experimental undertaking. The deitic we, however, is actually an I: We received this offering from the dictionaries compiler, Han Shaogong, a renowned gentleman of letters whose oeuvre includes and a host of other hugely influential works, and whose mighty skills in penmanship extend to both fiction and essays; not, however, to dictionaries. But having considered the specialized content of this dictionary as well as the opportunity that a lexicon affords for exploration and discussion, we encouraged this brave experiment and permitted him to retain his own distinctive literary style within the work. With this, he is given leeway to subvert the dictionary form, and for us to read it as a curious hybrid of dictionary, storytelling, vignette and anecdote. He is thus absolved from strictness and prescription. You could argue that while we know that this is a fictional work, he would know that we would have in mind our associations with dictionaries as a genre, especially the specialised-information dictionary. Many entries start off with a definition of a word, which is followed by a story to explain how this word means. Many of the definitions are actually found to be antonyms or bizarre interpretations of the usual dictionary denotation. In translation, if there is no close concept in the target language, that is the language of the translation, a near equivalent is found. Han seems to be saying to us that since I cannot tell you the correct meaning, I will tell you a story. 61

75 ASIALEX 2005 Antonymy as an overall structural framework The most striking thing about the Dictionary of Maqiao is its use of antonymy. Antonymy is a lexical-semantic relation that unites words with opposite and contrastive meanings and it has several characteristics that set it apart from other such relations. There seems to be something mystical about antonymy because opposites seem as different as they can possibly be yet they still seem to have something in common. This simultaneous closeness and distance can be partly explained by semantic dimension. If you know the exact antonym of a word, for not only will it give us a firmer grasp of the meaning of the word to which it is opposed, but inversely of itself (Egan 1968, 28a). This is what Han Shaogong plays on throughout the entire work. The opening paragraph from the entry for beginning (pp ), states: In Maqiao dialect, the word for end (pronounced wan in Mandarin) is pronounced the same as the word for beginning (yuan). Two temporal extremes are thus phonetically linked. In that case, when Maqiao people say yuan, do they mean end? Or do they mean beginning? You could consider antonyms to be gradable opposites with Time being the dimension, with beginning being the first base on the time continuum which ultimately finishes at the s end. Antonyms in the next except are linked by the conjunction and, in most cases, meaning together with or along with, in addition to; as well as again demonstrating the concept of antonyms as something lying on a continuum. This has the effect of shifting meanings backwards and forwards: Time always advances in a straight line, never repeating itself, with forward and back, this and that, right and wrong permanently in diametric opposition to each other, implying a certain standpoint for making comparisons. Even the protagonists are antonyms, History s optimists and history s pessimists in paragraph three of beginning being the noted example. In the middle of the piece we are posed the question, which is used in an almost a copula fashion: which yuan would Maqiao people choose? Beginning or end? Then the town of Maqiao is considered an antonym of itself: Maqiao has a great many stones and a great deal of soil, stones and you look, you won t see it changing. Every particle is a testament to eternity which is contrasted later with: One the other hand, Maqiao is not, of course, the Maqiao of former days, or even the Maqiao of a moment ago. A wrinkle has appeared, a white hair has floated to the ground, a withered hand has turned cold, everything moves silently on. Just as we think something is one thing, it becomes its opposite. Han says that the meaning has to be taken from your point of view. In the entry for resentment (pp ) the author frames our reading of word-change by baldly stating: Some words undergo a bizarre transformation once they pass into actual usage: their opposite meaning gestates and grows within until it bursts out of them, until they end up annihilating, totally negating themselves. They harbour shadows that are very hard to glimpse. Here the treatment of antonymy is more obviously stated in the next sentence, which is a relational clause: the hidden meaning of expose for example is in fact hide. Later on in the text it goes on to say, criticism is the hidden meaning of praise, again using two opposite concepts and parodying the relational clauses utilised in the 62

76 WORDS IN ASIAN CULTURAL CONTEXTS traditional dictionary format. The identifier and identified are at odds with each other. Under each of these definitions are stories to illustrate just how expose can in fact mean hide, and how criticism can mean praise. In some ways this is like a dictionary with its language usage notes, and the examples of situations in which the word is used. Having lives of their own: words as material agents The transitivity structure of a clause is concerned with who does what to whom, who it is done to and why (Halliday, 1994: 106). The process is affected by the participants and the circumstances involved setting the scene. Since words have lives of their own in the Dictionary of Maqiao, they act on their own behalf and the attempts of man to label them are futile. The entry for resentment deals with how words change of their own accord and this is why material clauses tend to dominate, while in beginning relational clauses dominate since here it is humans that have given them labels according to their particular point of view. Material processes are more like doing in the normal sense of the word, they are concrete changes that can be perceived and they can also serve as a model for construing our experience of change in abstract phenomena (Martin et al., 1997: 103). The one inherent participant is the Actor and it does not necessarily have to have the aspect of consciousness. However in the entry, resentment, they are almost given a consciousness of their own. The process of word transformation is a violent one, in that words gestate and grow, burst out and end up annihilating and totally negating themselves as mentioned on page three above. They (i.e. words and their transformations) pile up on humans, crushing them. This violent birthing process was one that, in the past, men like Mao Zedong and other authoritarian figures tried to harness for their own ends (McDougall, 1980). However, here the tables are turned. It is the people who are powerless; they almost know what is going on; but they are not acting, they remain passive. Therefore, most of the phrases that deal with actual human protagonists are mental or behavioural ones they realise, they watch, they ponder on the words as they change form. One gets the sense of helplessness, as humans appear as mere onlookers to this change. Resentment. In resentment, words are the main protagonists and actors. And this effect is intensified by the fact that Actors/subjects are few; the main players are words and their meanings, for example, some words, their opposite meaning or they are referred to anaphorically as they or them. There is also use of specific words such as praise and criticism, being the physical entity of the word (interestingly these are used in relational identifying clauses). The words act on themselves first and foremost: carrying within themselves thei r meaning. They self-transform. The material verbs used in conjunction with word-as-subject are almost like phrasal verbs as they are linked with either with verbs-in-series ( shock and stun, yawn and yawn and enables words to shift mooring and endlessly transform ) while having a physical effect on the people who read, right and speak them. Words are moving from an inner existence to being out in the world where they have lives of their own according to different circumstances, which are usually related to their location ( [ wo rds] coming out of your ears.) This also gives a sense of drift, of movement in the text. This also, in som e respects, is like the list of various meanings in a regular dictionary (meanings a, b, c and d, and when they should be used. 63

77 ASIALEX 2005 After this birthing process a relational clause is foregrounded: its new identity has been formed (to a certain degree) and they have power now that their meanings can be opposite, which the author emphasizes by the words for example and in fact in the phrase the hidden meaning of expose for example is in fact hide ; you could almost relate this to the Maoist era when things were read between the lines and where indeed hide was in fact the meaning of expose. Beginning. In the entry, beginning, it is not so much that words are changing themselves but people are trying to make sense of them, or more accurately trying to pin meanings that do not necessarily fit. At first, a series of relational clauses give us the identity of the word, but the linking of one thing to its opposite meaning subverts it and causes confusion in the people using it and perhaps to us as readers too. When Han says when Maqiao people say yuan, do they mean end?, it makes sense that verbal or behavioural clauses follow the identity-establishment of the word, since words are primarily verbal tools. Meaning here can be classified as a verbal use rather than a mental one since you could argue that meaning is equivalent to what they are really saying is, while Han Shaogong himself wades into the debate with his own view of how history s optimists and history s pessimists thinks what beginning means a little bit later with the authorial I in as I see it, a behavioural clause, a bodily action that really is a kind of mental one. What is said or insisted on keeps changing (is it this? Or is it that?) and what is viewed or seen is perhaps not what is actually there. This is highlighted by the second-last and last paragraph in which we see Maqiao town as seemingly an unchanging entity ( consider Maqiao ), but on the other hand it is not so (see also my discussion on page 3 above). As mentioned previously, humans rather than words are the protagonists here, they being history s optimists and history s pessimists, Han Shaogong, Maqiao People, and perhaps us. They may be behaving, saying and thinking certain things, but they still seem passive, perhaps because they are onlookers and not really actors. When we are asked to consider Maqiao towards the end of the entry, Maqiao itself is defined and also acts apart from the existence of humans in its environs. In addition, the juxtaposition of opposing views (of history s optimists and history s pessimists) in a parallel structure in long clause complexes also makes this decision-making difficult. Maqiao, acting apart from its human inhabitants, is later dissected into its component parts: sand, stones, withered hands etc. with these items undergoing their own physical transformation (just like words do) and we are told that no power can stop and no power can prevent this change; while things appear and fade, we get a sense of a testament to eternity but that is really only on the surface this could relate to everything about life words and also actions will disappear with time. Going with the flow: above and around the clause The text really does undulate and move like ocean current with various aspects held in its slipstream. The limited number of protagonists, placed in parallel structures, either of repetition or of antonymy help to create this effect. The use of ellipsed subjects, too, helps us to focus on the long line of processes in series and carries the stories onwards. There are also many long clause complexes in a paratactic relationship of coordination. In principle, paratactic relations are logically symmetrical with various elements being of equal status. This is exemplified by the conjunction and a kind of oh, by the way, it could be also this as well which is a particular feature of the entry 64

78 WORDS IN ASIAN CULTURAL CONTEXTS for beginning. However hypotaxis is also used to effect such as in the second paragraph of beginning when the subordinating structure if then... is used in parallel with another to highlight diametrically opposed concepts (with each aspect of the respective sentence being the polar opposite of the other). Punctuation helps us define parataxis and it is worth noting that dictionaries frequently use various punctuation marks such as the colon and semi-colon to mark off definitions. In this case, the use of these marks helps the drift or flow of the text onto the next point, it enables us to pause, to catch our breath and ponder on the implications of what the author has said as one continuing changing thought. Conclusion Such a brief overview cannot do justice to such a dense work. Its clever use of semantic and lexical devices show us that words indeed are slippery creatures and that we cannot always domesticate them. Meaning as a political statement has always been used by various individuals and groups throughout history. Han Shaogong, by subverting both the normal narrative structure of fiction and the authoritative nature of reference works firmly opposes set guidelines or dogma on thinking, art and literature. All our meaning-making is thus somewhat arbitrary and despite our attempts at trying to press names on them. As Han concludes that this, is of course, my own individual dictionary, it possesses no standardising significance for other people (Han, 2003: 319). This is something that dictionary compilers should think about when selecting words their next volume. Note 1. Pressing names is a definition in Dictionary of Maqiao (pp ), to denote a ritual carried out to formalise entry to a cl an after the funeral of an adopted father. In this ritual, the adopted relatives sang the long list of genealogical names to the person entering the clan to prevent the individual taking any wealth back to their original family later on. Adopted children had no inheritance rights before they did this, or pressed names. For me, the association with pressing names relates to the fixed de finitions given in dictionaries. 2. All subsequent references to this work is taken from Han Shaogong (2003), Dictionary of maqiao, trans. Julia Lovell (Columbia Press, New York). The original Chinese version was published in References Cruse, D. A. (1986), Lexical semantics (Cambridge: Cambridge University Press). Egan, R. F. (1968), Survey of the History of Antonymy, in P.B. Grove (ed), Webster s new dictionary of synonyms (Springfield, MA: Merriam-Webster), pp. 5s-31a. Halliday, Michael (1994), Introduction to functional grammar (UK/USA: Edward Arnold Press). Martin, J. R. et al. (1997), Working with functional grammar (London: Arnold). McDougall, Bonnie S. (1980), Mao Zedong s Talks at the Yan an conference on literature and art : a translation of the 1943 text with commentary (Ann Arbor: Center for Chinese Studies, University of Michigan). McDougall, Bonnie S. and Kam Louie (1997), The literature of China in the twentieth century (Hong Kong: Hong Kong University Press). 65

79 ASIALEX 2005 Abstract Verb-to-noun argument-sharing Charles Fillmore University of California, Berkeley The direct valents of a valence-bearing word in English - shown in brackets in the following - can be found inside the phrase headed by the word (fond [of babies], argue [that the world is flat], decision [to retire]), or in a subject or possessor relation to the word ([my] decision to retire, [Lucy] argued that the world is flat). Indirect valents can be found for verbs and copula-supported adjectives in Control or Raising contexts - and here we might include the subjects of auxiliaries ([Lucy] decided to retire, [she] has argued that the world is flat, [the stranger] seems to understand us, [we] are unlikely to be successful), as well as in easy -type constructions where the shared argument in the embedded expression is a non-subject ([the passage] was hard to understand, [this book] is worth reading). In this paper I will concentrate on how governing verbs share arguments with those of a directly or indirectly governed noun. The simplest cases will be that of the so-called light verbs, whose semantic contribution is minimal but which at least serve the role of providing a slot for one or more of the valents for the frame-bearing noun (FBN), and allowing tense and aspect information to be associated with the frame evoked by the noun. Examples, with the light verbs italicized, are [I] have a headache, [she] gives [me] good advice, [the president] made an announcement, [we] should say a prayer, [they] sang a song, etc. For these sentences it is important to realize that it is the noun, not the verb, that evokes the conceptual structure - the frame - which is at the center of the clause. The more interesting cases are those in which the governing verb of an FBN contributes semantic structure on its own, in addition to providing an understood valent of the noun. Following the lead of such authors as Laurence Danlos, Maurice Gross, and Igor Mel cuk, I wish to examine the full roster of such structures for English, provide definitions and criteria for their identification, explore their relationships with familiar control structures, and possibly offer brief comments on analogous phenomena in other languages, in particular French, German and Japanese. We will identify the syntactic patterns within which support constructions occur, including existential structures, where the supporting word does not provide a shared valent (there was a riot, a riot ensued; there is a relationship, a relationship holds); frame-bearing noun (FBN) as object (take a bath, deliver a lecture, have respect for), FBN as second object (give someone a scolding), FBN as prepositional object (be at risk, be in possession of), FBN as prepositional object of transitive support verb (hold you in contempt). The syntactic details will concern the difference between count and non-count property in the FBN, the presence or absence, and role, of determiners or possessives, the possibility of modification, the ellipsis potential of the nominal containing the FBN, etc. 66

80 WORDS IN ASIAN CULTURAL CONTEXTS The study will explore the value of these constructions for teasing out the full frame structure of the FBN nouns. The frame for explain and its nominalization explanation has a communicative sense and a cognitive sense. In the former there are four essential valents: Speaker (the one doing the explaining), Experiencer (the person who might come to understand whatever it is that needs to be explained), the Mystery (that is, the phenomenon which needs to be explained), and the Explanation proper (the account or story which interprets the mystery). In the cognitive sense, as [His drunken state] explains [the way he behaved that day], the subject counts as the Explanation, the object is the Mystery, the Experiencer is unexpressed, and there is no understood Speaker. We can now consider the valent-sharing functions of the verb need in respect to this frame. In [you] need to explain this, where the frame-bearing word is a verb, the subject of need is construed as the Speaker in the explanation frame; in [this] needs explanation, where the FBN occurs as a non-count noun, and without an article, the subject of need is construed as the Mystery; and in [The survivors] need an explanation, with an indefinite article, the subject of need is understood as the Experiencer. (In the case of an explanation, the implication is one-way: one can also say [This phenomenon] needs a better explanation than that, where the subject is the Mystery.) The lexicon-building and text-annotation parts of the FrameNet project ( are seeking to develop efficient ways for representing such information in both the properties of lexical entries and in mechanisms for semantic composition. Second-language learners and academic language: how can teachers and dictionaries help? Abstract Lily Wong Fillmore University of California, Berkeley wongfill@berkeley.edu For second-language learners of English, both the language of textbooks and teacher-talk in school include words and formations not easily learned through conversations with peers and other speakers of the language. The learner s task is to identify the part of an expression which is unfamiliar, to recognize its salient context, and to know how to follow the teacher s instructions to Look it up in the dictionary! Some of the locutions found in academic texts use familiar words in unfamiliar contexts; some use unfamiliar words that have a special relation to their linguistic context and cannot be understood independently of that context. Since a major part of learning a new word is understanding its capacity to combine with other words and phrases in its context of use, educational materials, and the teachers who use them, need to be able to highlight those stretches of text within which the word s function can be explained. This 67

81 ASIALEX 2005 paper will examine the language of math and science textbooks and examinations aimed at middle-school students (ages 10-13), materials I have been analyzing over the past several years in my work on the learning of Academic English by school age children who learning English as a second language. I will discuss the performance of existing dictionaries in dealing with the words that many students find unfamiliar or difficult, and will propose ways in which teachers and print or electronic linguistic resources can assist in the second language learners mastery of the academic register of the language they are learning. Abstract Macmillan English dictionaries new dictionaries for a new era Gwyneth Fox University of Birmingham and Macmillan Pjacqueline@pansing.com Macmillan is a new entrant to the field of learners dictionaries. This allowed us to learn from what had gone before. As all credible dictionaries should be, the dictionaries are based on up-to-date corpora. There is a general corpus of 200 million words made up of books, newspapers, magazines, speech etc from Britain, the United States and Australia. A learners corpus has also been built, allowing the lexicographers to see the typical errors of advanced learners, and the findings from this have informed both individual entries and longer usage notes. A corpus of ELT books has also been built, ensuring that the language of the classroom is fully covered. And finally, for the time being at any rate, for the Macmillan School Dictionary we built a 20 million word corpus of school text books used in countries where English is the medium of instruction. Using these corpora has allowed us to focus on the words that are typically used at different levels of language learning, and also on those meanings that are relevant in different learning environments. The Macmillan School Dictionary was a particularly interesting project, and I shall focus much of my attention on it during this talk. Abstract Metaphor in learners dictionaries Gwyneth Fox University of Birmingham and Macmillan Pjacqueline@pansing.com Language as it is taught in the classroom is overwhelmingly literal and concrete. Yet 68

82 WORDS IN ASIAN CULTURAL CONTEXTS language in real life is much more abstract, figurative, and metaphorical. In particular, English is full of institutionalised non-literary metaphors, which we use frequently without necessarily being aware of what we are doing. This can be confusing for learners who know only the literal meaning of words and phrases. It is therefore important to show learners how important metaphor is in the language and why they need metaphorical competence as well as their other competences. One reason is that metaphor is culturally determined and often culture specific. Although there are universal metaphors for example, those that draw on the concepts of movement and of up and down many conceptual metaphors central to English are not found in other languages and are therefore not easy to understand. It is important to realise how productive metaphor is and how frequently it is used in, for example, academic writing, journalism, and even conversation. Until recently dictionaries have dealt scantily with metaphor and indeed figurative language in general and have not helped users of the dictionaries to see the metaphoric patterns in English. As is the nature of a dictionary, words have been discussed in isolation and thus important facts about the language have not been expressed. This paper will discuss some of the major conceptual metaphors in English and will show how dictionaries, particularly learners dictionaries, are starting to deal with them. Abstract Learner perspective for effective dictionary skill training and lexicography Tetsuro Fujii Jikei Medical University, Japan t.fujii@jikei.ac.jp For lexicographers, the purpose of studying language is to improve a dictionary, but for EFL learners, the purpose of using a dictionary is to improve their language competence. H owever, for some instructors, actively discouraging students from using a dictionary while reading extensively is a common practice. Moreover, teacher-researchers often postulate that not using a dictionary is an indication of students improvement in their language learning strategy. Thus, there appears to be a wide gap among lexicographers, learners and instructors. Of course, the re have been attempts to bridge this gap by compiling a user-friendlier dictionary a nd attaching dictionary exercises for mastering the use of the dictionary. However, according to a preliminary survey conducted by the author, now that the majority of students use some forms of electronic dictionaries instead of a paper dictionaries, the pressing need is to provide dictionary use exercises, not just for one particular dictionary, but for a variety of dictionaries. At the same time, since the author believes English dictionaries are not yet optimally adapted to learners practical needs, their usability has to be objectively measured from the viewpoints of their users, in order for lexicographers to constantly improve the quality of dictionaries. In this paper, the author first reports the results of a survey on students perception of English learner s dictionaries, and second presents 69

83 ASIALEX 2005 dictionary skill exercises to accommodate the needs of EFL students, and finally offers lexicographers an investigative tool to measure the comprehensibility of dictionary definitions and example sentences. Computer-mediated language: a new challenge for dictionary-making? Wengao Gong National University of Singapore g @nus.edu.sg Abstract The advent of the Internet and computer-mediated communication (CMC) has brought into existence a new linguistic dimension computer-mediated language or Netspeak (Crystal, 2001). With the language of Internet in mind, Crystal holds that Netspeak takes the form of English and it is characteristic of its rich vocabulary closely related to Internet-restricted situations, operations, activities, and personnel. He also regards this aspect of Netspeak as a most creative lexical domain in contemporary English (2001:82). Of course, English is not the only form Netspeak can take. In fact, if we take a local perspective, we will find that Netspeak also takes the forms of Asian languages such as Chinese, Malay, and Japanese. The popularity of computer-mediated language worldwide raises the question of how lexicographers should treat lexis related to or originated from computer-mediated communication. The author examines the treatment of ten popular lexical items in computer-mediated English and Chinese respectively in the latest English and Chine se dictionaries an d finds that on the whole English dictionaries seem to be more responsive to new linguistic developments than Chinese di ctionaries. While this may be attributed to the fact that English is the dominant lan guage on the Internet, the author holds that this is more likely to be resulted from the fact that in C hina comput er-mediated language is still considered less orthodox than the language used in conventional media such as the press. Another contributing factor would be that English lexicographers have different perspectives about dictionary-making from their Chinese counterparts. Due to the fact that computer-mediated language has changed the linguistic environments worldwide, it is high time for lexicographers to face the new challenge posed by computer-mediated language and adjust their practice of dictionary-making. Introduction The rapid development and easy availability of information and communication technology have contributed considerably to the flourish of computer-mediated communication (CMC). The popularity of CMC has not only expanded our conceptions about communication but also blurred the boundary between spoken language and written language. As a result, there emerges a new linguistic dimension Netspeak in Crystal s term (2001). According to Crystal, Netspeak is not simply a new variety of 70

84 WORDS IN ASIAN CULTURAL CONTEXTS English, but a whole new medium, comparable to speech and writing in its distinctiveness and generality, and subsuming a great deal of linguistic variation (2003: 426). Of course, Netspeak does not exist in just one linguistic form English. In fact, if we take a local perspective, we will find that it can take many other forms such as Chinese, Malay, Korean, and Japanese. The popularity of this new medium has brought about many interesting changes to computer-mediated language. One aspect of these changes is the rapid increase in CMC-related lexis. How to treat these new linguistic developments has become a cursor of the lexicographers linguistic perspectives and their perspectives of dictionary-making. As English and Chinese are the two languages used by large population of Internet users, a sample analysis of the treatment of CMC-related lexis in the latest dictionaries in both languages will give us some rough idea about the current status of computer-mediated language in lexicography. Method In order to find out how CMC-related lexis is treated in current lexicographic practice and whether computer-mediated language has already been recognized as an important source of dictionary-making, the author selected ten English words and ten Chinese words and consulted them in three latest English dictionaries and three Chinese dictionaries respectively. The words consulted are listed in the lexis column of Table 1 and Table 2 below. The English dictionaries consulted are: the Chambers Dictionary (2002), the Penguin English Dictionary (2003), and Collins English Dictionary (Anderson, et al., 2004). The Chinese dictionaries are: Xiandai hanyu guifan cidian (Li, 2004) ( Standard Modern Chinese Dictionary), Xiandai hanyu xinci cidian (New Word Dictionary of Modern Chinese) (Wang, 2003), and Xiandai hanyu xinci cidian (New Word Dictionary of Modern Chinese) (Qu and Han, 2004). These dictionaries have been chosen for this study mainly because they are the only latest dictionaries available to the author. Two pieces of information about each word were sought for. One is about whether a word appears in the dictionary and the other is about whether the CMC-related meaning is cited. Findings and discussions The findings concerning the ten English words and their treatment in three mainstream English dictionaries are represented in Table 1. Table 1. CMC-related English words and their treatment in latest dictionaries Lexis The Chambers The Penguin English Collins English Dictionary (03) Dictionary 2nd (03) Dictionary (04) Online Offline Weblog X Blog X BBS X Chat X X Webcam X Spam Netspeak X X LOL X X As can be observed from Table 1, the three English dictionaries vary in their 71

85 ASIALEX 2005 treatment of the same CMC-related words. Eight out of the ten words can be found in Collins English Dictionary. In fact, the word BBS can also be found in this dictionary but it is listed as the abbreviation of be back soon, a phrase frequently used in Internet Relay Chat (a form of CMC). BBS usually stands for bulletin board system, another popular form of CMC where people exchange information through s. One striking feature of Collins English Dictionary is its inclusion of abbreviations such as LOL (laughing out loud) and BBS (be back soon). As a general-purpose mainstream dictionary, Collins Dictionary is a pioneer to include lexical items purely originated from computer-mediated language. The Penguin English Dictionary is also quite responsive to the new lexical developments in the English language. It has included seven out of the ten words, although it does not go that far as Collins to include abbreviations specifically associated with chatroom English. The Chambers Dictionary seems to be less adventurous than the other two. It has only included five out of the ten words which are more of technical colour. In fact, some of the five words included in this dictionary are very often used in CMC setting. For example, online and offline are used in ICQ, MSN, Skype, etc. to show the status the participants. Table 2 (see below) is a summary of the findings concerning the distribution of ten words which are related to computer-mediated Chinese in three recent Chinese dictionaries. From this table we can see a quite different picture from Table 1. All the three Chinese dictionaries tend to be rather conservative in including words originated from computer-mediated language. The first dictionary is the most authoritative among the three, but it has only included two of the ten words. That might be a reflection of its prescriptive nature, as its Chinese title (gui1fan4, meaning standard) suggests. The other two dictionaries have failed to include more words originated from computer-mediated Chinese either, though the titles of both dictionaries indicate that they are dictionaries for collecting new words in Chinese. One thing worth noting here is that all the examples about those CMC-related Chinese words included in these two dictionaries come from conventional media such as newspapers and magazines. Two inferences can be made from this phenomenon. One is that terms originated from computer-mediated language using are actually spreading to conventional media or discourse. The other is that lexicographers in China would only consider including those CMC-related words into their dictionaries after they have appeared in conventional media. Table 2. CMC-originated Chinese lexis and their treatment in latest dictionaries Lexis Meaning Xiandai hanyu Dangdai hanyu Dangdai hanyu Guifan Zidian xinci cidian ( 04) xinci cidian (03) zai4xian4 Online X X li2xian4 Offline X X X wang3you3 Web friends X wang3zhi4 Weblog X X X bo2ke4 Blogger X X mei3mei2 Pretty girls X X X shuai4ge1 Cool boys X wang3lian4 Web love X you2xiang1 box X X X zhuan3tie3 Quoted message X X X The different practices in selecting words between the English and Chinese dictionaries may be attributed to the diffe rent statuses of the two languages. As the 72

86 WORDS IN ASIAN CULTURAL CONTEXTS language of the Internet and a global lingua franca, English is the medium which carries most of the new vocabularies which are related to the IT industry in general and computer-mediated communication in particular. As a result, CMC-related vocabulary fits well with the English language itself. The impact of the Internet and computer-mediated communication on the Chinese language comes late but striking. Associated with western technology and western culture, computer-mediated communication in Chinese has generated quite a number of words and expressions which are often considered foreign, rebellious, and unconventional. This might have made Chinese lexicographers more cautious in deciding which words to include in their dictionaries. Of course, whether a new word or expression is to be included into a mainstream dictionary is mainly determined by the lexicographers perspectives of language and their perspectives of lexicography. In both cases, if they hold a prescriptive perspective, they will feel more comfortable to take the so-called conventional or established sources as the basis of lexis selection, thus being more cautious in absorbing new lexical developments into their dictionaries. If they adopt a descriptive perspective, they will tend to be more open-minded and would be more responsive to the latest lexical developments. The English dictionaries used in this research seem to be on the descriptive side, while their Chinese counterparts tend to be on the more prescriptive side. Here we are only talking about dictionaries for general purposes. Implications of the findings Like it or not, the advent of computers and the Internet has considerably changed our lin guistic environment. It is high time for lexicographers to re-examine some of the basic issues in lexicography in the light of an already changed linguistic and technological environment. One issue is of immediate relevance here, that is, the relationship between authority and representativeness in dictionary-making. Lexicographers are always in a dilemma when they come to the issues of authority and representativeness. As Hartmann rightly points out, the lexicographer (especially the one who sets out to document current usage) is faced with the awkward choice between uncommitted recording of everything (thus disappointing the appeal to the dictionary as an authority) or being deliberately selective (and at the same time abrogating his responsibility to scientific objectivity) (1983: 9). Nevertheless, dated from Samuel Johnson, dictionary-making has mainly been practiced along the prescriptive line. This practice has its own merits, which needs little explanation. However it also has its limitations. The biggest problem with such practice is that dictionaries so compiled can only reflect the stable, established features of the lexis. They can not reflect the dynamics of lexical developments in time. Being responsive to current developments is a very basic expectation from the part of the dictionary users. It will be too late if we only include certain words into a dictionary after they have already become an established part of everybody s language. After all, one of the most practical reasons why people consult dictionaries is that they want to look up for new or unknown words. A dictionary for general-purpose which has well-chosen entries and good definitions may not be very authoritative if its size is too limited. Inadequate c overage and slow responses to latest linguistic developments will definitely undermine the authority of a dictionary. In order to be quicker in response to recent linguistic changes, lexicographers should be more aware of the changes in people s use of 73

87 ASIALEX 2005 language and hold a descriptive perspective. Fortunately, we do have another school of thought in lexicography whose major interest is to describe what kind of words people are actually using in their daily communication and how they are actually using them. Of course, dictionaries may not necessarily be in hardcopy format. The ever-increasing availability of computers and the Internet has begun to change the ways how dictionaries are being published. The concept of digital dictionary can contribute to easing the tension between being authoritative and being representative. The following quotation from Fitzgerald (2000) helps to illustrate my point. It would be an ambitious pursuit, to say the least, for a dictionary to contain all of a language, but the elimination of spatial restrictions will enable dictionaries to more accurately and fully record a language. All manner of neologisms could be included, from the mainstream to the esoteric. This would be without paying the price of losing any of the historical nature of a dictionary, so that the digitised dictionary will be not only a more functional tool for the user, but also a better reflection of language (cited in Gibbs and Krause, 2000: 50). One condition for a dictionary to better reflect the lexical reality of a language is that the lexicographer keeps good watch on the new developments in language use and records them timely. Conclusion From this small research we can see that computer-mediated language is a new frontier of language use which has its unique features. Its existence and popularity has changed the linguistic environments we are in. Lexicographers should be fully aware of its existence and adjust their practice in dictionary-making so that they can better reflect the reality of our language use. References Anderson, Sandra, et al. (2004), Collins English dictionary. Glasgow: HarperCollins. Crystal, David. (2001), Language and the Internet. Cambridge: CUP. Crystal, David. (2003), The Cambridge Encyclopedia of the English Language. Cambridge: CUP. Gibbs, D. and Krause, K. L. (2000), Cyberlines: language and cultures of the Internet. Sydney: James Nicholas. Hartmann, R. K. (1983), Lexicography: principles and practice. London: Academic Press. Li, Xingjian. (2004), Xiandai hanyu guifan cidian (Standard Modern Chinese Dictionary). Beijing: Foreign Language Teaching and Research Press and Yuwen chubanshe Chinese Language Press. Qu, Wei. and Han, Mingan. (2004), Dangdai hanyu xinci cidian (New word dictionary of contemporary Chinese). China Encyclopedia Publishing House. The Chambers Dictionary. Edinburgh: Chambers Harrap The Penguin English dictionary. London: Penguin Wang, Junxi. (2003), Dangdai hanyu xinci cidian (New word dictionary of contemporary Chinese). Hanyu dacidian chubanshe (Chinese Dictionary Press). 74

88 WORDS IN ASIAN CULTURAL CONTEXTS The selection, presentation and treatment of cultural phrases in a multicultural dictionary Abstract Rufus H Gouws Universiteit van Stellenbosch, Privaatsak X1 Matieland rhg@sun.ac.za The user-perspective, so prevalent in modern-day lexicography, compels lexicographers to be well aware of the needs and reference skills of the target users of their dictionaries. Where a dictionary, monolingual or bilingual, is compiled for users living in a multicultural community, the lexicographer should carefully negotiate the selection, presentation and treatment of cultural items. This endeavour should not only be directed at the macrostructure of the dictionary and the selection of lexical items to be included as lemmata, but also at specific microstructural items. Phrases represent one of these types of microstructural items. This paper focuses on the way in which lexicographers select, present and treat cultural phrases in a multicultural dictionary. The emphasis is on the type of phrases that needs to be incl uded to empower the intended target user in his/her attempts to achieve communicative success for both text production and text recep tion as lexicographic functions. Attention will be given to various aspects of the structure of dictionaries. Utilizing the frame structure suggestions will be made for an interactive relation between the central list and the outer texts with thematically ordered cultural items to be included in a back matter text, ensuring a poly-accessible dictionary that can direct the user in an unambiguous way to the desired entries in the central list. Although the paper will not focus on a specific language the multicultural South African community with its different languages, including Asian languages, will be used to illustrate the underlying theory. Lexicon-based orthographic disambiguation in CJK intelligent information retrieval Jack Halpern The CJK Dictionary Institute, Japan jack@cjk.org Abstract The orthographical complexity of Chinese, Japanese and Korean (CJK) poses a special challenge to the developers of computational linguistic tools, especially in the area of intelligent information retrieval. These difficulties are exacerbated by the lack of a standardized orthography in these languages, especially the highly irregular Japanese orthography. This paper focuses on the typology of CJK orthographic variation, provides a brief analysis of the linguistic issues, and discusses why lexical databases should play a central role in the disambiguation process. 75

89 ASIALEX 2005 Interlingual lexicography, with special reference to research priorities Abstract Reinhard Hartmann University of Exeter & University of Birmingham Interlingual lexicography, which can be defined as being concerned with the design, compilation, use and evaluation of interlingual dictionaries, deals not with conventional unilingual, monolingual, or intralingual, but with bilingual, bilingualised, multilingual and polyglot reference works. In contrast to the former, the latter have been relatively limited, not only in their provision, but also in their reception as reflected by publications such as conference proceedings, textbooks and periodicals, by training facilities, and by academic research generally. This neglect will be reviewed and illustrated in the plenary lecture in terms of the six main perspectives of metalexicography: Dictionary history is devoted to the investigation of lexicographic traditions, practices and theories of various languages, cultures and countries. However, such studies are only available for the major world languages, and the coverage of specific language pairs varies considerably, being usually restricted to general-purpose bilingual dictionaries. Dictionary criticism is still an underdeveloped specialism. No universally agreed criteria exist for the evaluation of dictionaries, the few guidelines proposed for reviewing bilingual dictionaries are not widely known, and comparative critical accounts of different interlingual traditions are extremely rare. Dictionary typology has not adequately addressed the topic of classifying interlingual dictionaries. Although there is a growing range of genres available, we know little about their diversity across the language pairs. Dictionary structure deals with the presentation of various information categories (words, phrases, meaning, grammar, usage, names, etc.) in various reference works, but hardly any systematic studies exist of the translation-related complexities of interlingual reference works. Dictionary use has been the subject of a number of detailed observations, but not all dictionary types, user groups and reference skills (other than some of those related to foreign-language learners) have been addressed by appropriate methods and for the numerous language pairs around the world. Dictionary IT has revolutionised lexicographic processes, but the many new electronic reference products that are becoming available have not yet been fully assessed. Some specific issues requiring more empirical study, such as intercultural diversity, translation equivalence and directionality, are discussed in their relevance to interlingual dictionary-making for different language pairs, which in any case often lags far behind the progress made in unilingual lexicography. The lecture ends with a list of ten research priorities and the hope that these can be addressed soon in all parts of the world, within a wider reference science. [Please turn to page 393 for the full paper] 76

90 WORDS IN ASIAN CULTURAL CONTEXTS Some interesting lexical features of Singapore Colloquial English Abstract Jisheng He National University of Singapore The purpose of this study is to explore some of the issues that surround the lexicon of Singapore Colloquial English (SCE). It seeks to address the linguistic variation shown by SCE in terms of lexicon and show evidence of nativization of English words caused by contacting with other language varieties in Singapore (Wee, 1998; Ooi, 1998; Low & Brown, 2003). For example, in addition to their normal usage, some of the English words, such as cut, take and follow, can be used differently in SCE as in Your car cut mine ; I ll send you home (Ooi 1998). Such different uses, as pointed by Wee (1998), often lead people to consider if this variation is to be regarded as a part of SCE as acceptable nativization or if it has to be considered as an error. What we attempt to do, then, is not just to study each variation in its own right, but also to disentangle the interaction between the various types of variation. In the present study, by following Ni and Ler (2000:159), we take it as axiomatic that any lexical feature shown by SCE can provide us some useful information about the linguistic characteristics of this variety of English, and that such features can reflect the social uses of language (Chambers 2002:3). A character-driven three-layer network model of meaning and conceptual representation Shu-Kai Hsieh Universität Tübingen, Germany kai@hanzinet.org Abstract This paper represents a theoretical framework underlying an on-going project of a specific Chinese characters-driven enriched conceptual network (HanziNet). The main assumption we made throughout this framework is that the conceptual and semantic knowledge information wired in Chinese characters can be systematically described and made useful in performing NLP tasks as well. An integrated 3-layer pyramid structure network model is proposed to elucidate the relationship between character, word, concept and meaning. In conclusion, we believe that this model will shed light on polysemy, morphological derivation and compounding, and even idioms as ways of expressing concepts in Chinese. Introduction 77

91 ASIALEX 2005 Writing systems, especially the ideographic one, i.e. Chinese characters (Hanzi), have been neglected in current linguistic studies. The neglect of the writing could be attributed to many reasons. One of these reasons might be due to the prevalent surrogational model (Harris 2000) in contemporary linguistics which states, writing is to be viewed as a surrogate or substitute for speech, the latter being the primary vehicle for human communication. However, if we look at the only ideographic script one one of basic models of writing systems in the world namely, the Chinese writing system, we can see that in contrast to phonographic writing systems, it still displays a considerable amount of semantic information at the character level. Chinese characters have survived over thousands of years; some proposed that the whole set of Chinese characters can be viewed as an encyclopedia in essence, or in terms of knowledge representation, as a kind of ontological knowledge. This distinctive feature also suggests that the system of Chinese characters might contain rich but concise system of inter-related concepts. WordNet and HanziNet: A Fusion Figure1. The Pyramid Structure Model In the past few years, there has been a growing interest in the field of lexical semantics. Linguists and psychologists have been especially interested in the study of word senses in order to shed light on important aspects of human communication, such as concept formation and language u se. Research in lexical semantics is rather heterogeneous as far as scope, metho ds, and results are concerned, but it shares the same starting point: the word-based perspective. Central to a natural language processing system is also a word-store, the lexicon. In China however, there has been a different philological and lexical tradition. Before the 19 th century, most linguistic study in China focused on the characters: their forms, meanings and phonological structures. Under the influence of the modern linguistic theoretical trend, word hood in Chinese has thus become an issue of urgency and many studies have ensued. 1 1 There are, of course, contras. The famous Chinese linguist, Shuxiang Lü (1981:45) had strong opinions,...the reason why one cannot find a satisfactory definition for the Chinese word is that there is no such thing in the first place. As a matter of fact, one does not need the notion word in order to discuss Chinese grammar. 78

92 WORDS IN ASIAN CULTURAL CONTEXTS We are not going to fall into a fixed position with regard to speculating about whether a word constitutes a real or an epiphenomenal construct, nor about the acquisition of the lexicon in the context of Chinese. The questions we are interested in this paper are rather, how can we create an explanatory framework for clarifying the different ways in which concepts are lexicalized in Chinese? How are these lexicalized concepts and their relations structured? And how can they (if exists) and WordNet be fused to make an illuminating explanation of Chinese morphology and lexical semantics both from the viewpoints of theoretical and computational linguistic perspectives? Hanzi, Conceptual Primitives and Word Meanings: A Proposed Pyramid Structure Model The model I want to propose here is a Chinese characters-driven 3-layer network model with respect to the meaning and conceptual representation in the context of Chinese. It is also called the integrated Pyramid Model for Chinese processing, whose key aspect involves a fusion of natural language and conceptual information processing with Hanzi. In other words, it is aimed at bridging word-based natural language semantics and conceptual knowledge representation via the HanziNet. Thus, some enriched linguistic information for some Natural Language Processing tasks is obtained. The main underlying assumptions are as follows: (1) Concepts and Meanings are different. The Meaning of a linguistic expression overlaps with the Concept to a certain extent, though not necessarily totally. The main difference lies in that Concepts are more abstract and prima facie, while Meanings are mostly determined and used depending on the pragmatic, social and cultural context. These two sets constitute a bipartite graph. (2) Characters do not have concepts in themselves, but they are cues to concepts. In Chinese, the basic conceptual units are implicitly encoded in characters. This means, we can interpret the set of Hanzi as the set of concept primitives. In other words, characters = concept genes (or atoms) = the core and original meaning units. The concept of a Chinese word which normally consists of two or more characters is therefore a resulting construct of combination of two or more concept genes. (3) The meaning of a Chinese word, when using it in different contexts, often leads to ambiguity ( with different meanings ) and polysemy ( with various senses ), but it can be explained by two mechanisms: conceptual regulating and contextual constraints. Based on these assumptions, an integrated 3-layer Pyramid Model for Chinese processing is proposed, through which I hope to overcome the individual inadequacy of the lexical resources, and thus obtain some enriched linguistic information for NLP tasks. Figure 1 shows a schematic diagram of the proposed model. Formally, these networks have several levels of structural organization, with distinct scales respectively. From the top down, the first layer structure is concept chaos, the second layer is a tree-like Hanzi-driven TOP level Ontology, 1 and the bottom layer concerns with the long-range highly-correlated random acyclic graphic structure of the word-based semantic network. Each of these levels determines a distinct set of properties of the networks. The semantic processing of Chinese (compound) words can be seen as a process of concept-regulating semantic wandering within a semantic field (or network). 1 For the technical reason, the tree-like structure is not clearly depicted. 79

93 ASIALEX 2005 From the viewpoint of alphabetic-based linguistics, it seems enough to have the first and third layers (e.g., WordNet). Characters can be just treated as morphemes, which is still a controversy issue in Chinese linguistics. The design of a character network (called HanziNet) is aimed to fill the gap with two objects in mind: (1), to give each character in use a rigorous conceptual location, and a character conceptual network thereof; (2), to anchor HanziNet as a coupling interface between Concept and WordNet in Chinese context, in order to facilitate the integrated lexical and knowledge processing. Figure 2 schematizes further the representation of the middle layer. In Aitchison s (2003) terms, for the character level, we take an atomic globule network viewpoint, where the characters realized as concept primitives which share similar conceptual information, cluster together (as do the clouds in Figure). Inspired by Chu s Ontology design ( the relationships between these concept primitives form a tree with 2ⁿ (n=1, 2, 3, 5, 8) branches. Characters are thus assigned to the leaves of the tree in terms of an assemblage of binary bits. For the word level, we take the cobweb viewpoint, as words- built up from a pool of characters- are connected to each other through semantic relations. 1 Conclusion In conclusion, we have proposed a theoretical framework of Chinese lexical semantics, within which Chinese characters can be settled in a meaningful way. Some experiments and statistical analysis will be performed to validity this model. We believe that, with the completion of HanziNet, this model will shed light on the polysemy, morphological derivation and compounding as ways of expressing concepts in Chinese context. References Aitchison, Jean. (2003). Words in the mind: an introduction to the mental lexicon. Blackwell publishing. Handke, Jürgen. (1995). The structure of the lexicon: Human versus machine. Mouton de Gruyter. Harris, Roy. (2000). Rethinking writing. The Athlone Press, London. Hsieh, Shu-Kai. (2005). HanziNet: An enriched conceptual network of Chinese characters. In The 6 th Works hop of Chinese Lexical Semantics. China: Xia-Men. Lua, K. T. (1993). A study of Chinese word semantics. In Computer Processing of Chinese and Oriental Languages, Vol. 7, No 1. Murphy, Gregory. (2002). The big book of concepts. The MIT Press. 1 For detailed information please refer to Hsieh (2005) and the website : 80

94 WORDS IN ASIAN CULTURAL CONTEXTS English dictionary use strategies by EFL learners in Taiwan: a re-examination of the system of foreign language learning strategies Da-Fu Huang Southern Taiwan University of Technology dfjhuang@mail.stut.edu.tw Abstract In the context of learning English as a foreign language in Taiwan where the influence of the grammar-translation teaching method is deep-rooted in learners beliefs, English dictionaries are considered a necessary and important learning aid in translation and expansion of English vocabulary. Although English learners in Taiwan clearly rely on dictionaries during different learning stages, and dictionary use constitutes an integral part of learners reading and writing activities, our understanding is fairly limited as to how learners use dictionaries and why they use them as they do. This study thus investigated Taiwanese university English majors dictionary strategies used in different English learning tasks. A total of 414 Taiwanese university English majors completed a survey instrument developed for this study to assess the participants dictionary strategies: the English Dictionary Strategy Questionnaire (EDSQ). To triangulate the quantitative data and clarify the contexts of students dictionary use and their lookup ashington, interviews were conducted with six proficient and six less proficient English students. The quantitative results suggest that learners used memory, metacognitive, general cognitive, and task-bound cognitive strategies widely for translation aid and conscious learning of vocabulary. The results of this study suggest a positive perspective on dictionary use in foreign language learning and engender noteworthy pedagogical implications. Understanding students dictionary strategy use may also enable EFL teachers to help students develop effective learning strategies and ultimately improve their English abilities. Introduction Research on dictionaries would be incomplete without research on dictionaries from the user perspective. Despite the increasing studies on dictionary use and some evidence of positive association between dictionary use and language learning or teaching (Tono, 1989; McCreary and Dolezal, 1999; Knight, 1994; Summers, 1988, Hustijin et al, 1996, Laufer and Hill, 2000, Notohara, 1987, Kipfer, 1987, Huang, 2003, etc.), our understanding of dictionary use by EFL learners and the contexts of dictionary 81

95 ASIALEX 2005 use in an EFL setting is fairly limited. Unlike ESL learners, EFL learners learn the target language in a linguistic environment where English is not used for communicative needs in the daily social, economic, or political lives. Typically situated in what Kouraogo (1993) dubbed an input-poor English learning environment and exposed to ashin target language input, EFL learners consequently have fewer opportunities to use and communicate in the target language, which in turn leads to less availability of authentic language input or exposure to authentic language materials. For EFL learners, in addition, reading and print exposure tend to be the major source of input of English and the dominant mode of language comprehension and learning. In an EFL learning or teaching context where learners lack the opportunities to use English in the authentic situations and negotiate meanings or resolve their learning difficulties by or in the direct use of the target language, it should come as no surprise that such learners use their native language to assist in learning English and turn to other self-directed options such as dictionaries or translation machines to gain more access to the language input and solicit more aid in various English learning tasks. Accordingly, EFL learners such as those in Taiwan are reasonably expected to experience more dictionary use than, and have different dictionary beliefs and dictionary strategy use from, ESL learners. Prior studies (e.g. Béjoint, 1981; Tono,1988; Atkins and Varantola,1998, etc.) were also unsatisfactory in that they focused only on dictionary skills while neglecting dictionary strategies and the distinction between the two. Skills are cognitive routines or capacities that have been honed or developed through practice and can be expressed in ashingto at any time whereas strategies involve conscious decisions to select, combine, or redesign those routines or implement those skills (Schmeck,1988; Kirby, 1988). In the same vein, dictionary skills involve understanding dictionary conventions or routines, using dictionaries the way the dictionary editors expected them to be used in order to efficiently arrive at the target linguistic units. Dictionary strategies, in contrast, may involve choice or decision making as to dictionary use which may necessitate implementation of dictionary skills, and tend to be task-driven or tied to specific learning contexts. Since language learners use dictionaries in language learning contexts or for the purposes of performing specific language tasks, it is important to understand not only how learners use dictionary skills to target and retrieve linguistic information in dictionaries, but also how they process, interpret, analyze, or transform the sought information to facilitate language learning or tackle contextually-based learning tasks. During reading or writing English, learners are likely to use lookup methods such as cognitive, note-taking, cross-referencing, memory, or social strategies to help complete a learning task. Unexplored in prior studies, this wide spectrum of dictionary strategy use can be fitted well into, and should be examined within, an overarching and more comprehensive learning strategy classification scheme such as the strategy systems in Oxford (1990), O Malley et al (1985a) and O Malley & Chamot (1990). The purpose of this paper is therefore to understand Taiwanese EFL learners lookup strategies and how they can be incorporated into the existing foreign language learning strategy system. Methodology This study employed quantitative and qualitative approaches to understand Taiw anese university English majors dictionary strategies. The major research method used in the study involved the administration of the English Dictionary Strategy 82

96 WORDS IN ASIAN CULTURAL CONTEXTS Questionnaire (EDSQ) to assess students English dictionary strategy use. Around 400 English majors of Southern Taiwan University were requested to respond to the items on dictionary strategy use on a Lickert scale from (1) Always to (5) Never, thus demonstrating how frequently they used the strategy. The mean score reflected how frequently students report using dictionary strategies in English learning. In addition, qualitative interviews of some participants and the participants responses to the open-ended questions included in the strategy questionnaire further delineated the context of the participants dictionary and hence cross-checked the quantitative data. Semi-structured interviews of 12 of the participants were conducted to understand their dictionary use while reading and writing English. Results the Table 1.1 presents the ten most frequently used dictionary strategies reported by participants in the present study (M < 2.50). Table 1.1: The ten most frequently used dictionary strategies reported by the Participants Item Description M 17. If a paper or electronic English-Chinese dictionary gives several senses for an English word I 1.46 am l ooking up, I try to select one most suitable sense to understand the text I am reading. 19. When reading an English passage, I check out meanings of an English word from a dictionary 1.60 and copy the Chinese glosses of the word in the margins of the passage. 25. When reading English, although a word is familiar to me, I look it up in a dictionary to 1.93 retrieve more senses or usage of the word. 26. When writing English, I use a paper or electronic Chinese-English dictionary to find out the 1.94 English for a Chinese expression every time I don t know how to express it in English. 11. When looking up an English word in a dictionary, I try to read each sense of the word When reading an English passage, I underline unknown words, and look them up in a 2.34 dictionary only after I have read one or several paragraphs. 12. When looking up a word in a dictionary, I try to read all the example sentences of the word Whenever I come across an unknown word in reading, I check out its meanings right away in 2.46 a dictionary. 10. When I look up a word in a dictionary, I try to learn the word by heart at the same time When reading an English passage, I look up an unknown word and write in the margins of the 2.50 passage all or a part of its phonetic symbols to help pronounce the word. The ten most often used dictionary strategies pertained exclusively to general cognitive (items 10, 11, 12) and task-bound cognitive strategies (items 17,19, 21, 22, 24, 25, 26). The use of general cognitive strategies reflected students prevalent lookup ashington of reading as many example sentences (item 12) and word senses (item 11) as possible as well as memorizing the word at the same time when looking it up in a dictionary (item 10). When engaging in English reading and writing tasks, these lookup ashington typically would go hand in hand with the follow-up task-bound strategies, such as selecting one most suitable sense to understand the context of a passage being read (item 17), copying and writing the checked out Chinese glosses (item 19) or phonetic information (item 21) in the margins of the text. Besides, while reading or writing, students tended to check out its meanings immediately upon encountering an unknown word (item 24) or when not knowing how to express in English a Chinese thought (item 26). The quantitative results suggest that students use general cognitive and task-bound cognitive strategies most frequently particularly for the purposes of conscious learning of English vocabulary as well as obtaining translation aid in reading 83

97 ASIALEX 2005 and writing tasks. In addition, when it comes to dictionary use for conscious learning of vocabulary, high proficiency learners tended to focus on sentence-level word information (e.g. example sentences, grammatical information, and multiple word senses) encoded in a dictionary, while low proficiency students tended to focus on the word-level information ( e.g. phonetic information, spelling, and word meanings). More importantly, among the ten most frequently used dictionary strategies reported in this study, high proficiency learners were more likely to use what EFL teachers in Taiwan consider as the thoughtful strategies such as many of the general cognitive strategies, and less likely to use what they consider as the less effective strategies such as several of the task-bound cognitive strategies, than low proficiency learners. The qualitative data also showed that interviewees took notes in the margins of a reading text or kept additional notebooks or index cards to collect word information, marking in or on a dictionary, reading and analyzing different word senses and determining a best sense to fit in a reading context, referring to more than one dictionary to solve reading difficulties, checking out phonetic transcriptions or example sentences to help learning words by rote, and so on. Besides, both proficient and less proficient students would typically use Chinese to a differing extent to help them write English by using Chinese-English dictionaries for the English translations of a Chinese thought which they had no idea of how to put it in English. However, proficient learners were more inclined to proceed to English-Chinese or English-English dictionaries to clarify and distinguish the meanings of the English translations, which were usually unclear for most students. Following Oxford (1990) s strategy system, proficient interviewees tended to use metacognitive strategies (e.g. cross-reference), general cognitive strategies (e.g. checking out example sentences or phonetic transcriptions for word memorization), and memory strategies (e.g. keeping additional notes to refresh word memories) than less proficient ones. Conclusion The results of the study suggested that the students used a wide variety of lookup strategies during English reading and writing, and that the higher the students English proficiency, the more thoughtful and reflective they tend to be in using dictionaries. Since learners were found in this study to use a series of cognitive strategies, including note-taking, which remained unexplored in previous dictionary use studies, any previous understanding of dictionary use only as a cognitive, referencing strategy (e.g. Scholfield, 1997) needs to be modified to include memory strategies, social strategies, and metacognitive strategies as evidenced in cross-referencing, self-evaluating and monitoring learning progress. These findings of learners dictionary strategy use add a new dimension to our understanding of the second/foreign language learning strategies, since previous strategy research did not take learners dictionary use into consideration. The findings of this study also suggested that the effective learning strategies typically used by the more proficient language learners reported in prior strategy research on good language learners (e.g. Rubin, 1975; Stern, 1975; Naiman et al, 1978) appears to parallel the case of effective dictionary strategies, thus implying that good language learners are also good dictionary users, and vice versa. 84

98 WORDS IN ASIAN CULTURAL CONTEXTS References Atkins, B. T. S. & K. Valantora (1998), Language learners using dictionaries: the final report on the EURLEX/AILA research project on dictionary use, in B. T. S. Atkins (ed), Using dictionaries: studies of dictionary use by language learners and translators (Tübingen: Max Niemeyer), pp Béjoint, H. (1981), The foreign student s use of monolingual English dictionaries: a study of language needs and reference skills, Applied Linguistics 2(3): Huang, Da-Fu (2003), Taiwanese university English majors beliefs about English dictionaries and their dictionary strategy use, unpublished doctoral dissertation (Austin: University of Texas at Austin). Hulstijn, J. H., M. Hollander & T. Greidanus (1996), Incidental vocabulary learning by advanced foreign language students: the influence of marginal glosses, dictionary use, and reoccurrence of unknown words, The Modern Language Journal 80(3): Kipfer, B. A. (1987), Dictionaries and the intermediate student: communicative needs and the development of user reference skills, in A. P. Cowie (ed), The Dictionary and the Language Learner (London: Longman), pp Kirby, J. R. (1988), Style, strategy, and skill in reading, in R. R. Schmeck (ed), Learning strategies and learning styles (New York: Plenum Press), pp Knight, S. (1994), Dictionary use while reading: the effects on comprehension and vocabulary acquisition for students of different verbal abilities, The Modern Language Journal 78(3): Kouraogo, P. (1993), Language learning strategies in input-poor environments, System 21(2): Laufer, B. & M. Hill (2000), What lexical information do L2 learners select in a CALL dictionary and how does it affect word retention? Language Learning & Technology 3(2): McCreary, D. R. & F. T. Dolezal (1999), A study of dictionary use by ESL students in an American university, International Journal of Lexicography 12(2): Naiman, N., M. Frohlich, H. H. Stern & A. Todesco (1978), The good language learner (Toronto: Ontario Institute for Studies in Education). Notohara, A. (1987), The results of the Dictionary Using Skills Test (I), paper presented at the 13 th Zenkoku Eigo Kyoiku Gakkai in Okayama. O Malley, J. M., A. U. Chamot, G. Stewner-Manzanares, L. Kupper & R. P. Russo (1985), Learning strategies used by beginning and intermediate ESL students, Language Learning 35: O Malley, J. M. & A. U.Chamot (1990), Learning strategies in second language acquisition (Cambridge: Cambridge University Press). Oxford, R. L. (1990), Language learning strategies: What every teacher should know (New York: Newbury House). Rubin, J. (1975), What the good language learner can teach us, TESOL Quarterly 9: Schmeck, R. R. (1988), An introduction to strategies and styles of learning, in R. R. Schmeck (ed), Learning strategies and learning styles (New York: Plenum Press), pp Scholfield, P. (1997), Vocabulary reference works in foreign language learning, in N. Schmitt & M. McCarthy (eds), Vocabulary: description, acquisition and pedagogy (Cambridge: Cambridge University Press), pp Stern, H. H. (1975), What can we learn from the good language learner? Canadian Modern Language Review 31: Summers, D. (1988), The role of dictionaries in language learning, in M. Carter and M. McCarthy (ed), Vocabulary and Language Learning (London: Longman), pp Tono, Y. (1988), Assessment of the EFL learners dictionary using skills, JACET Bulletin 19: Tono, Y. (1989), Can a dictionary help one read better? in G. James (ed), Lexicographers and their works (University of Exeter), pp

99 ASIALEX 2005 Go and do, go to do, and go do Table 3 a corpus-based phraseological analysis Ai Inoue Kwansei Gakuin University, Japan narudo24@hotmail.com Abstract I will propose that the phraseological units go and do, go to do, go do should be differentiated semanti cally and syntactically. This will be proved by the fact that syntactically go and do takes an object of do and an adverbial of go at the same time. It follows that the original meaning of do seems to have been lost. By contrast, syntactically go to do and go do can take only one object of do. In terms of semantics, the original meaning of go seems to have been lost. These considerations would lead us to believe that go do should be regarded as the ellipsis of to from go to do instead of the ellipsis of and from go and do. This proposal is based on the theory of semantic syntax extensively discussed in Yagi (1996, 1999, and 2004). Introduction The purpose of this study is to discuss the semantic and syntactic differences between the phraseological units go and do, go to do, and go do based on the theory of the s emantic syntax, which is the assumption that part of the meaning of a word is closely related to the syntactic characteristics of the word (Yagi 2005:16). There are countless examples of word combinations like these in English where meanings of component words are simple and easy but they start to show their existence and ins ist on their roles as meaning carriers of subtle differences in the discourse. Details aside, let me call such units as phraseological units (hence Pus). I will analyze go and do, go to do, and go do as Pus using a spoken corpus. The result of the analysis reveals some interesting facts. Syntactically, go do should be regarded as the ellipsis of to from go to do instead of the ellipsis of and from go and do, which is evidenced by the similar distributional properties of go do and go to do. Semantic ally, the original locomotive meaning of go seems to have been lost from go do and go to do, while go of go and do seems to be retained and thus, there are two events of go and do. This is evident from the fact that syntactically go and do can take an object of do and an adverbial of go at the same time. By contrast, go to do and go do do not have the time interval between the events of go and do and focus on showing the purp ose of do, I will argue that go and do, go to do and go do should be separately listed in dictionaries with detailed explanations. The origin and current situation of Phraseology Sinclair (1991) proposes a theory that speakers formulate sentences not by using the ability to generate sentences infinitely but by using semi-preconstructed chunks, or Pus. At this point care must be taken not to confuse Sinclair (1991) with the origin of phraseology. Phraseology had already been recognized from three different standpoints. 86

100 WORDS IN ASIAN CULTURAL CONTEXTS The first is that the pedagogical standpoint of Palmer and Hornby. It was initiated in Tokyo from the 1920s. At that time, Pus were termed as comings-together-of words or formally and technically collocations (cf. Cowie 1999). The second is the phraseological studies carried out by Russian national policy in the 1930s by Vinogradov and Amosova. It was for the first time that the term phraseology in its present sense was used. Russian phraseology greatly influenced Makkai s works (1975, 1987). The t hird is the collocational study by the members of the Neo-Firthian school of linguists such as Mitchell, Halliday and Sinclair. They studied phras eology in the light o f grammatical category. Recently these three movements have come to converge due to the advancement of corpus. This will be evident from the fact that the BBI (1986, 1997), which succeeded the fruits of Palmer, Hornby, and Russian phraseologogists, was published and that research on phraseology using corpora is being actively carried out. The data the significance of using the Larry King Live Corpus The Larry King Live Corpus (for short LKL Corpus) is a spoken corpus and consists of interview programs of the Larry King Live and Larry King Weekend on CNN. The amount of words as of March in 2003 that I used in this study is about 13,000,000 words. Reasons for not using BNC and WordBanks for this research are as follows: one is that BNC and WordBanks on one hand and the LKL Corpus on the other have different aims. BNC and WordBanks are compiled aiming at general purpose studies such as dictionary-making, I therefore refer to them as general purpose corpus (hence GPC). On the other hand, this research focuses on Pus used in spoken English. Inevitably, I will utilize a corpus which is built for the studies of specific purpose corpus (fo r short SPC ). The other is the difference of quality in spoken sections between BNC, WordBanks, and the LKL Corpus. The contents of spoken sections in BNC and WordBanks are prepared speech. Thus, the spoken sections in BNC and WordBanks do not subsume the following features of spoken English denoted in Leech and Svartvik (2002:11): silent pauses, voice-filled pauses such as erm, repetitions, false starts, discourse markers and short forms. Additionally, authentic spoken data must be impromptu and spontaneous. The LKL Corpus includes all these features. Thus, I regard the interview as a worthy source for research on spoken English. A review of previous research on go and do, go to do, and go do A vast amount of scholarship has been devoted to the problem of go and do, go to do, and go do. Research on them concludes that go and do, go to do, and go do are almost synonym. For example, Declerck (1991) mentions that stop and think and stop to think are much the same semantically. Similarly, many dictionaries such as CALD, LDOCE 4, and COB 4 do not make a clear distinction between go and do, go to do, and go do. These dictionaries explain that go and do is used in colloquial style instead of using go to do. In addition, go do is considered to be an omission of either and or to. By contrast, LAAD regards go do as the ellipsis of to from go to do. MED describes go and do and go to do separately. Both dictionaries do not, however, explain how go and do and go to do are distinguished. Persuasive research has not been done on go and do, go to do, and go do. 87

101 ASIALEX 2005 Thus, this paper will focus on making a clear semantic and syntactic distinction between them, taking the function of and and to-infinitive into consideration. Go and do According to Huddleston and Pullum (2002), the sequence of X and Y implicates X and then Y, which means that there is a time interval between the events of X and Y. To put this another way, and in go and do operates as a sequent in Qui rk et al. (1985) s terminology. Consequently, go and do syntactically takes an object of do and an adverbial of go at the same time as is shown in (1). This means that the original meaning of go is retained and there are two events in go and do. Go to do (1) KING: Who came up with idea of, good-bye? ROBINSON: That s mine. We used to have to go and see her in big bedroom. And when she d had enough of us, she would say, Thank you. Good-bye. [ LKL, May, 2001] It is fairly generally agreed that to-infinitive is used to clarify the meaning of purpose (cf. Jespersen 1940). In such a case, there is no time interval between go and to-infinitive. Thus, considering the syntactic fact that an object of to-infinitive appear in go to do in the following example, the original meaning of go seems to have been lost. Go do (2) KING: Well, how do you break this down? What s your choice mechanism? ARNOLD: Well KING: I will go to see her, I won t go to see her. ARNOLD: Well, my dad is helping me. He likes women with jobs. That s his whole he doesn t care what they look like if they ve got a job. Our friend Steve Tisch is helping me. He s a [ LKL, Mar. 2000] (3) KING: Cause it is a rom it is certainly a light romantic, funny movie that s also pretty rough CLOONEY: It is. KING: It has some hard edges. CLOONEY: And it s funny. I m not one - - I m not all that apt to go see a romantic comedy. I m fairly tainted. And I but this isn t necessarily a romance. KING: You can t type it. [ LKL, Jun. 1998] As represented in (3), go do takes only one object of do as well as go to do, so the events of go and do are carried out at the same time and go do also shows the purpose of do. Semantically, the lexical meaning of go seems not to be retained. These considerations would lead us to believe that go do seems to be the ellipsis of to from go to do instead of the ellipsis of and from go and do. 88

102 WORDS IN ASIAN CULTURAL CONTEXTS The differences between go and do, go to do, and go do Table 1 gives a summary of the differences between go and do, go to do, and go do syntactically and semantically. Table 1 The differences between go and do, go to do, and go do syntactic feature semantic feature go and do the object of do and the a time interval between go and do adverbial of go go to do the object of do no time interval between go and do, the original meaning of go seems to have been lost. Go do the object of do no time interval between go and do, the original meaning of go seems to have been lost. Analysis of the result of informant elicitation On the basis of Table 1, I inquired into which of these sentences in each example is often used of ten informants (five Americans and five British). (4) a. I ll go and pick up my friends at the airport this afternoon. b. I ll go to pick up my friends at the airport this afternoon. c. I ll go pick up my friends at the airport this afternoon. (5) a. I ll go and pick up my friends this afternoon. b. I ll go to pick up my friends this afternoon. c. I ll go pick up my friends this afternoon. Table 2 An answer to (4) Table 3 An answer to (5) informants U.K. U.S. informants U.K. U.S. (4a) 3 2 (5a) 0 0 (4b) 2 2 (5b) 5 4 (4c) 0 1 (5c) 0 1 total 5 5 total 5 5 The tables 2 and 3 indirectly support the discussion above and the result as summarized as Table 1. Conclusion Summarizing the above discussion, Pus go and do, go to do, go do are differentiated syntactically and semantically. Therefore, I venture to throw out the suggestion that go and do, go to do, go do should be separately listed in dictionaries. References Cowie, Anthony, P English dictionaries for foreign learners: a history. OUP. Declerck, Rennat A comprehensive descriptive grammar of English. Tokyo: Kaitakusha. Huddleston Rodney and G. K. Pullum The Cambridge grammar of the English language. Cambridge: Cambridge Univ. Press. Jespersen, Otto A modern English grammar on historical principles part V. Copenhagen: Eijnar 89

103 ASIALEX 2005 Munksgaard. Leech, Geoffrey & J. Svartvik A communicative grammar of English. Third edition. Longman: Pearson Education. Quirk, Randolph, S. Greembaum, G. Leech & J. Svartvik A comprehensive grammar of the English language. London: Longman. Sinclair, John Corpus, concordance, collocation. OUP. Yagi, Katsumasa Neitibu-no chokkan-ni semaru goho kenkyu (A descriptive approach to modern English). Tokyo: Kenkyusha. Yagi, Katsumasa Eigo-no bunpo to goho (A semantic-descriptive approach to modern English). Tokyo: Kenkyusha. Yagi, Katsumasa Imiteki tougoron kara mita want no hobunkouzou (The complementation of want from the viewpoint of semantic syntax). H. Goto. (ed.). Eigo kenkyu no syosou Gengo, kyouiku, bungaku (Some aspects of English studies language, pedagogy, literature). Sapporo: Kyoudoubunkasha, Yagi, Katusmasa <±HUMAN> and which/who alteration from the semantic syntactic viewpoint. Language and culture. No.8. Kwansei Gakuin University, Language Educational Center, Dictionaries BBI 1,2 : The BBI Combinatory Dictionary of English , The BBI Combinatory of English Words Combinations Amsterdam and Philadelphia: John Benhamins. CALD : Cambridge Advanced Learner s Dictionary COB 4 : Collins COBUILD Advanced Learner s English Dictionary LAAD : Longman Advanced American Dictionary LDOCE 4 : Longman Dictionary of Contemporary English. 4 th edition MED : Macmillan English Dictionary Lexical features of the English composition by Japanese EFL learners: a quantitative analysis of the online discussion messages posted by Japanese and Canadian college students Abstract Shin ichiro Ishikawa Kobe University iskwshin@vanilla.fre .ne.jp The development of and online BBS system has created more and more opportunities for EFL learners to communicate with native speakers in English, but the composition by non-native speakers seems to be somewhat deviant from the one by native speakers. There have been several studies on the analysis of English composition by Japanese college students. The researchers have discussed the vocabulary size (Nonaka, 2004), paragraph organization (Sakagami, 2004), collocation (Hasegawa & Chujo, 2003), and topic (Inoue, 2003). In Ishikawa (in press), the author conducted a quantitative analysis of the English essays written by Japanese college students, which illuminated several interesting facts in their use of vocabulary. However, in order to discuss the lexical deviance typically seen in the writing by Japanese learners, we need to have a comparative study. Therefore, in this paper, we will compare the online discussion messages posted by Japanese college students with those by their Canadian 90

104 WORDS IN ASIAN CULTURAL CONTEXTS partners (approx. 24,000 tokens each). Our research focus is mainly on the quantitative indexes, such as the number of tokens, type/token ratio, word length, word level (in comparison with the defining vocabulary of EFL dictionaries), and so on. Also, we will investigate the statistical keywords and the word class distribution in each data, which will show us a more detailed lexical profile of English composition by Japanese EFL learners. Usage of cultural keywords seen in English dictionaries and corpora Abstract Yuka Ishikawa Nagoya Institute of Technology, Japan y_uka@catnip.fre .ne.jp Are there any words which have peculiarly significant meanings to one culture and society, but not so important to others? We can find a number of dictionaries of Western culture and society in bookstores, and in recent years, some dictionaries of Asian cultures have also been published. Many of them contain just some dozens, or, at most, around a hundred headwords and associated phrases that seem to represent the culture (e.g. William s, 1983). Western intellectual culture is said to be linked to a few dozen words (Benvenist e, 1954 ). The lex icon of a language is surely a key to the culture, but such keyword s, th rough which a culture can be understood, are not necessarily so many. Several researche rs have argued that studying the usages of some particular words will identify or reve al some aspects of the complex and elusive concept of a culture (Stubbs, 1996/2002; Wier zbicka, 1999). In this paper, we will firstly focus on the historical change of the meaning of such keywords by studying some definitions and examples cited in dictionaries. Words acquire new connotations in a society. We will therefore compare their uses in English dictionaries published some decades ago and those in the newest one s. Secondly we will investigate their usage by using various kinds of contemporary English corpora that may reveal hidden attitudes or moral views that are shared by the people in the same culture. Brythonic loanwords in English and the receptive knowledge maintained by native speakers of English Mitsuhiko Ito Toyohashi University of Technology mitsu@hse.tut.ac.jp Abstract The research objectives are: (1) to see how many Brythonic words native speakers of English maintain, (2) to see if English dictionaries and corpuses keep Brythonic words which are us ed by native speakers of English, and (3) to see if age difference and gender 91

105 ASIALEX 2005 difference appe ar among native speakers of English. The research er used the SOD ( 1973), and he collected 48 Brythonic words from the SOD according to etymological descriptions for each entry. Then, he made a matching type of questionnaire comprised of the 48 words, and delivered it to native speakers of English in Japan, males and females, most of whom are college or university English teachers. The researcher analyzed the results of the 25 replies from male subjects and the 25 replies from female subjects to examine the three main objectives outlined above. Introduction It is obvious that native speakers of English do not know all of the entries in dictionaries. Furthermore, they do not know all of the Celtic words in English even though they know that the English language has adopted Celtic words throughout the history of the language. Native speakers of English might be aware of some of those Celtic words found in dictionaries. Ito (1993, 1997, 2000a, 2000b, 2001, 2003, 2004) has studied native speakers knowledge of Irish words, French words of Celtic origin, and Scottish Gaelic words by using words from the Shorter Oxford Dictionary (1973). Methods for the Present Study The main purposes of the present study are (1) to see how familiar native speakers of English are with Brythonic words, (2) to see if native speakers of English maintain more receptive knowledge of Brythonic words than Irish words, (3) to see if native speakers of English keep more receptive knowledge of Brythonic words than Scottish Gaelic words, and (4) to see if native speakers of English retain more receptive knowledge of Brythonic words than French words of Celtic origin in English. It should be noted that an etymological discussion will not be conducted in this study. The questionnaire, composed of 48 words from the SOD, required the participants to match each word with its appropriate definition given in the questionnaire. The questionnaires were sent out to native speakers of English in Japan until twenty-five male replies and twenty five female replies were returned to the researcher. The 48 words for the present study are comprised of 29 Welsh words (bragget, brut, cader, cambrel, cockabondy, commot, coracle, corgi, cromlech, crowd, cwm, eisteddfod, flannel, flummery, gossan, grig, grouse, guillem, gull, gwiniad, kibe, kistvaen, lech, metheglin, morglay, pendragon, pennill, pikelet, tope, and towan), 12 Cornish words (bowssen, costean, dillue, elvan, grouan, killas, morgay, puffin, vug, and wrasse), 6 Brythonic words (bannock, bin, brock, crag, pen, and tor), and 2 Breton words (menhir, and peulvan). Results The results and discussion are based on the twenty-five male responses and the twenty-five female responses to the questionnaire which were returned to the researcher. The participants were all native speakers of English. The age of the participants ranged from their 30s to their 70s. All of the 50 participants held various teaching positions in colleges or universities. Furthermo re, a ll of the participants had higher education. The researcher checked to see if the 48 words appeared in corpuses based on materials used in daily life as well as other dictionaries which were compiled for language learners and used in daily life by native speakers of English. These included: 92

106 WORDS IN ASIAN CULTURAL CONTEXTS Frequency Analysis of English Vocabulary and Grammar Based on the LOB Corpus vol. 1. (1988), the Collins COBUILD English Language Dictionary (1978), the Longman Dictionary of Contemporary English for Advanced Learners, third edition (2001), and the Longman Dictionary of American English (1983). Thirteen words came from the LDOCE, nine words appear in the COBUILD, seven words were found in the ADLOCE, and five words are contained in the LOB Corpus. However, 23 of the words are included in the American Heritage Dictionary of the English Language, fourth edition (2000) [hereafter the Heritage CD]. It is noted that in the Random House Unabridged Dictionary, Second Edition CD-ROM, which the researcher considers the American counterpart of the SOD, 25 words were adopted. The number of answers and correct answers varied widely. The number of answers by male participants ranged from a high of 48 to a low of 3 (14.68 average), whereas the number of answers by female participants range from a high of 48 to a low of 3 (19.2 average). The highest number of correct answers by male subjects was 17, and the lowest was 1 (7.76 average), whereas the highest number of correct answers by female participants was 18 and the lowest was 3 (8.28 average). The number of words with at least one correct answer ashingt 36. Correct answers, their frequencies and the percentage of participants who had them are noted below: List 1 36 correct answers of the 48 Brythonic words word male female total frequency corgi flannel grouse crag puffin bin coracle gull eisteddfod bannock pendragon tor brock cwm pikelet metheglin cromlech flummery pen dillue kibe kistvaen menhir brut cockabondy lech tope vug wras se elvan gossan

107 ASIALEX 2005 grig grouan killas morglay pennill It is clear that out of the 48 words, some were answered by both male and female participants, some others were answered by only one gender, and the others were not answered at all. Discussion All of the Brythonic words which appear in the COBUILD, the LDOCE, the ALDOCE, and the LOB appear in the identified 36 words. Twenty-one of 23 words, except for crowd, in the Heritage CD-ROM appear in the correctly answered 36 words, and 22 of 25 words, except for costean, and crowd, in the Random House CD-ROM. As in the case o f Scottish Gaelic, however, even the Webster s New Third International Dictionary, Unabridged CD-ROM (2000) does not include 11 words from the 48 words: cader, cockabondy, elvan, grouan, guillem, gwiniad, morgay, morglay, pen, pennill, and peulvan. Thus, 37 words of the 48 words appear in the Webster. It is clear that 30 of the identified 36 words appear in the Webster and that six of the 11 words which do not appear in the Webster were recognized by the participants (cockabondy, elvan, grouan, morglay, pen, and pennill). Many of the 12 unidentified words seem to be unfamiliar to native speakers of English. None of the unidentified 12 words are contained in the LOB, the LDOCE, the ALDOCE, and the COBUILD. Eleven of the 12 words are not included in the Heritage CD-ROM; one word which appears in the dictionary is crowd. Ten of the 12 words are not included in the Webster; the two words which do appear in the dictionary are costean and crowd. Thus, it might be natural that native speakers of English can not match the appropriate definitions for the 12 words. Statistical analysis of the results of male and female differences in correct answers is as follows: gender number of subjects average standard deviation male female: The t score for gender difference is with df = 48, which means that there is no significant difference between the two groups. In the similar study of Irish loanwords and that of Scottish Gaelic loanwords, there were no gender differences (Ito, 2000b, 2003). Statistical analysis of the results of age differences in correct answers is as follows: age number of participants average standard deviation 30s: s + over: There were no participants in their 20s in the present study; therefore, the participants were divided into two groups: 30s versus 40s and over. The t score 94

108 WORDS IN ASIAN CULTURAL CONTEXTS for age difference is with df = 48, which means that there is no significant difference between the two groups. In the similar study of Irish loanwords, there w as no significant difference (Ito, 2000b) ; however, in the similar study of Scottish Gaelic, there was a significant difference (Ito, 2003). Mention should be made of the matter of familiarity the participants had of Brythonic words by comparing the results of the present study with the results in previous studies about Celtic words in English by way of French (Ito, 2000a), Irish words in English (Ito, 2000b) and Scottish Gaelic words in English (Ito, 2003) even though it has been discussed to some extent in earlier sections. Native speakers of English correctly identified 66 (98%) out of 68 words, and more than half of them could identify 25 (37%) out of the 68 words when they were given Celtic words in English by way of French. They correctly identified 43 (53%) out of 81 words and more than half of them could identify 5 (6%) out of the 81 words when Irish words in English were considered. Finally, they correctly identified 46 (69%) out of 67 words and more than half of them could identify 8 (12%) out of the 67 words when they were given Scottish Gaelic words. In the present study, words correctly identified by at least one participant were 36 (75%) out of 48 words from Welsh, Cornish, Breton and Brythonic, and words identified by more than half of them were 7 (15%) out of the 48 words. Accordingly, it seems that native speakers of English are more familiar with French words of Celtic origin than Brythonic words, and that they are more familiar with Brythonic words than the other two source groups. Conclusion A few findings are drawn from the results and discussion of the present study. First, native speakers of English are more familiar with Welsh, Cornish, Breton, and Brythonic words in English than Irish words and Scottish Gaelic words. This is probably because England has kept a geographically and politically stronger tie with Wales than Scotland. Second, native speakers of English are less familiar with Welsh, Cornish, Breton, and Brythonic words than French words of Celtic origin because French words of Celtic origin have been more naturalized than Brythonic words. Third, the present study confirms the appropriateness of word selection from the dictionaries or the corpus used in the study: the COBUILD, the LDOCE, the ALDOCE, the Heritage CD-ROM, and the LOB. Finally, statistics show variety for gender difference and age difference. The analysis reveals neither gender difference nor age difference as far as Brythonic words in English are concerned. References Ito, M. (1993). Irish loanwords familiar and unfamiliar to the native speakers of English, Bulletin of Toyohashi University of Technology, 15, Ito, M. (1997). Irish loan-words well-known by native speakers of English, Celtic Forum, 2, Ito, M. (2000a). Native speakers receptive knowledge of French words of Celtic origin in English, Bulletin of Toyohashi University of Technology, 22, Ito, M. (2000b). Receptive and productive knowledge of Irish loanwords maintained by native speakers of English, Éire, 20, Ito, M. (2001). Native speakers productive knowledge of French words of Celtic origin in English, Bulletin of Toyohashi University of Technology, 23,

109 ASIALEX 2005 Ito, M. (2003). Native speakers receptive knowledge of Scottish Gaelic in English, Éire, 23, Ito, M. (2004). Native speakers productive knowledge of Scottish Gaelic in English, Bulletin of Language Center of Okayama Prefecture University, 2, Johansson, S. & Hofland, J. (1988). Frequency analysis of English vocabulary and grammar based on the LOB corpus. Vol. 1 (Oxford: the Clarendon Press). Dictionaries The American Heritage dictionary of the English language, fourth edition CD-ROM. (2000). Boston: Houghton Mifflin. Collins COBUILD English language dictionary for advanced learners. (1978). London: HarperCollins. Longman dictionary of American English. (1983). NY: Longman. Longman dictionary of contemporary English for advanced learners, third edition. (2001). Essex: Longman. Random House unabridged dictionary, second edition. CD-ROM (2001). NY: Random House. The Shorter Oxford English dictionary (1975). Oxford: The Clarendon Press. Webster s third new international dictionary unabridged v2.5 CD-ROM (2000). MA: Merriam-Webster. Designing a pedagogical dictionary of English academic vocabulary for Chinese speakers Abstract Gregory James Hong Kong University of Science and Technology This paper will introduce a project to determine a basic academic vocabulary for first-year undergraduates in Hong Kong entering an English-medium university after a Chinese-medium secondary education. It will describe the compilation of a 5,000,000-word corpus of extracts from texts in Science, Engineering, Business, Humanities and Social Science, the selection of headwords, and the provision of definitions and exercises. Once the corpus was collected, the frequency list was lemmatized and main and derived items from the Longman Defining Vocabulary were deleted (assuming the students would not need to look these up, after 15 years of English!). The remaining words constituted the basis for the dictionary headwords with subjective decisions made by the editors in a few cases. The list was expanded to include significant word combinations found to occur frequently in the corpus. A user-friendly relational database was designed to help users navigate the various fields. The headwords are enhanced by indications of frequency, discipline-specific definitions and contextualized exemplification, based on the corpus evidence, as well as explanations of extended uses. The headwords are matched to uses in a 20,000,000-word corpus of Hong Kong students writing in English, and notes prepared, where difficulties are apparent, for incorporation into the Help sections. An on-line vocabulary training tool is being designed to supplement the glossary, to help students use and recognize the headword vocabulary. 96

110 WORDS IN ASIAN CULTURAL CONTEXTS The Changes in Second Language Attrition Lexicon and Syntax Tse-Chun Jao National Taipei Teachers College, Taiwan Shu-Chen Chiu National Taipei Teachers College, Taiwan Abstract This case study examines the changes in the subject s lexicon and syntax that occurred during the process of natural second language (English) attrition in the first language (Mandarin) environment. The subject spent two years in the United States and returned to Taiwan at 8:10. In terviews and storytelling were used to track language attrition from the sixth month of the subject s return to the eighteenth month. The results show that there was language attrition including (1) a 31.7% attrition or change in use of vocabulary, (2) errors in the usage of prepositions, (3) changes in the expression of the possessive case, (4) modification of nouns, from post-nominal to pre-nominal, and (5) increasing errors in the usage of tense. There is evidence to show that language attrition may be due to language interference as well as reduced exposure to English. Introduction Language attrition usually accompanies migration and the differences between language contexts can make it worse. Since 1980, the issue of language attrition has received attention from many researchers (Edwards, 1976; Bahrick, 1984; Welten et al., 1986; T omiyama, 1999; Tomiyama, 2000). Their efforts have provided useful answers to parts of the issue and given helpful advice in the teaching and learning of second language. However, the attrition process between different language contexts may be varied and little attention has been given to the impact on native Mandarin speakers. This study aims to examine English attrition in the Mandarin context, providing some baseline d ata to countries such as Taiwan, Japan and Korea for their EFL policy and education. Research question Th is study is intended to provide a qualitative view of the subject s L2 lexical and syntax attrition. The following is the question: How is his lexical and syntax affected over time? Review of language attrition research This section aims to introduce the possible factors and evidence on language attrition, including language interference, lexical and syntax changes. In the first place, the differences between languages may cause errors or attrition. This phenomenon is called language inte rference. Galbraith (1981) found in his study that around 5/6 th of errors could be explained by this attrition factor. Tomiyama (2000) also discovered evidence of how English is affected by Japanese. The modification of nouns will change, from post-nominal (e.g. a mom who is a tiger) to pre-nominal (e.g. a tiger mom). In the second place, the issue of lexicon, Andersen (1982) thought that the difficulties confronted usually were concerned with the loss of lexicon ability. The research in 97

111 ASIALEX 2005 Tomiyama s study shows a 16% lexicon loss after her subject returned for 17 months. The attrition in lexicon also is reflected in long pauses, codeswitching and the use of compensat ory strategies (Tomiyama, 1999). Furthermore, with respect to syntax, Cohen (1974) found that the subjects errors increased and length of spoken sentences shortened after a summer break. Tomiyama (2000) also gave evidence on syntax attrition, such as the omission of relativization, modification of nouns, misuse in prepositions, as well as the use of past tense and plural morphemes. Finally, from the point of view of memory, the difficulty, or the blocking of memory, that the attriter shows in communication can be characterized as being temporarily inaccessible or forgotten permanently. As Craik and Lockhart (1972) pointed out, the depth of processing might impact what remains in memory. Superficial elements (e.g. FL words learned only by repetition but not applied in context even once) are much more prone to loss than those deeper processed elements (e.g. FL words or syntax that have been applied). Method Subject The subject, Kevin, was born in Taiwan, and is a male Chinese child. He comes from an upper-class family. His parents are college-educated and proficient in English, and he has a younger sister. Kevin left Taiwan for New York at the age of 6;10. He returned to Taiwan after two years, at the age of 8;10. In Taiwan, and in America, Kevin was always enrolled in public elementary schools without extra English courses after school. His English competence was acquired but not learned. According to his mother, Kevin s silent period was around six months. After that, Kevin was able and willing to speak English at school. By the time he returned to Taiwan, his dominant language was English with a native speaker s accent. After returning to Taiwan the school that he attended had English classes for one hour per week, the only formal English course he took. However, Kevin read English books and watched English cartoons very often at home. When watching the programs, he preferred to listen to the dialogue rather than read the captions. This way it was easier for him to enjoy the cartoons. Data collection This study started from the sixth month after Kevin returned to Taiwan. The data was gathered every other month by the researcher visiting the subject s home, for a period of one year. Each session was audio-taped. Two picture books were employed as instruments to elicit the data. First, Kevin had to describe all the strange parts of Wacky Wednesday (LeSieg, 1974). Second, he would tell the story of A boy, A dog, A frog and A friend (Mayer & Mayer, 1971) in English. Every effort, for example, warm-up games, was employed to keep these interviews as natural as possible. Data analysis The audio-taped materials were transcribed for analysis. The analysis procedures were as follows. The changes in lexicon were observed by tracing 57 items in Wacky Wednesday. The changes in syntax were measured in both picture books by analysing prepositions, possessive cases, modification of nouns, and the errors in the usage of tenses. Modifications of nouns were divided into two types: post-nominal (e.g. a mom who is a tiger) and pre-nominal (e.g. a tiger mom). The wrong usage of tense fell 98

112 WORDS IN ASIAN CULTURAL CONTEXTS into two categories: present tense replaced by past tense (e.g. John said, say goodness, that turtle was (is) gone.) and past tense replaced by present tense (e.g. She don t (didn t) know that turtle will bite her again). Findings and discussion Lexicon A mong 57 items observed in Wacky Wednesday, 18 items (31.7%) changed. The result is higher than the study of Tomiyama (1999) (16%). The following shows the items that changed: socks, ceiling, sink, broken, pot, hook, disappeared, wheel, lose, skate, bea rd, desk, upside down, line, crocodile, none, down stairs, backwards, and bush. For example, the description below details the subject s change in use of the item wheel. Table 4 Session 1 ( 6 mos) Session 2 ( 8 mos) Session 3 (10 mos) Session 4 (12 mos) Session 5 (14 mos) Session 6 (16 mos) Session 7 (18 mos) wheel (Kevin described a baby carriage with only one wheel) one of the baby chair miss one wheel one baby chair had one wheels someone s baby chair has losing tires another baby chair has no two (pause) none of tires one baby car here was missing two wheels two tires was missing on the baby chair someone s baby chair has missing two tires The parts in italics show the change of the item from wheel to tire over time. It is important to know that the pronunciation of wheel in Mandarin, LÚN TAI, is much more similar to tire [tair] than to wheel [hwil]. Language interference might be the reason why Kevin replaced wheel with tire gradually. The item socks is another example. Table 5 socks (describing a boy taking a shower with socks on feet) Session 1 ( 6 mos) the socks is on my feet Session 2 ( 8 mos) there was a sock on his feet Session 3 (10 mos) not elicited Session 4 (12 mos) socks on my feet Session 5 (14 mos) and a wocks (pause) and a socks (pause) is (pause) on my leg Session 6 (16 mos) my feet has a wocks (pause) socks Session 7 (18 mos) a socks is in # on my feet In session 5 and 6, Kevin employed wocks first. After self-checking, he corrected the word wocks to socks. Because the pronunciation of socks in Mandarin, WÀ ZǏH, is similar to wocks [waks], it might provide evidence that language interference occured (Galbraith,1981). This example also reveals the increasing misuse of the article a. The next change is with the word beard. Table 6 beard (describing a student with a beard in an elementary school classroom) Session 1 ( 6 mos) one of the baby has beard. Session 2 ( 8 mos) not elicited Session 3 (10 mos) a mother was taking a walk with a grandpa Session 4 (12 mos) a mother is taking a walk with a old baby 99

113 ASIALEX 2005 Session 5 (14 mos) Session 6 (16 mos) Session 7 (18 mos) and the dwarf in (pause) the baby chair there s a old dwarf in that baby chair someone s baby had whisker The parts in italics show that Kevin employed compensatory strategies, using a circumlocution or synonym, e.g. grandpa, old baby, dwarf and whisker, to compensate for the loss of the item beard. Evidence of using strategies to compensate for attrition in the lexicon was also found in the studies of Andersen (1982) and Tomiyama (1999). The following is an example using the item backwards. (4) backwards (describing the statue of George Washington sitting backwards on a horse) Session 1 ( 6 mos) the george ashington seat were wrong. Session 2 ( 8 mos) george ashington sitting on the horse was wrong Session 3 (10 mos) the george ashington was sitting backwards Session 4 (12 mos) the george ashington status was upside down Session 5 (14 mos) the george ashington rock was upside down Session 6 (16 mos) the george ashington sitting upside down on his horse Session 7 (18 mos) the george ashington statue are upside down In session 2 and 3, Kevin simply used the item wrong to describe this wacky scene, providing evidence of employing simplification compensatory strategies. In session 3, the item backwards showed up precisely. However, backwards was replaced by the item upside down from sessions 4 to 7. The likely explanation is that he was using a circumlocution or synonym compensatory strategy. Other items also showed changes happening: ceiling was replaced by wall, sink was rep laced by tub, broken was replaced by hole, pot was replaced by soup, hook was replaced by holder, in the air, disappeared and lose were replaced by missing, desk was replaced by table, line became wire, crocodile became alligator, down stairs became place, route or something, bush became tree and finally there was the mixed use of none and no. It is clear that attrition occurred in Kevin s English lexicon. Syntax We consider the subject under the following four points: prepositions, the possessive case, modification of nouns, and errors in the usage of tense. First, with regard to prepositions, there were errors in Kevin s usage (especially the misuse of in and on). The following example shows this misusage in detail (describing a man sitting in the back seat driving a car, the parts in italics showing the preposition that Kevin used, the parts in parentheses showing the correct usage). Session 1 ( 6 mos) the driver drove in the back place Session 2 ( 8 mos) the driver was out on (in) the back Session 3 (10 mos) the driver of the car is on (in) the back Session 4 (12 mos) the driver was on (in) the back of the car. Session 5 (14 mos) the driver was back on (in) a car Session 6 (16 mos) the driver driving the car in the back Session 7 (18 mos) the driver is on the back seat Language interference may give a possible account of his misuse because the preposition cases in and on, and even at, are all pronounced ZÀI in Mandarin. It may explain why Kevin couldn t distinguish from in and on from time to time. 100

114 WORDS IN ASIAN CULTURAL CONTEXTS Second, there seemed to be three steps by which Kevin s usage of the possessive case changed. To begin with, he employed the item of to express the possessive case (e.g. session 1 and 4 below, the parts in italics). Next, he used the possessive morepheme s to express the case (e.g. session 2, 3 and 6). Finally, Kevin employed multiple noun modification to replace the possessive case (e.g. session 5 and 7). Session 1 ( 6 mos) Session 2 ( 8 mos) Session 3 (10 mos) Session 4 (12 mos) Session 5 (14 mos) Session 6 (16 mos) Session 7 (18 mos) one of the car (pause) wheels are wrong george ashington s picture was wrong again the flag s color were wrong on the back of the car (pause) someone is walking the george ashington picture was wrong that car s back the people walks the flag sign are upside down Third, part of Kevin s gradual change in the expression of the possessive case points to the third change in syntax of this study, noun modification, from post-nominal to pre-nominal. Table 1 shows the rate of change. Table 1:Rate of pre-noun modification Session # 1 # 2 # 3 # 4 # 5 # 6 # 7 6 mos 8 mos 10 mos 12 mos 14 mos 16 mos 18 mos Pre-nominal rate 44.0% 68.3% 76.3% 75.0% 69.2% 83.3% 91.1% (37/84) (28/41) (29/38) (27/36) (62/183) (35/42) (41/45) (In session 1, the sum of pre- and post-noun modification is 84, including 37 pre-noun modification. The rate of pre-nominal is 44.0% (37/84).) As table 1 indicates, the pre-nominal rate increased from 44.0% (session 1) to 91.1% (session 7). Research in a Japanese context found the same result (Tomiyama, 1999). The following example seeks to capture the change (describing a picture upside down, hanging on a wall, the parts in italics showing the method of noun modification Kevin employed). Session 1 ( 6 mos) post one of my pictures is upside down Session 2 ( 8 mos) not elicited one picture was falling down Session 3 (10 mos) pre my dad s picture was upside down Session 4 (12 mos) pre my uncle s picture was upside down Session 5 (14 mos) post a upside down picture of my uncle Session 6 (16 mos) pre my uncle s picture was upside down Session 7 (18 mos) pre someone s picture is upside down As for the possible reason as to why Kevin employed pre-noun modification gradually, language interference may again play an important role. In English, post-noun modification is more common than pre-noun modification, however, quite the opposite is the case in Mandarin. Finally, Kevin s misusage of tense changed over time. Table 2 summarizes the change in the rate of tense misuse. 101

115 ASIALEX 2005 Table 2 : Rate of tense misuse Session # 1 # 2 # 3 # 4 # 5 # 6 # 7 6 mos 8 mos 10 mos 12 mos 14 mos 16 mos 18 mos Rate of tense misuse 20.7% 4.8% 34.7% 39.0% 33.9% 38.2% 75.2% (31/150) (7/147) (58/167)(54/136)(62/183)(52/136)(115/ 153) (In session 1, the sum of tense use is 150, incluing 31 misuse. The rate of tense misuse is 20.7%(31/150)) Table 2 shows the increasing errors in the usage of tense, from 20.7% (session 1) to 75.2% (session7). Below is an example of the misuse (the parts in italics showing the Kevin s misuse, the parts in parentheses showing the correct usage). Session 1 ( 6 mos) He said I m so sorry I pull (ed) you out. Session 2 ( 8 mos) All right, I got (get) rid of the turtle. Session 3 (10 mos) One day they are (were) fishing. Session 4 (12 mos) They think (thought) to shake off the turtle. Session 5 (14 mos) Then they go (went) home. Session 6 (16 mos) One day when they go to (went) fishing, Session 7 (18 mos) but the turtle do (did) not move As seen from the example above, most of Kevin s misuse in tense is past tense replaced by present tense (except session 2). It may indicate that Kevin intended to employ the original verb to replace its past tense form. It is also possible that there was attrition in Kevin s past tense verb morphemes. Table 3 presents the increasing rate of his past tense verb morpheme errors. It shows a similar result to Tomiyama s study (Tomiyama, 2000). In Mandarin, tenses are not emphasized as much as in English. Language interference might be showing its influence on Kevin again. Table 3:Rate of past tense verb morpheme errors Session # 1 # 2 # 3 # 4 # 5 # 6 # 7 Rate of errors 6 mos 26.4% (14/53) 8 mos 22.7% (15/36) 10 mos 19.5% (8/41) 12 mos 34.3% (12/35) 14 mos 26.2% (11/42) 16 mos 43.6% (17/39) 18 mos 51.7% (15/29) (In session 1, the sum of past tense morphemes used is 53, including 14 errors. The rate of past tense morpheme errors is 26.4%(14/53).) Conclusion and suggestion In conclusion, this case study details considerable evidence of the subject s Englis h language attrition in a Mandarin environment. The subject experienced (1) a 31.7% attrition or change in use of vocabulary, (2) errors in the usage of prepositions, (3) changes in the expression of the possessive case, (4) modification of nouns, from post-nominal to pre-nominal, and (5) increasing errors in the usage of tense. From this case study it is apparent that language attrition may be due to language interference as well as reduced exposure to English. 102

116 WORDS IN ASIAN CULTURAL CONTEXTS The results of this case study provide information on Kevin s language attrition process. However, the results may not necessarily be similar for other children with similar backgrounds. In addition, the impact of memory caused by employing the same instruments over time may have affected the results. This study started from the sixth months after Kevin s return to Taiwan. Future research may want to start from the beginning of subject s return to provide even more precise results to those who are concerned about the issue of language attrition. References Anderse n, R. W. (1982). Determining the linguistic attributes of language attrition. In R.D. Lambert & B.F. Freed (eds.). The loss of language skills, (pp ). Rowley, Mass: Newbury House. Bahrick, H. (1984). Fifty years of second language attrition: implications for programmatic research. Modern Language Journal, 68, Cohen, A. D. (1974). Culver City Spanish immersion program: How does summer recess affect Spanish speaking ability? Language Learning, 25, Craik, F., & Lockhart, R. (1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behavior, 11, Edwards, G. (1976). Second language tetention in the Canadian public service. The Canadian Modern Language Review, 32, Galbraith, N. (1981). A study of transfer in language loss: Some problems and considerations. Unpublished master s thesis, University of Utrecht. LeSieg, T. (1974). Wacky Wednesday. New York: Random House. Mayer, M. and M. Mayer. (1971). A boy, a dog, a frog and a friend. New York: Dial Books. Tomiyama, M. (1999). The first stage of second language attrition: A case study of a Japanese returnee. In Lynne Hansen (Ed.), Second language attrition: Evidence from Japanese contexts (pp ). Oxford: Oxford University Press. Tomiyama, M. (2000). Child language attrition: A longitudinal case study. Applied Linguistics; 21 (3), Weltens, B., & van Els, T. (1986). The attrition of French as a foreign language: Interim results. In Bert Weltens et al. (Eds.), Language attrition in progress (pp ). Dordrecht: Foris Publications. The socio-psychological motivation for code-switching/mixing: the Maithili-Hindi situation Abstract Girish Jha Jawaharlal Nehru University, India girishj@mail.jnu.ac.in Code-switching/mixing (CS/CM) is a very common phenomenon among bi/multi-linguals under different socio-psychological motivations. In a country like India, where multi-lingualism is the norm, switching from one language to another or mixing two codes is part of people s everyday use of spoken language. The Mithila region in north India presents a curious socio-linguistic case for linguists. Maithili has a 2.3 million strong linguistic community. It has been recently included in the Indian constitution as one of the 18 national languages. It is also recognized as one of the 22 Indian languages of the Indian academy of literature, and is the second official 103

117 ASIALEX 2005 language of Nepal. For linguists, the sociolinguistic diversity of Mithila offers an interest ing case-study. Maithili is a language known for its caste dialects. People living in the same village are known to have distinct linguistic variations based on their caste and com munity. The use of Hindi in this region follows a different course as the state and ce ntral official government communication medium, as the primary medium of instruction in most schools and colleges, as the vehicle of print and electronic media including films and music, as a tool for bilingual communication, and also, by way of CS/CM. The goal of the paper is to understand the phenomena of code-switching and mixing by the Maithili speakers of this region and try to ascertain the complex socio-psychological or contextual reasons which characterize its prevalence and persistence. [Please turn to page 401 for the full paper] Information Technology applications for Sanskrit lexicography: the case of Amarakosha Girish Nath Jha Sudhir K Mishra R. Chandrashekhar Jawaharlal Nehru University, India girishj@mail.jnu.ac.in mishra_skumar78@yahoo.com.in chandra_shekar_r@yahoo.com Abstract India has seen amazing strides in Information and Communication Technology (ICT) applications for Indian languages in general and for Sanskrit in particular. Since Machine Translation from Sanskrit to other Indian languages is often the desired goal, traditional Sanskrit lexicography has attracted a lot of attention of ICT and Computational Linguistics community. While several attempts are being made to build word-nets on traditional Indian epistemological and logical principles, the need for generating a multilingual lexical resource for Amarakosha, the Sanskrit lexicon built on ontological principles, has been largely ignored. Amarakosha, the 4th CE lexicon build by Amarasimha has influenced modern lexicographic techniques in quite the same way as Panini and Chomsky have done to generative linguistics. The present work is an effort in this direction to build a multilingual lexical resource for Amarakosha which not only lets users store ontological equivalents of Sanskrit concepts in their languages, it also lets them search and edit. The work has tremendous applications in Word Sense Disambiguation (WSD) process of Machine Translation Systems, in Knowledge Representation, and in language pedagogy. The present work is an ongoing project with a Java front-end and relational database server as backend. At present, it has the following features facility for online multilingual data entry in Sanskrit, Hindi, Kannada and English. Currently, the data is being stored in itrans. Scope for Unicode is provided stores up to 50 synonyms with grammatical information and detailed glosses cross-referencing among synonyms 104

118 WORDS IN ASIAN CULTURAL CONTEXTS search capability in the supported Indian language Structure of Amarakosha As a text, AK has three kaandas (chapters), each subdivided into vargas (classes). The first and second kaandas have 10 vargas each. The third kaanda has 6 vargas. The following diagram illustrates the structure of AK amarakosha prathamakaanda dvitiyakaanda tritiyakaanda svarga bhuumi visheshyanighna vyoma pura sankirna dig shaila naanaartha kaala vanaushadhi avyaya dhi sinhaadi lingaadisangraha shabdaadi manushya vyaakhyaakaropasamhaara naatya brahma paataalabhogi naraka vaari kshatriya vaishya shuudra There have been attempts to put the text of AK online or in digital formats. But there has been no attempt to create a version of this work which not only lets user interactively build a database of AK but also lets them search and test. Previous work There have been some efforts in building a bilingual Sanskrit electronic dictionary. Gerard Huet (2004) is working on creating a Sanskrit French electronic dictionary. Louis Bontes (2005) had built a stand-alone PC based version of Monier Williams dictionary. The Cologne Digital Sanskrit Lexicon contains Monier-Williams Sanskrit-English Dictionary with approx. 160,000 main entries. It has an online search facility in both Sanskrit and English. Capeller s Sanskrit-English Dictionary has been converted to a digital format similar to the Cologne project and has online search facility in both Sanskrit and English. It however has only 50,000 entries. Apte Sanskrit Dictionary Search is a web Sanskrit dictionary based on the famous work of V. S. Apte - The Practical Sanskrit-English Dictionary. Andre Signoret s French-Sanskrit dictionary is freely downloadable from the net. The BharatiyaBhasha multilingual dictionary built by Central Hindi Directorate, New Delhi under TDIL, Govt. of India funding consists of nearly 5000 common words in 14 different languages. It is available for download from the TDIL site. The Sanskrit Dictionary-Database being prepared by Jong-cheol Lee (2005), Academy of Korean Studies, Seoul, Korea will include mappings among Sanskrit, Tibetan, Chinese and Korean. Mohanthy et al (2004) have done some work on representing ontologies of Sanskrit words using Navya Nyaaya methodology. As we can see, none of these works focuses on AK and its rich semantic ontologies. In terms of search, the Cologne and Capeller s works are comparable. 105

119 ASIALEX 2005 System Components The online system is being developed using Java servlets as front-end hosted on Tomcat platform, and MS SQL server as backend. The system has the following components - the relational database the data entry component search component detail search component data editing component The AK Database The AK database is a relational database designed using MS SQL server objects and procedures. The database includes intricate relationship between the base words, synonyms and multilingual glosses. Data entry component This component lets users enter data in their language by way of user validation by password checking. Users can select the basic Sanskrit entry and provide information in their language (at present only English, Hindi included). The screen capture for this page is given below - after the basic data, the synonyms (up to a maximum of 50) can be relevant information. Data Search entered with other Search is carried out by the base word (Sanskrit), English, Hindi and Kannada at present. There is scope for more languages in future. After entering the word and clicking search button as shown in the following image 106

120 WORDS IN ASIAN CULTURAL CONTEXTS the system will deliver the following results including all the synonyms it found for the search string - Clicking on the synonym link takes to the details page where each synonym entry is explained according to AK. The image that follows shows the synonym details for the English search string horse - Data Edit This module lets language experts edit wrongly entered data 107

121 ASIALEX 2005 Limitations and future enhancements Currently, the system has only Romanized data entry scheme. Future upgrade will include multilingual scripting with Unicode. At present, we have only three languages, and we would like to include as many as possible depending on the resources at our hand. A Sanskrit textual analysis based on AK database will be added in future. The system thus presented here is the first Sanskrit Shastra made available in this format. We aim to include more and more texts in future. Conclusion The system thus presented here is an attempt to make available a Sanskrit shastraic text online in interactive form. The resulting database can be used for Sanskrit and Indian languages WSD and for Sanskrit to Indian languages Machine Translation. It can also be used for multilingual concept acquisition. References Huet Gerard (2004), Sanskrit French dictionary, (accessed : 23 March 2005) Bontes Louis, Monier William Digital Dictionary, (accessed : 23 March 2005) Cologne Digital Sanskrit Lexicon, (accessed : 23 March 2005) Capeller s Sanskrit-English Dictionary, (accessed : 23 March 2005) Apte Sanskrit Dictionary Search, Andre Signoret s French-Sanskrit dictionary, (accessed : 23 March 2005) BharatiyaBhasha multilingual dictionary, (accessed : 23 March 2005) Sanskrit Dictionary-Database of Jong-cheol Lee, Academy of Korean Studies, Seoul, Korea (accessed : 23 March 2005) Mohanty et al (2004), Ontological analysis in Sanskrit wordnet, Procs of ICSLT-O-COCOSDA, New Delhi, 2004 Representing and describing words flexibly with the dictionary application TshwaneLex David Joffe TshwaneDJe HLT david.joffe@tshwanedje.com Abstract Gilles-Maurice de Schryver Ghent University & TshwaneDJe HLT gillesmaurice.deschryver@ugent.be, gillesmaurice.deschryver@tshwanedje.com This paper describes the fully customisable and built-in DTD (Document Type Definition) editor of TshwaneLex. This powerful tool, which is based on XML 108

122 WORDS IN ASIAN CULTURAL CONTEXTS (extensible Markup Language) standards, allows lexicographers to tailor the dictionary grammar of any project, and thus to truly represent and describe words flexibly. The demands placed on modern dictionary databases A lexicographer typically summarises each word s analysis in the form of a meticulously constructed dictionary article. Thousands of such analyses are then brought together in reference works. Today s reference works are electronic, so are the databases underlying them. With representations and descriptions of words becoming increasingly multifaceted and interlinked, the demands placed on modern dictionary databases grow exponentially. In this paper it is shown how one of the main challenges, namely that of providing a customisable DTD to lexicographers, was met in the dictionary application TshwaneLex. TshwaneLex is a true hybrid in that it allows for the creation of monolingual, bilingual and semi-bilingual dictionaries, for virtually any language, thanks to full Unicode support, as well as support for the Windows Input Method Editors ( soft keyboards ). At the heart of TshwaneLex lies a fully customisable DTD with which the dictionary grammar may be modified by the lexicographers themselves, and this without the need for an IT expert. Although the mindset that creating and modifying a DTD is something you give to your IT expert to do for you is in the dictionary industry already, TshwaneLex offers users the choice of doing this themselves if they want / need to without an IT expert. Basics of hierarchical dictionary data modelling in TshwaneLex The basic structure of each dictionary article is hierarchical. One lemma may contain several word senses, and each of those word senses may in turn contain sub-senses. Any of the senses or sub-senses may contain usage examples or MWUs, where the latter may again in turn contain one or more senses, etc. When creating a dictionary article, the lexicographer s task can be seen as consisting of two essentially separate but closely related sub-tasks: (1) to specify the basic skeletal structure or layout of each dictionary article (represented as a Tree View in TshwaneLex), and (2) to flesh out that basic structure with content (which may be inputted in or selected from sub-windows accessible with the function keys F1 and F2 in TshwaneLex). Borrowing terminology from XML, the root and branches of the Tree View may be called elements, the text content values attributes. There are different types of elements, and every element type has its own set of attributes that may be associated with it. For example, a Lemma element could consist of the lemma sign itself, a pronunciation field, and an etymology field. For a usage-example element, say E.G., there may be three attributes, viz. the usage example itself, a translation of that example (if a bilingual dictionary), and optionally the source (citation). When one clicks on any element in the Tree View, the Attributes (F1) sub-window shows the attributes associated with the type of element one has clicked on, and one may immediately proceed with entering or modifying the relevant contents. These links between (1) internal article structure, (2) Tree View, and (3) boxes / fields to be filled in by the user, constitute one of the crucial design features to ensure smooth and sound compilation. Only the relevant attributes are seen at any given point, so that the screen does not get clogged, and so that the potential for errors is minimised as only attributes allowed by the DTD are editable. 109

123 ASIALEX 2005 The Document Type Definition in TshwaneLex A DTD, also known as a schema, is a description of the structure of the articles in a particular dictionary, or essentially, a description of the types of elements that a dictionary has, as well as the attributes that each of those elements may have. Every dictionary document needs a DTD, and this is typically mostly set up as completely as possible at the start of a dictionary project, with perhaps minor modifications later on. The DTD is then used while creating the dictionary, and by enforcing conformance to a well-designed DTD, one can ensure that the final dictionary follows a logical and consistent structure throughout. When work on a new dictionary begins, TshwaneLex creates a basic default DTD. This may then be customised by the user, although a few DTD element and attribute types are fundamental to TshwaneLex and should not be changed, such as the Lemma element s LemmaSign and HomonymNumber attributes (as the program automatically calculates homonym numbers). Three basic aspects are defined by a DTD: (1) valid element types, (2) element attributes, and (3) element child relations. In Figure 1, a screenshot of the DTD structure tab of the TshwaneLex DTD editor dialog is shown. The full list of all the valid element types that may appear in this particular dictionary, appears on the left under Element types. Each element in the DTD has its own element attributes or set of attribute types. When an element is selected in the list of all the element types, the Attributes of this field section displays a list of the attributes for that element. For example, if the Lemma elemen t is selected, the basic Lemma attributes are shown, such as LemmaSign, Pronunciation, Etymology, etc. (These are of course also the attributes that are shown under Attributes (F1) when a node of this element type is selected in the Tree View while editing the dictionary. The names and order for these attributes is also configured here.) Figure 1. TshwaneLex DTD structure editor dialog 110

124 WORDS IN ASIAN CULTURAL CONTEXTS The list of elements at the left of the DTD structure editor dialog is a flat list, and does not describe which element types are allowed to be added under which other element types when adding elements in the Tree View. Clearly only certain elements may be added to others. For example, one cannot add a Lemma element beneath a usage example. One therefore needs to define element child relations, which describe the allowed child element types of each element. This is done in the Allowed child fields section in the DTD structure editor dialog. This section contains two lists. The left one, with header Child fields, is a list of the currently allowed child elements for the selected element type. The list on the right is simply a list of all element types, from which to choose to add new child types. Each parent-child element relation may also have a constraint associated with it. There are four possibilities, which may be selected using the Child type drop-down list in the Allowed child fields section of the dialog: (1) One child of this type only, (2) Zero or more children of this type, (3) One or more children of this type (i.e. at least one ), and (4) Zero or one child of this type. When compilation proper proceeds in the Tree View, TshwaneLex will of course not allow these constraints to be broken and will grey out certain options when needed, preventing lexicographers from creating invalid articles and thus ensuring consistency throughout the dictionary. The styles system in TshwaneLex The styles system in TshwaneLex is used to transform the structured data of each dictionary article into the corresponding output in the preview area, and is also sed when exporting to formats such as RTF or HTML. Figure 2. TshwaneLex styles/formatting editor dialog 111

125 ASIALEX 2005 Every element and every attribute in TshwaneLex may have its own style. A style allows various formatting options (such as font face, font size, text colour, bold, italics, underline, superscript, subscript, etc.) to be used for that element or attribute. The styles in TshwaneLex can be modified in the styles editor dialog, as seen in Figure 2. Figure 2 shows a list of all available styles in a particular dictionary, in the column headed by Styles. Element styles are highlighted in grey, and the attribute styles for each element are shown beneath it. The style of a selected element or attribute from this list may be modified in the Style properties section on the right. Style changes are applied immediately in the preview area, thus one may instantaneously see the effects of style changes while changing the styles. These changes are moreover effected throughout the entire dictionary. In addition to formatting options such as bold and italics, common text and punctuation that should appear before or after specific elements or attributes can be configured with the style. This can be used, for example, to place the POS between round brackets, the etymology between square brackets, etc. By using styles, this sort of punctuation may be changed throughout the entire dictionary at once, simply by changing the style. The lexicographer is furthermore also freed from entering such text or punctuation manually. Note that there are, for each style, two Text before and Text after options. The first of these, labelled uses style formatting, will be output with the other formatting options of the style (e.g. bold) applied. The other options, labelled before style formatting and after style formatting will be output before / after the formatting options of the style are applied. Element styles have additional properties that attribute styles do not have. These are shown in the bottom right area of the styles editor dialog under Style properties for groups of multiple child elements or multiple list items. These are options that are used for groups of elements. One is dealing with a group of elements when there are multiple child nodes of that element type under another node in the tree (e.g. if the Lemma element has multiple word senses then there is a group of Sense elements, if a Sense element has multiple Definition elements then there is a group of Definition elements). The element group style options allow specified text to appear before or after the entire group of child nodes of the same type, as well as between each element, allowing, for instance, a comma or semi-colon to be used to separate multiple definitions or usage examples. Element and attribute output order in TshwaneLex The output order of elements and attributes in the preview area (and in formatted output such as RTF or HTML) is configured separately to the DTD. This allows some separation of the underlying data structure and the final visual output. The output order of elements is configured using the Output (display) order tab in the DTD editor dialog. For each element type, one can configure the output order of both child elements of that element type, and attributes of that type all relative to one another. Compare in this regard Figure

126 WORDS IN ASIAN CULTURAL CONTEXTS Figure 3. TshwaneLex output (display) order editor dialog In Figure 3, the output (display) order for the Lemma element type is shown. This element has a number of attributes such as LemmaSign, Pronunciation and Etymology, as well as, among others, a number of child element types, including child senses ( Sense ). One sees that some of the Lemma attributes (such as LemmaSign and Pronunciation) appear before those child senses, while other Lemma attributes (such as Etymology) appear after the child senses. The output order of this list of attributes and child fields for Lemma can now easily be modified by using the Move up and Move down buttons. As another example, whether the HomonymNumber attribute should be placed before or after the LemmaSign attribute for all articles throughout the central text for the production of a certain dictionary, may simply be decided at the output stage. Each child element or attribute in the output order configuration may furthermore also be selected to be hidden in the preview area or output. This can be done by toggling the state of the Visible checkbox. This allows any fields to be marked as non-printing fields, thus fields that should not appear in the output of a certain dictionary. Customisable DTDs in TshwaneLex: Rather formidable and intimidating? No doubt, to new users the DTD editor dialog as a whole will appear to be rather formidable and intimidating. Yet it is worth investing some time in studying its mechanics. With this customisable DTD users now have an entire printing house at their fingertips. Compare in this regard the following advice by Zgusta in the recent anthology of lexicography: A matter of great practical importance is the decision concerning the font types in which the dictionary will be printed; this problem should be given attention from the 113

127 ASIALEX 2005 very beginning of the work and should be finally decided as soon as the lexicographer begins to write entries. It would seem that there is no problem in this decision, that the headword is simply printed in bold type, the rest of the lemma in smaller italics, the definitions in italics, the examples in roman, etc. In reality, it is not an easy task to make all these decisions; they should be made after a thorough consultation with the prospective printer. (Zgusta 2003: I 77; being a revision of Zgusta 1971: 353) This might indeed have been the case until a few decades / years ago, and might actually still be the case at various dictionary publishing houses where the bonnet cannot be taken off their typesetting programs. However, as should be clear from the discussion of the styles system in TshwaneLex, the layout aspects Zgusta mentions can trivially be changed at any point in time, and can also be adapted for the production of any type of dictionary. Moreover, TshwaneLex offers a true database in the literal sense of this word, in that various products may be derived from it that contain selections of the base of data, and these data may be presented in whatever order is appropriate for the production of a particular dictionary. On top of all this, the customisable DTD safeguards the dictionary grammar at all times. References TshwaneLex ( ), Online info, (accessed: 31 March 2005). Zgusta, Ladislav (1971), Manual of lexicography (Prague: Academia). Zgusta, Ladislav (2003), Planning and organization of lexicographic work, in Reinhard R.K. Hartmann (ed.), Lexicography: critical concepts (London: Routledge), pp. I (Ch. 3). Lexical association network maps for basic Japanese vocabulary Abstract Terry Joyce Tokyo Institute of Technology terry@valdes.titech.ac.jp Extending psycholinguistic research into the lexical representation of two-kanji compound words within the Japanese mental lexicon (Joyce, 2002, 2004), this paper reports on a large-scale word association survey for basic Japanese vocabulary. The database of word association norms, which is being compiled from various survey formats including a web-based version of the survey, supplements existing databases concerning the lexical features of Japanese vocabulary (Amano & Kondo, 1999; Yokoyama, Sasahara, Nozaki & Long, 1998), such as familiarity ratings and frequency counts, which are essential for cognitive science research. A particularly promising application of the word association norms data, however, is the creation of lexical association network maps that capture important proprieties of words and their interconnectivity. These maps complement other approaches that attempt to tap into aspects of lexical knowledge, such as WordNet, thesauri, ontologies, and collocation data, while avoiding some of their problems. There are also direct and interesting lexicographical and Japanese language learning applications of the 114

128 WORDS IN ASIAN CULTURAL CONTEXTS database of word association norms and the lexical association maps. In addition to providing more user-friendly means of consulting the lexical entries of electronic dictionaries, the inclusion of the word associative normative data within the lexical entry will also greatly enrich the variety of lexical information. Given the importance of contextual relevance for second language learning, the lexical associative networks also open up extremely effective learning strategies for Japanese language education. 1. Introduction Many areas of cognitive science, such as psychology, artificial intelligence, as well as natural language processing and computational linguistics, are contributing to our understanding of the mental lexicon and the representation of lexical knowledge. Reflecting, on the one hand, the conviction that associative processes are basic mechanisms underlying cognition (e.g., Cramer, 1968) and, on the other hand, notions of activation spreading through lexical networks (e.g., Collins & Loftus, 1975), much attention has been devoted to investigating the rich networks of word associations that connect related words together. For instance, Nelson and McEvoy (2003) have recently demonstrated the influence of word associations on cognition, showing that differences in the associative structures of known words in terms of associate set size, resonance (or backward association), and the connectivity within an associate set effect performance on memory tasks. This paper reports on a project to compile a large-scale database of word association norms for basic Japanese vocabulary and to utilize the data to create lexical association network maps that capture important properties of words and their connectivity. After briefly noting some psycholinguistic research into the representation of two-kanji compound words within the Japanese mental lexicon (Joyce, 2002, 2004), which highlights the need for comprehensive word association data for cognitive science experimentation, the paper outlines the compilation and coding of an initial survey corpus and data-collection work. Finally, the paper touches on utilizing the normative data to create lexical association network maps for basic Japanese vocabulary, singling out two promising applications of the maps: namely, as an approach to modeling the semantic representations of connectionist models and in the areas of Japanese lexicography and language learning. 2. The Japanese lemma-unit model (Joyce 2002, 2004) Compounding is an extremely productive word-formation process in Japanese, making the language particularly important for examining the extent of morphological involvement in the organization of the mental lexicon. In that context, Joyce (2002, 2004) has examined the lexical representation and retrieval of two-kanji compound words by conducting constituent-morpheme priming experiments comparing the facilitation from component kanji in the lexical decision task with various kinds of compound words, such as modifier + modified, verb + complement and the reversed order of complement + verb, and synonymous pairs. The general results from the experiments are that both constituent prime conditions facilitate lexical decisions and in the majority of cases at similar levels, indicating that morphology is important in the organization of the mental lexicon. The research has also provided findings pointing to the importance of verbal information. In experiments that manipulated the positional frequency of the verbal constituents in verb + complement and complement + verb compound words, reversed patterns of 115

129 ASIALEX 2005 priming were observed in high positional frequency conditions where the reaction times for verbal constituent conditions were significantly faster than in the respective complement conditions. To account for the lexical representation and retrieval of two kanji compound words, Joyce (2002, 2004) has proposed adapting for the Japanese mental lexicon a version of the multilevel interactive-activation framework. A feature of the model is the incorporation of lemma-unit representations that mediate the connections between access representations (both orthographic and phonological form representations) and semantic representations, which is especially appealing for handling the complex nature of the Japanese writing system. While this visual word recognition research indicates a central role for morphological information in the organization of the Japanese mental lexicon, clearly possible confounding factors, such as association effects, need to be examined further. However, because comprehensive Japanese word association data is currently not available, a central objective of the present research is to construct a large-scale database of word association norms for basic Japanese vocabulary. 3. Word association norms for Japanese After noting some word association databases for English and Japanese, this section outlines the creation of the initial survey corpus and data-collection preparations. While Moss and Older (1996) have collected between responses for some 2,400 words of British English, Nelson, McEvoy and Schreiber (1998) have compiled the largest database for American English covering some 5,000 words with approximately 150 responses per item. In terms of Japanese language surveys, although the early survey by Umemoto (1969) is well-known and has responses from 1,000 university students, the word corpus is very small with only 210 words. More recently, Ishizaki (2004) has collected word associations for 1,656 nouns for use in building an associative concept dictionary. A drawback with this data, however, is the fact that response category was specified, so it tells us little about free association norms. In creating an initial corpus of basic Japanese vocabulary for the word association survey, three primary reference sources were used. The first was the survey of basic vocabulary for Japanese language education conducted by the National Language Research Institute (1984), with 6,800 words including a core set of about 2,200 words. The second source was a more recent list of about 4,000 words prepared by Tamamura (2003), while the third reference was a handbook of Japanese orthography (Sanseidō Henshūjo, 1991) listing sanctioned readings for all Jōyō kanji. Based on comparisons of these sources, an initial survey corpus of 5,000 kanji and words has been created, and this will be expanded through the inclusion of associate responses as data collection proceeds. The discrete free word association task is relatively straightforward the respondent is simply asked to provide the first meaningfully-related word that comes to mind when presented with a stimulus word. However, a major concern for the present project in seeking to efficiently construct a large-scale database has been to devise an automatic method of generating multiple individual respondent survey lists from the survey corpus, while minimizing as far as practically possible the effects of intra-list association. Accordingly, much of the preparatory efforts have been devoted to coding the survey corpus with information relating to the following criteria: (1) Pronunciation; 116

130 WORDS IN ASIAN CULTURAL CONTEXTS necessary due to the high incidence of homophones and to explore the influence of orthographic variants (i.e., kanji versus kana). (2) Orthographic type (i.e., single kanji, multi-kanji, and mixed kanji-kana words); to reduce respondent strategies, orthographic type is mixed in each survey list. (3) Component kanji codes; to ensure that a given kanji only appears once in a survey list. (4) Semantic category codes based on the NTT thesaurus (Ikehara, et al, 1999); to ensure that survey lists do not contain multiple words from a given semantic category. (5) Unique ID code; assigned ID codes will serve as additional measure to eliminate intra-list associations as data collection proceeds. In order to obtain the large-scale quantities of responses required for the database, a web-based format of the survey is being developed (which should be available shortly). However, data collection is already underway employing traditional paper questionnaires. For the first block of data, a random sample of 2,000 items was taken from the corpus, and approximately 100,000 responses were collected from about 1,000 undergraduates (100 items per questionnaire). While the questionnaire format involves data inputting burdens, it is important for addressing reliability issues. A second block of data is currently being collected with questionnaires covering the remaining items in the survey corpus. VENUS PLUTO UNIVERSE STAR PLANET MARS MOON Figure 1. Basic concept of lexical association network maps (based on Nelson & McEvoy, 2003) SPACE SATURN EARTH 4. Association data applications The large-scale database of free word association norms for basic Japanese vocabulary will be a valuable resource for cognitive science research, such as memory and visual word recognition experiments. The database of association norms also makes it possible to create lexical association network maps as a means of representing associate sets and their connectivity. Figure 1 illustrates the basic concept of lexical association network maps, with an example for the English word planet based Nelson and McEvoy (2003). In addition to representing associate set size, the maps can highlight differences in association strengths, both forward and backwards (which are independent), as well as the association density of the associate set. The lexical association network maps can complement other approaches to capturing aspects of lexical knowledge, while avoiding some of their problems. For example, unlike the lexicographer s expertise required to define synonym sets in WordNet (Fellbaum, 1998), the maps are based on the free responses from respondents. Thus, the structure in the network emerges from tapping into free associations, rather than being theory-driven or constrained by specific relationships, which are concerns for associative concept dictionaries and ontologies where the hierarchical structure is pre-defined. The use of free associations also avoids domain biases that are an issue 117

131 ASIALEX 2005 for collocation data. Figure 2 resents two examples of associates sets, for the Japanese words 冬 winter and 集める gather, collect, based on the free word association responses collected so far. The enclosed figures on the arrow connections represent the percentage of responses. As the figure s hows, 冬 winter has a very strong primary associate with the word 寒い さむい cold, which accounts for 44 percent of all responses. The second associate of 雪 snow represents only 15 percent of the responses, followed by 夏 summer and 冬至 winter solstice, both at 6 percent, and 白 白い white at 4 percent. Thus, 冬 winter has a relatively small set of core associates with one particularly strong associate. こたつかまくら北休み 休息春 寒い さむい雪 冬 くま冬眠越冬 夏 冬至白 白い氷 冬将軍切ない フィギュア 密集標本コレクトカン 大人買い ガラクタおち葉 集める 収める集まる 趣味捨てる 人集合ゴミコレクターコレクション 集会 Figure 2. Associate sets for 冬 winter and 集める gather, collect In contrast, the verb 集める gather, collect has a larger set of core associates, お金 切手収集 but naturally with weaker association strengths. The primary associate here is お金 金 money accounting for 15 percent of the responses. There are also two secondary responses at 10 percent; namely, 切手 stamps and 収集 collection. Some of the remaining core associates are 人 people (8%), 集合 set (6%), ゴミ rubbish, trash (6%), and コレクター collector (6%). Consistent with their respective word classes of noun and verb, these two words exhibit different kinds of syntagmatic responses. Compared to the very strong association between the adjective 寒い さむい cold and the noun 冬 winter, more of the core responses for the verb 集める gather, collect are nouns that could either occupy the direct object slot (i.e., 切手 stamps, 人 people, ゴミ rubbish, trash ) or the subject slot (i.e., コレクター collector ). In capturing the networks of associations that exist between words an important aspect of lexical knowledge there are promising applications for the lexical association network maps. One such application is in modeling the semantic representations within connectionist models of the Japanese mental lexicon, like the Japanese lemma-unit model (Joyce, 2002, 2004). However, there are also direct and interesting lexicographical and Japanese language learning applications of the word association database and the lexical 118

132 WORDS IN ASIAN CULTURAL CONTEXTS association network maps. For instance, the inclusion of word association norms within the lexical entry would greatly enrich the variety of lexical information presented to the dictionary user. So, in addition to listing information relating to the entry word s pronunciation, its definitions, inflectional or derivational forms, and idiomatic expressions, the dictionary could also provide word association information, which together with response frequency data, would assist in identifying for a given entry word the relative importance of different kinds of associative relations, such as synonyms, antonyms, and hyponyms, as well as common modifiers and complements, and various other relations. Supplementing electronic dictionaries with the word association normative data and the lexical association network maps could also support more user-friendly search functions (Zock & Bilac, 2004). In daily life, people regularly consult dictionaries to check the spelling of a known word (or the strokes of a kanji character) or to find out the meaning of a newly encountered word. However, electronic dictionaries, and even thesauri organized primarily according to synonyms, are of little help to the user experiencing the common tip of the tongue phenomenon (Brown & McNeill, 1966). Typically in this situation, while the individual is unable to retrieve the desired word from their mental lexicon, they are often able to provide information about some of the word s form features, such as its initial and/or final letters, or part of its pronunciation. Unfortunately, these are not the kinds of inform ation that can be used effectively to search electronic dictionaries. Ho wever, in addition to knowing some of the target word s form characteristics, the indiv idual will also be aware of various kinds of semantic information related to the target word, such as the associative relations the word has with other w ords. In contrast to partial information ab out form features, which is too impr ecise to be of any value for retrieval, known semantic information is potentially much more useful. If the word association normative data and the lexical association network maps were incorporated into electronic dictionaries, then inputting an associated word (either single or multiple word entry) could provide access to the relevant lexical association network map, from which the user could follow appropriate association links until the target word is identified. Given the importance of contextual relevance for second language learning, the lexical associative networks also open up extremely effective learning strategies for Japanese language education. Memory researchers have long demonstrated that the categorization and semantic organization of stimulus materials have dramatic effects on retrieval performance (e.g., Bower, Clark, Winzenz, & Lesgold, 1969). Accordingly, the lexical association network maps for basic Japanese vocabulary could be a very useful resource in the context of teaching Japanese as a foreign language. To the extent that the goal of second language teaching is to instill in the learner lexical knowledge for the target language that approaches that of the native speaker, the lexical association network maps, depicting sets of associatively-related words based on the free word association responses of native Japanese speakers, represent a kind of authentic study material and a value reference source for the second language learner in judging the naturalness of lexical combinations. In pursuing these Japanese lexicographical and language learning applications of the database of word association norms and the lexical association network maps, this research project is also working to create a comprehensive kanji database and integrated kanji instruction system. The key concept behind the integrated kanji instruction system is an electronic study notebook that would build into a personalized dictionary 119

133 ASIALEX 2005 by drawing on the database reference source through various learning assignments, ranging from basic look-up tasks to more advanced expansion tasks that would help the learner develop a deeper understanding of kanji structure and the morphological principles of compound words. Acknowledgements This research project is part of the 21st Century COE Program Framework for Systematization and Application of Large-scale Knowledge Resources (program leader; Professor Sadaoki Furui) at the Tokyo Institute of Technology, Japan. References Amano, Shigeaki, and Kondō, Tadahisa (Eds.), (1999), Nihongo no goitokusei [Lexical properties of Japanese] Vols. 1-6, NTT database series. Tokyo: Sanseidō. Bower, G. H., Clark, M. C., Winzenz, D., and Lesgold, A. (1969), Hierarchical retrieval schemes in recall of categorized word lists, Journal of Verbal Learning and Verbal Behavior, 8, Brown, R., and McNeill, D., (1966), The tip of the tongue phenomenon. Journal of Verbal Learning and Verbal Behavior, 5, Collins, Allan. M., and Loftus, Elizabeth. F. (1975), A spreading-activation theory of semantic processing, Psychological Review, 82, Cramer, Phebe, (1968), Word association. New York & London: Academic Press. Fellbaum, Christiane (Ed.), (1998), WordNet: An electronic lexical database, Cambridge: MIT Press. Ikehara, Satoru, Miyazaki, Masahiro, Shirai, Satoshi, Yokoo, Akio, Nakaiwa, Hiromi, Ogura, Kentaro, Ōyama, Yoshifumi, and Hayashi, Yoshihiko, (1999), Nihongo goi taikei [Goi-taikei-A Japanese lexicon] (CD-Rom), Tokyo: Iwanami Shoten. Ishizaki, Shun, (2004), Rensō gainen jisho Version 1.0 CD. Joyce, Terry, (2002), Constituent-morpheme priming: Implications from the morphology of two-kanji compound words, Japanese Psychological Research, Blackwell: Japan, pp Joyce, Terry, (2004), Modeling the Japanese mental lexicon: Morphological, orthographic and phonological considerations, In Serge P. Shohov (Ed.), Advances in Psychological Research, Vol ume 31, (pp ). Hauppauge, NY: Nova Science. Moss, Helen, and Older, Lianne, (1996), Birkbeck word association norms, Hove: Psychological Press. National Language Research Institute, (1984), Nihongo kyōiku no tame no kihon goi chōsa, Tokyo: Shuei Shuppan. Nelson, Douglas L., and McEvoy, Cathy L., (2003), Implicitly activated memories: The missing links of remembering, Fourth Tsukuba International Conference of Memory (Human learning and memory: Advances in theory and application), January, Epochal Congress Center, Tsukuba, Japan. Nelson, Douglas L., McEvoy, Cathy L., and Schreiber, Thomas A., (1998), The University of South Florida word association, rhyme, and word fragment norms, Sanseidō Henshūjo, (1991), Atarashii kokugo hyōki handobukku (Dai yonhan), Tokyo: Sanseidō. Tamamura, Fumio, (2003), Chūkyūyō goi: Kihon 4000 go, Nihongo Kyōiku, Tokyo, pp Umemoto, T., (1969), Rensō kijunhyō: Daigakusei 1000 nin no jiyū rensō ni yoru, Tokyo: Tokyo Daigaku Shuppankai. Yokoyama, Shōichi., Sasahara, Hiroyuki, Nozaki, Hironari, and Long. Eric, (1998), Shinbun denshi media no kanji: Asahi Shinbun CD-R OM ni yoru kanji hindohyō [Electronic newspaper media kanji: Kanji frequency lists based on Asahi Newspaper CD-ROM]. Tokyo: Sanseidō. Zock, Michael, and Bilac, Sl aven, (2004), Word lookup on the basi s of associations: From an idea to a roadmap, COLING2004 Workshop on Enhancin g and using electronic dictionaries, August, Geneva. 120

134 WORDS IN ASIAN CULTURAL CONTEXTS Multi-lingual (KJCE) lexical database using ILI Beom-mo Kang Jae-eun Cha Korea University Korea University Abstract This paper introduces the Multi-lingual (Korean, Japanese, Chinese and English) Lexical Database which has been carried out by the Institute of Korean Culture of Korea University since the year The aim of this Project is to construct a M ultilingual Lexical Database con taining 50,000 entries until August, Since the WordNet and the EuroWordNet were developed, a concern about the Multi-lingual database has been increased in Korea, but East Asian languages have not yet any ground to use each wordnet. So we made a list of Korean word senses by way of refining the senses in the Korean lexica l DB of the Sejong project and then made each sense be connected to the corresponding English, Chinese and Japanese sense. We use EWN s ILI in order to connect the Korean with other languages. Since it is not possible to automa tically get the offset number from ILI, we made an input tool which makes it possible to get it semi-automatically. We compared the 5409 Korean noun senses and 8032 Korean noun collocations with corresponding Chinese, Japanese, English, and ILI in this paper. There are relatively many word-to-phrase correspondence between Korean-English: a Korean word saengseon corresponds to the Chinese word xiān yú, or Japanese word sengyo, but, in English it corresponds to a phrase as raw fish. We found no concept in ILI which corresponds to saengseon. It is supposed that since East Asian people eat sliced raw fish or dried fish, they need a concept which expresses the freshness of fish like saengseon, xiān yú, sengyo. Objectives The Institute of Korean Culture of Korea University has been carrying out the Central Support Research Institute with the aid of the Korea Research Foundation(KRF) under the title of Research of Multi-Lingual Information Processing Methodologies and Database Development for Globalizing Korean Studies since the year *This research was supported by KRF Grant (code: A20002) As the main task, we have pursued this research project entitled Research and Model Development for a Multi-Lingual Lexical Database (hereinafter referred to as KJCE DB). The aim of the KJCE DB Project is to construct a practical Multi-Lingual (Korean, Japanese, Chinese, and English) Lexical Database which will contain about 50,000 entries until August, The structure of the KJCE DB Most European languages have developed its own wordnet. However, East Asian languages have not set any ground on using each wordnet. Therefore, the Multi-Lingual Lexical Database which is in construction arranges the information of 121

135 ASIALEX 2005 corresponding terms for English, Chinese, Japanese based on the Korean language. By making use of the ILI from the EuroWordNet, the database has the following structure in attempt to expand other languages. Figure 1. The structure of the KJCE DB Top EWN ILI Other languages KJCE DB Concretely, the design for noun and verb in the KJCE DB is as the following table (* means repeatable item). noun Table 1. The design for noun and verb in the KJCE DB verb entry MntGrp entry MntGrp HeadGrp HeadGrp Sense* IdiomGrp Lr ILI coll* Idiom* Sense* IdiomGrp MetaphorGrp Metaphor* Lr caseframe* Idiom* Methods The KJCE DB is structured by the following process. First, we made a list of Korean entries. The entry of the KJCE DB is selected from the frequency data based on the ambiguity-solving corpus of the Institute of Korean Culture. Then, we made a list of Korean word senses, collocations, idioms, and metaphors. When we made the list, we refined the sources of the Sejong Electronic Dictionary from the 21st century Sejong Program which is promoted by the Korean Ministry of Culture and Tourism. Then, we made each sense, collocation, and idiom connected to corresponding English, Chinese and Japanese. Each one was refined from Korean-English, Korean-Japanese, and Chinese-Korean electronic dictionary. Finally, we use EWN s ILI in order to connect Korean with other languages, mainly European. We got vast information, such as the part of speech, synonym, hypernym, hyponym, domain, collocation, idiom, case-frame, etc. from the Sejong Electronic Dictionary. Its structure and form enables the user to use it along with the computer, so the information was automatically extracted from here and the rest of the data was supplemented by using a paper dictionary. We are attempting to connect the European languages with the Korean, Chinese and Japanese languages by using the EuroWordNet ILI. The ILI itself is an unstructured index so it has a structure possible to preserve and make continuous 122

136 WORDS IN ASIAN CULTURAL CONTEXTS modifications. ILI contains proper Offset number, Gloss, and Literal (which shows English vocabulary). Since it is not possible to get automatically the offset number from ILI, we made an input tool which makes it possible to get it semi-automatically. This tool suggests the alternatives of corresponding English sense to Korean sense so that we could choose the proper sense. When we chose the proper ILI, the offset number and gloss of the ILI is to be inserted into KJCE DB automatically. The next picture shows the input tool. Figure 2. The input tool of the KJCE DB In order to handle 50,000 vocabulary entries in this task, the Multi-Lingual input devise was developed to make information efficiently. Results Until now, we have constructed 20,000 noun entries and 500 verb entries. Among those entries, we selected 4,301 noun entries (5,409 senses) by the order of frequency to check the Multi-Lingual Lexical correspondence for this paper. Firstly, consider the total correspondence of Korean, English, Japanese, and ILI. Table 2. Multi-Lingual Correspondence in the KJCE DB type frequency ratio(%) type frequency ratio(%) KJCE-ILI KC-ILI 2 0 KJCE KJ-ILI 0 0 KCE-ILI KE-ILI 1 0 KJE-ILI KC 0 0 KJC-ILI KJ 0 0 KCE 1 0 KE 0 0 KJE 2 0 K-ILI 0 0 KJC K

137 ASIALEX 2005 According to the table above, the ratio of Korean, English, Chinese, Japanese and ILI are in correspondence with the total of 5,409 Korean senses is the highest by 86.2%. The reason why all languages are in correspondence with Korean is that the noun lexical items are high-frequency words which are included in the highest 5,000 ranks in frequency. It is 12.5% that Korean, English, Chinese and Japanese are in correspondence, but a corresponding concept is not found in ILI. A part of corresponding examples is provided. Table 3. Sense Correspondence in the KJCE DB Korean Japanese Chinese English 논지 #1_1 論旨ろんし论旨 point [gist] of an argument 생선 #1_1 鮮魚せんぎょ 鲜鱼 raw [fresh] fish 마님 #1_1 夫人ふじん 夫人 ; 太太 madam; My Lady 막걸리 #1_1 マッコルリ ; 独酒どぶざけ 马格利酒 ; 米酒 unstrained [raw, crude] rice wine; makkolli 노래방 #1_1 カラオケ 练歌房 Noraebang, the Korean commercial singing establishment The following analysis is possible about the factors of such examples. The first factor is that English is in correspondence by phrase and proper literal is not found in ILI while Chinese and Japanese words are in correspondence with Korean words by word. For example, 논지 (nonji) is in correspondence with 논지 (nonji) - 论旨 (lùn zhǐ) - 論旨ろんし (ronji) in Korean, Chinese and Japanese words by word. However, it is in correspondence with the phrase point [gist] of an argument in English. ILI includes several concepts that are realized in phrases, but concepts are mostly realized in words. Therefore, proper lexical items are not found. There are relatively many word-to-phrase correspondence between Korean-English: a Korean word 생선 (saengseon) corresponds to the Chinese word 鲜鱼 (xiān yú), or the Japanese word 鮮魚せんぎょ (sengyo), but, in English it corresponds to a phrase as raw fish. The second factor is cultural importance or lexical importance. We found no concept in ILI which corresponds to 생선 (saengseon), or 마님 (manim). It can be concluded that the Korean sense is not an important concept in the ILI of EuroWordNet. In the case of saengseon, it is supposed that since East Asian people eat sliced raw fish or dried fish, they need a concept which express the freshness of fish like saengseon, xiān yú, sengyo, while this is not an important concept in the ILI of EuroWordNet. Such analysis is also applied to 막걸리 (makkeolli). The lexical item exists in the Eastern Asian culture in which rice is a principal food source. In Korean, Chinese and Japanese, the lexical item is realized in a word like 막걸리 (makkeolli)- 米酒 (mǐ jiǔ) - 独酒どぶざけ (dobujake) while it is realized in a proper noun like makkolli or in a phrase like unstrained [raw, crude] rice wine in the western culture and it is not regarded as a major concept in ILI. Similar analysis is possibly applied to 노래방 (noraebang). This concept is related to an entertainment culture while is originated from Japan, swept Korea, and 124

138 WORDS IN ASIAN CULTURAL CONTEXTS introduced to China. On the other hand, such culture is not yet settled in Western Europe. This reflects the cultural difference between Western Europe and Eastern Asia. Then, how are such concepts which are regarded to be important received in the Multi-Lingual Lexical Database? Two methods are considered even though concrete managemen t can differ depending on individual senses. The f irst method is to bind the group of synonyms of individual languages with one of quasi-equivalent relations (quasi synonym equivalent relation, hypernym equivalent relation, metonym equivalent relation, variation equivalent relation, generalization equivalent relation, etc.) of EuroWordNet. For example, 막걸리 (makkeolli), for which a corresponding concept is not found in ILI is corresponded with the alcohol using hypernym equivalent in relation. But this method includes a fault, incapable of fully reflecting the cultural differences of individual languages. The second method is to register a new separate concept which is not found in an existing ILI-Record. This method is already used in EuroWordNet which is made based on WordNet. In this case, it is effective to put a separate identification mark for the differentiation of the new record from the existing one. For instance, ILI-Record is newly given to 노래방 (noraebang) as the following. 1 WORDNET_OFFSET : _k 1 GLOSS : the commercial singing establishments & ndOrderEntity 1 PART_OF_SPEECH : n 1 VARIANTS 2 LITERAL : [Noraebang] It means that Noraebang is registered as a new concept with the meaning of the commercial singing establishments in the identification number, _k. Likewise, after the addition of ILI record, a group of synonyms of each language is related with a new concept by the equivalent relation 8,032 collocations are used to inspect the Multi-Lingual correspondent relation. The collocation is important in simplifyin g algorithm for machine translation. The case of Korean, English, Chinese and Japanese that are in corresp ondence, forms 85.5% of correspondence. The case that Korean, Chinese, and Japanese are in correspondence forms 10%. The case that Korean, English and Chinese a re in correspondence is 1.5%. It is remarkable that the probability that Korean, Chines e and Japanese are in correspondence is higher than the probability that other three languages are in correspondence. Conclusion Since the WordNet and the EuroWordNet were developed, a concern about Multi-lingual Concept Network has been increased in Korea. Constructing a Korean Concept Network is a work to be done sometime even though it will be a vast effort. The hierarchical relation of the KJCE DB is a fundamental work for it. If a completed WordNet of Korean, Chinese and Japanese is connected with the Multi-Lingual Lexical Database of this project - the database is designed considering this connection,- Multi-Lingual meaning search on web will be supported. 125

139 ASIALEX 2005 Once a construction of 50,000 terms and related collocations are completed, the KJCE DB might sim ultaneously show common concept and concept gap in four languages sense by sense. Moreover, Korean Concept Network can be realized to be shown by the Periscope of EuroWordNet through ILI information. Accordingly, important language information for foreig n language learners is p rovided by indicating coll ocations, conceptual hierarchical relation between terms, and transition or variation of meaning in Multi-Lingual mapping. References Beom-mo, Kang (1999), The structure and Representation of Lexical Semantic Information, Korean Semantics, 5: Fellbaum, Christiane (1998), WordNet. An Electronic Lexical Database (Cambridge, Ma: The MIT Press). Beom-mo Kang, Yu-sun Lee, Jae-eun Cha (2002), Research and Model Development for a Multi-Lingual Lexicai Database(1) (Seoul: Insititute of Korean Culture of Korea University). Peters, W., P. Vossen, P. Diez-Orzas and G. Adriaens (1998), Cross-linguistic Alignment of Wordnets with an Inter_Lingual_Index, Computers and the Humanities 32(2-3): Improving English-Japanese dictionaries with particular reference to 1 systems of labelling Takashi Kanazashi Nihon University, Japan kanazashi@bus.nihon-u.ac.jp Abstract User-friendliness does not seem to be fully achieved in modern English-Japanese dictionaries for learners. This is partly because dictionaries tend to widen the range of their intended users. Suggestions are made as to the necessity of notes for phonetically difficult words and ways of improving the systems of labelling (grammar labels and subject field labels). With the use of a model entry from a fictitious dictionary, which restricts its intended users to Japanese students who are preparing for the entrance examinations to Japanese universities, I will show how the high restriction of intended users might enhance the reliability and the user-friendliness of the information provided in a dictionary. A fictitious dictionary for the entrance examinations to universities There is no doubt that user-friendliness is one of the top priorities in dictionary-making. More and more lexicographers have discussed the user perspective recently: Diab (1990), Atkins (ed) (1998), Hartmann (2001), Tono (2001) to name a few. Nakamoto (1998: 35) suggests the possibility of analyzing dictionaries from the viewpoint of the dictionary users, and encouraged by his suggestion, Dohi et al. (2002) and Ichikawa et al. (forthcoming) allot one section of their dictionary reviews to user research. In this article, I will deal with an imaginary entry, triple, from a fictitious model 126

140 WORDS IN ASIAN CULTURAL CONTEXTS English-Japanese dictionary, EXIONARY, by comparing it with the corresponding entry in existing English-Japanese dictionaries, GENIUS, GLOBAL, and C-LIGHTHOUSE. First, EXIONARY is aimed at Japanese learners of English, and can be consulted by them for both active and passive use. In this respect, it is much the same as three other dictionaries, some of whose intended users are high school students. What is really unique to EXIONARY is that it restricts its intended users to Japanese students who are preparing for the entrance examinations to Japanese universities (henceforth UEEs). Dictionaries tend to widen the range of their intended users, principally for commercial reasons. But this sometimes leads to user-antagonism, because their needs differ according to their proficiency in the target language. One might still wonder why there should be a dictionary specifically designed for the UEEs. In order to understand this, one has to know how fiercely competitive the examinations are, how biased English teaching in Japan is, and how awkwardly the students proficiency in English is assessed in the UEEs. Many Japanese children start private preparatory schools years before they take UEEs, because there is still an overemphasis on educational qualifications, if not as much as before. With so many applicants for admission to universities, written questions in true-false or multiple-choice format are preferred to reduce the time of testing and marking. This applies to all subjects, and English is no exception. Besides, ever since the Japanese started studying classical Chinese many centuries before the first Europeans drifted to Japan, the deep-rooted tradition of the grammar-translation method of teaching foreign languages has made many Japanese feel insecure if deprived of translation equivalents. A student could well be discouraged from practicing spoken English because there are few universities that require their applicants to take a listening comprehension test, much less an oral examination. This is indeed a sad situation, but dictionaries cannot change it. Although C-LIGHTHOUSE and LUMINOUS provide many sample dialogues, these do not make much difference because they are written-to-be-spoken. All that the students preparing for the UEEs can expect of a dictionary is that it will be a helpful tool for them. While many vocabulary books, textbooks and grammar books have been published in Japan to help high school students prepare for the UEEs, there is no dictionary for this very purpose as far as I know. It is against such a background that the need for EXIONARY might be evaluated. triple /tripl/ adj. [<hikaku nashi>] [<meishi no mae>] 1 <3 bubun kara naru> 2 <3 bai no> 3 [<on>] <3 byoushi no> n. [C] 1 <3 bai no kazu, ryou> 2 [yakyuu] <3 ruida> 2 Pronunciation In the above entry for the word triple in EXIONARY, immediately after the headword, its pronunciation is give between slashes. The phonetic symbols in this dictionary are largely based on the IPA, but some symbols are more simplified than those in existing dictionaries. For example, / / is substituted by /i/ here, which usually represents the long vowel as in tree. C-LIGHTHOUSE and LUMINOUS discriminate between them, but GENIUS and N-GLOBAL do not. Presumably it is difficult to explain to the users at this level the difference in their sound value using just these two slightly different symbols. If this discrimination is not sufficiently taught in high schools, nor is it demanded in UEEs, there is not much point in describing it in this dictionary. It is the policy of EXIONARY not to divert the users attention from what would indeed be useful 127

141 ASIALEX 2005 for their preparation for UEEs. There are three features regarding pronunciation in the form of notes which will prevent the users mispronunciations (or, in practice, prevent the users from memorizing incorrect pronunciation): (1) words that are pronounced in breach of the rules of English phonics (2) words that are likely to be pronounced by analogy with Japanese loanwords from English (3) words that are likely to be pronounced by analogy with Japanese loanwords from foreign languages other than English The first type of words, for example heart and indict, may be phonetically diff icult not only for the Japanese but also students of English in other countries, because the rules of English phonics tell them to pronounce the vowel in heart as in hear or heard, and to pronounce the second syllable of indict just like the second syllable of predict. The second and the third types of words may well be referred to as phonetic false friends, whose mispronunciations can be attributed to analogies with Japanese loanwords. When the Japanese borrow a foreign word, they naturally try to approximate their pronunciation of the word to that in the donor language, but sometimes the result is heavily Japanized. Pageant, patron, and wool are examples of type (2): Japanese students of English tend to pronounce the stressed vowels of pageant and patron as in patron and pageant, respectively, while the vowel in wool is likely to be pronounced with a long /u:/, all by analogy with loanwords. Examples of type (3) are theme and virus. As Japanese loanwords, these are pronounced more like German Thema and Latin virus, respectively, than English because of the direct derivation of these words. It is desirable to provide notes for all these words in addition to phonetic symbols. 3 Grammar labels Following the indication of part of speech, two grammar labels are provided: no comparison and usually before noun. Let us contrast the former with those used in existing English-Japanese dictionaries. Table 1: the labels for adjectives with no comparison dictionaries grammar label literal explanation GENIUS2, GENIUS3 φ<hikaku> phi comparison N-GLOBAL1, N-GLOBAL2 c with a slash through it C-LIGHTHOUSE, LUMINOUS <hikaku nashi> no comparison EXIONARY <hikaku nashi> no comparison Both GENIUS and N-GLOBAL include a letter that is difficult to read and understand, which represents a decrease in transparency. This problem has not been eliminated in their revised editions. How could an average Japanese high school student (or even an adult) be expected to read a Greek letter and understand it? As regards the letter in N-GLOBAL, users are required either to read the user s guide carefully, or to infer from the fact that it is a label for an adjective that c is the initial of comparison and that the slash means its opposite. One could even go so far as to say that these dictionaries seem to aim at pedantry rather than user-friendliness. In contrast, EXIONARY, in keeping with C-LIGHTHOUSE, provides a note written in plain Japanese, thus saving the effort of turning to the user s guide. 128

142 WORDS IN ASIAN CULTURAL CONTEXTS Table 2: labels for adjectives that function attributively grammar label literal explanation GENIUS2, GENIUS3 <gentei> attributive N-GLOBAL, N-GLOBAL2 <gentei> attributive C-LIGHTHOUSE, LUMINOUS A A EXIONARY <meishi no mae> before noun The grammatical term <gentei> (attributive) will cause a problem because not many high school students are familiar with it (and its antonym predicative). 4 The label A in C-LIGHTHOUSE is even more problematic, because A is also an abbreviation for adjective in the description of sentence patterns in the same dictionary. Hence the label <meishi no mae> (before noun) in EXIONARY, a label that is also used in LDOCE4. For predicative adjectives, LDOCE4 gives the label [not before noun] and CALD [after v] (adjective that only follows a verb). But a direct translation of these labels would cause some inaccuracy. Instead, EXIONARY uses <hogo to shite> (as complement). Complement is a term that is more familiar to Japanese high school students than predicative, mainly because the five-sentence-pattern system, according to which all English sentences are classified into five patterns, with S(ubject), V(erb), C(omplement), and O(bject) as their compulsory components, is still widespread among high school textbooks, grammar books, and even dictionaries such as N-GLOBAL2. Subject field labels There are two kinds of subject labels in the entry for triple in EXIONARY: <on> (music) and <yakyuu> (baseball). Most subject labels in this dictionary are composed of one or two kanji, or Chinese characters, as in existing English-Japanese dictionaries. There are two main reasons for this. First, each kanji is an ideogram with its own meaning. Second, many Japanese technical terms, including those denoting subject fields, are easy to understand if written in kanji, because each of them contains at least one kanji that characteristically provides the meaning. For example, the first character <gyo> of <gyoruigaku> (ichthyology) means fish. Such terms markedly contrast with English technical terms typically derived from classical Greek or Latin (cf. Suzuki 1990: ). A Japanese might wonder why no native speaker of English has invented such words as * fishology (for ichthyology) and *birdology (for ornithology) for the sake of communicative effectiveness. The kanji <on> means sound and tone, which are closely connected with music. However, for musical terms N-GLOBAL, C-LIGHTHOUSE and LUMINOUS use <gaku>, which is the other component of the compound character <ongaku> (music) and which has only a relatively weak association with it. Presumably these dictionaries do so in order to avoid confusion between music and phonetics, whose Japanese equivalents share the same initial kanji. Since EXIONARY is aimed at Japanese high school students, many of whom are not familiar with the term <onseigaku> (phonetics), only giving <on> would be sufficient. This illustrates that the transparency may depend on the intended users. <Ongaku> (music) can be abbreviated to its initial, but <yakyuu> (baseball) should not, because none of its components has a strong association with the term as a whole. Thus, lexicographers should consider the transparency of abbreviated terms. This is even more important when kana, or Japanese syllabograms, are used as abbreviations, because they will not be more transparent than abbreviations based on the Roman 129

143 ASIALEX 2005 alphabet. An example commonly found in GENIUS, N-GLOBAL, C-LIGHTHOUSE, and LUMINOUS is the use of <ro> as an abbreviation for <rooma> (Rome), although users might frequently misinterpret it as an abbreviation for <roshia> (Russia). 5 User-friendliness could be enhanced by eliminating this kind of opaque labelling. Conclusion I have described three features of EXIONARY. Because of its restriction to a particular type of intended users, it has some unique characters. Best of all, it pays considerable attention to transparency in the coding of grammar and subject field labels, which alone could be evidence of the user-friendliness of this dictionary. It is hoped that a dictionary would be a useful tool for its specially intended users in close harmony with textbooks, grammar books, and other dictionaries. Notes 1 I should like to thank Dr R. R. K. Hartmann for carefully reading an earlier version of this article and giving me valuable comments. Any shortcomings that remain are, however, my responsibility. The earlier version was reported orally at the Graduate Conference in English Studies at the University of Exeter in May, As all the dictionaries evaluated have been revised since then, I decided to revise the article, only to find that the problems concerning the labels have not been eliminated from the new editions. 2 Under this model entry and elsewhere, labels of any kind are placed between [square brackets], while transcriptions from Japanese characters are placed between <angle brackets>. 3 GENIUS gives a note on all these seven phonetically difficult words, but C-LIGHTHOUSE and LUMINOUS only gives a note on heart, and N-GLOBAL does so on none of them. 4 The results of my tests and interviews with Japanese informants have convinced me that few university and technical school students know what attributive and predicative mean, even if they are translated into Japanese. See Kanazashi (2001: 74-75) and Dohi et al. (2002: 74-75). 5 Some of my informants misinterpreted <roshin> as Russian myth instead of Roman myth, although none misinterpreted <gishin> (Greek myth). References (a) dictionaries CALD (2003): Cambridge Advanced Learner s Dictionary. Woodford, K. ed. Cambridge: Cambridge University Press. C-LIGHTHOUSE (1995): Kenkyusha College Lighthouse English-Japanese Dictionary. Takebayashi, S. et al. ed. Tokyo: Kenkyusha. GENIUS2 (1994): Taishukan s Genius English-Japanese Dictionary, 2nd edn Konishi, T. ed. Tokyo: Taishukan. GENIUS3 (2001): Taishukan s Genius English-Japanese Dictionary, 3rd edn. Konishi, T., Minamide K. ed. Tokyo: Taishukan. LDOCE4 (2003): Longman Dictionary of Contemporary English, 4th edn. Bullon, S. ed. Harlow: Pearson Education. LUMINOUS (2001): Kenkyusha Luminous English-Japanese Dictionary. Takebayashi, S. et al. ed. Tokyo: Kenkyusha. N-GLOBAL1 (1994): The New Global English-Japanese Dictionary. Kihara, K. ed. Tokyo: Sanseido. N-GLOBAL2 (2001): The New Global English-Japanese Dictionary, 2nd edn Kihara, K. ed. Tokyo: Sanseido. (b) other references Atkins, B. T. S. ed. (1998), Using dictionaries: studies of dictionary use by language learners and translators (Lexicographica Series Maior 88) (Tübingen: Niemeyer). 130

144 WORDS IN ASIAN CULTURAL CONTEXTS Diab, T. (1990), Pedagogical lexicography: a case study of Arab nurses as dictionary users (Lexicographica Series Maior 31) (Tübingen: Max Niemeyer). Dohi, K. et al. (2002), An analysis of Longman Advanced American Dictionary, Lexicon 32: Hartmann, R. R. K. (2001), Teaching and researching lexicography (Harlow: Pearson Education). Ichikawa, Y. et al. (forthcoming), An analysis of Longman Dictionary of Contemporary English, 4th edition, Lexicon 35. Kanazashi, T. (2001), User s guides in English-Japanese dictionaries for learners, with particular reference to grammatical information, unpublished PhD thesis (Exeter: University of Exeter). Nakamoto, K. (1998), An analysis of ILC s Dictionary Analyses, Lexicon 28: Suzuki, T. (1990), Nihongo to gaikokugo [Japanese and foreign languages] (Tokyo: Iwanami Shoten). Tono, Y. (2001), Research on dictionary use in the context of foreign language learning: focus on reading comprehension (Lexicographica Series Maior 106) (Tübingen: Niemeyer). Effective ways of using native language in English classrooms: suggestions for non-native teachers Madoka Kawano Hoshi University, Tokyo, Japan kawano@hoshi.ac.jp Abstract Teachers who are non-native English speakers have their strengths and weaknesses. In Japan, where most teachers of English are Japanese nationals, Ministry of Education, Culture, Sports, Science and Technology launched a plan to improve the qualifications of English instructors. Improving their language proficiency is a very important mission when the teachers attempt to sustain classroom interactions in English. However, at the same time, it has been pointed out that Japanese (L1) can help students learn English (L2) efficiently. Depending on the level of the students, use of L1 facilitates language learning and cognitive development of the learners. In this paper, some ways to adopt L1 in L2 classes are presented and demonstrated. First, at the beginner s English classes, instruction in Japanese (L1) has an advantage because it will help students expand vocabulary and learn English syntax. The students will learn how to express their thoughts in English words by juxtaposing two languages. Second, DVDs can contribute to language classrooms by bilingual/trilingual subtitles and sounds. They can be authentic resources of various listening and reading activities. Third, activities in which students compare and contrast an English newspaper article with a Japanese article on the same topic are demonstrated. Since English mass media and Japanese counterpart have different perspectives, comparing them is beneficial not only in improving the target language, but also in raising cultural awareness and in promoting critical thinking. Introduction Non-native English speaking teachers in Japan have strengths and weaknesses, which need to be analyzed and discussed. Most of the English teachers in Japan are 131

145 ASIALEX 2005 Japanese nationals, whose L1 is Japanese. Consequently, Japanese is used to a greater extent in most English classrooms. Although English classes conducted only in English are considered preferable, there are several situations in which Japanese or L1 can be used effectively and efficiently. Depending on the level of the students, use of L1 can save class time, help motivate students in becoming bilingual, and heighten a student s awareness of different languages. The Japanese Ministry of Education, Culture, Sports, Science and Technology launched an initiative to cultivate and enhance English abilities in Japanese students ( The ministry proposed plans to improve English classes by improving the English teachers teaching abilities, upgrading the teaching system, and improving methods to motivate students to learn English more diligently. It is notable that the ministry also considered L1 development and improvement important in promoting L2 development. This initiative makes sense when one considers the Japanese proverb, Silence is golden, and eloquence is silver. In Japanese culture, speaking out in public (expressing personal opinions openly) is considered bad manners, and in Japanese schools, children are not taught otherwise! When children are not trained to express their opinions freely and logically in their native language, it is near impossible to expect Japanese children to speak up in a foreign language. With this in mind, children should be taught critical thinking strategies as well as methods of expressing themselves effectively in foreign languages. The relationship of L1 and L2 is explained in the theory of Common Underlying Proficiency, by Cummins (1980). There is a central cognitive system, which operates language systems in every language. In other words, if one develops certain cognitive functions in a particular language, those functions will work well in other languages, too. Pinker suggests that there is a universal language system, which is common in all languages (1994). Taking these theories into consideration, it can be said that enhancement of L1 leads to development of L2 and vise versa. In this paper, I would like to discuss the role of L1 in L2 classrooms, and suggest a few ways to incorporate L1 into classroom activities: (1) writing activities (2) DVD related activities, and (3) reading newspapers and/or journal article activities. Activities to adopt L1 in L2 classes writing In beginner level English classes, the use of Japanese (L1) is helpful in soliciting students to write in the target language. When students are required to write or to express themselves in the target language, such students often feel helpless, because they have limited vocabularies and knowledge of English syntax. In such cases, students should first be allowed to jot down their opinions and feelings in Japanese. This activity is meant to give students a sense of safety and freedom. In this way, students can more easily reorganize and formulate their own thoughts or at least, speculate on what they really want to say or write. Students can then attempt to express their main points in English before having to write an actual paragraph in English. I have noticed that without brainstorming in one s L1, beginners usually end up with shallow and superficial expressions of their ideas. Another approach for beginners and intermediate students is a protocol approach in which students write down their thoughts about the writing process in L1. For example, after students write a paragraph in English, they could write their comments, recalling their reflections about what they wrote, in the margins (on either 132

146 WORDS IN ASIAN CULTURAL CONTEXTS side of their English compositions). They could write such comments that come to mind like, I wanted to say such and such a thing in English or I wish I had known the English equivalent of a certain word in Japanese. In this, students can note various problems they had encountered in L2 writing. These problems can be discussed in peer groups and solutions can be found. Bilingual teachers could help students find solutions or at least, provide hints. Consequently, students would learn how to better express their thoughts in English by juxtaposing two languages. Activities to adopt L1 in L2 classes DVDs DVD content can contribute to language learning by reading bilingual/trilingual subtitles, as well as listening to the actual voices of the speakers (actors and actresses). Students in Japan have very limited exposure to target (foreign) languages in their daily lives; therefore, DVD content could offer authentic resources for various listening and reading activities. As a listening activity, students could first watch a DVD movie without subtitles, followed by a second viewing with subtitles. As a method for improving reading skills, students could watch Japanese movies with English subtitles. When my students watched, Sen to Chihiro no Kamikakushi (Spirited Away), a cartoon which won a prize at the Berlin Film Festival, they became absorbed in the English subtitles. Since the students are psychologically relaxed when watching movies in L1, they are able to concentrate more on reading the English subtitles. DVDs could be used in speaking and writing activities as well. Advanced students could attempt to make up unique subtitles, as if they were translators. Scenes from films could be used in role-playing activities. Students could summarize stories, or attempt to predict what will happen next. Advanced students could also write critiques about DVD movies they have seen. Activities to adopt L1 in L2 classes reading newspaper/journal articles Lastly, I would like to recommend an activity in which students could compare and contrast English newspaper/journal articles with Japanese articles on the same topic. Students could compare articles in English on such topics as the Olympic Games in Athens with similar articles found in Japanese newspapers, though the coverage of games and athletes may differ slightly between the two articles. In another related activit y, students could read both Japanese and English articles about a similar topic, such as a successor s problems in Japanese Royal family; one article in Japanese from the Asahi Shinbun (a Japanese newspaper), and the other from the English edition of the Japan Times. Again, students could compare and contrast the two articles note the different viewpoints. This activity has multiple educational values, in that students will pick up English vocabulary more easily, because they will be reading the same content in their L1 Japanese article. Hence, students will be more likely to be able to guess the meaning of unfamiliar words in the L2 article. Students will also be able to transfer knowledge, which was gained by way of L1 into L2 and vice versa. In other words, the use of two languages will reinforce the development of both languages. By comparing and contrasting perspectives in such articles, students will raise their cultural awareness. Lastly, but not least, a close analysis of written discourses will promote critical thinking. Conclusion 133

147 ASIALEX 2005 I have suggested several viable ways in which Japanese English teachers can incorporate native language (L1) in their English language (L2) curricula. My suggestions include various writing activities, the use DVD content, and the comparison and contrasting of newspaper articles in L1 and in L2. Japanese students study English as a means of communication and as an international language. In other words, the true goals of language instruction in Japan are for students to become bilingual and able to function effectively and efficiently in two languages, rather than to become monolingual English speakers. Therefore, the use of L1 should be included in L2 classes. Ideally, students should aim to expand their language proficiency in two languages through L2 classes. References Casey, J. (2004), A place for first language in the ESOL classroom, Essential Teacher, Cummins, J. (1980), The construct of language proficiency in bilingual education, in J.E.Alatis (ed.) Georgetown University Round Table on Languages and Linguistics (Georgetown University Press). Cummins, J. (1984), Bilingualism and special education: issues in assessment and pedagogy (Clevedon: Multilingual Matters). Ministry of Education, Culture, Sports, Science and Technology. (2003), Regarding the establishment of an action plan to cultivate Japanese with English abilities, (accessed: 29 December 2004) Pinker, S. (1994), The language instinct: how the mind creates language (William Morrow & Co.). Abstract Dictionaries and translation, from art to science? Ilan Kernerman ilan@kdictionaries.com Language is a means of communication that conveys culture and knowledge, thoughts, feelings and much more translated into words. Translating one language into another is, literally, a translation of a transla tion. Dictionari es trans late and define chunks of language, usually in the form of words and phrases. In the past they were usually compiled by one person, single-handedly, thus e xpressing that person s inner world and outlook, like an artist s creati on. But times are changing. Nowadays dictionaries are often developed by teams and based on corpora, aiming for real language and impartial science. Still, the personal factor remains vital, to select the input, interpret the output and determine the outcome. The internet may be an ultimate such corpus today and offer instant multilingual corpora tomorrow, will dictionaries eventually become incorporated in machine translation with no human t ouch? Linking dictionary and corpus 134

148 WORDS IN ASIAN CULTURAL CONTEXTS Abstract Adam Kilgarriff Lexical Computing Ltd A corpus is an arbitrary sample of language, whereas a dictionary aims to be a systematic account of the lexicon of a language. Children learn language through encountering arbitrary samples, a nd using them to build systematic representations. These banal observations suggest a relationship between corpus and dictionary in which the former is a provisional and dispensable resource used to develop the latter. In this paper we use th e idea to, first, review the Word Sense Disambiguation (WSD) research paradigm, and second, guide our current activ ity in the development of the Sketch Engine, a corpus query tool. We develop a model in which a database of mappings between collocations and meani ngs acts as an interface between corpus and dictionary. 1 Putting the dictionary in the corpus. Consider SEMCOR, a hugely successful project and resource, very widely used and stimulating large amounts of WSD work. It is clearly a dynamic and important model, only exceeded in its take-up and impact by the WordNet database itself. SEMCOR inserts dictionary sense labels into the corpus. It puts the dictionary into the corpus ; like our title, but the other way round. Call this the PDIC model, in contrast with one which puts the corpus into the dictionary, the PCID model. If one thinks of WSD as a task on the verge of hitting the marketplace and being widely used in applications, then PDIC is appropriate, as it represents the WSD task successfully done, and can be used as a model for what a system should do. However it is widely acknowledged that WSD is not in any such near-to-market situation (as 1 shown by discussions at the SENSEVAL-3 workshop ) and we stand by our deep reservations about the nature of the WSD task (Kilgarriff 1997a, 1997b) which imply this is unlikely to change. An alternative model, closer to the observations of the opening paragraph, is that the larger task is at a further remove from applications and is better seen as lexical acquisition. We are not yet at a stage (and probably never will be) at which a general purpose WSD module is a relevant goal, but there are many language interpretation tasks which cannot be done without richer lexical representations. In the PCID model, the corpus serves to enrich the lexicon. 1.1 Levels of abstraction The direct approach to corpus-dictionary linkage is to put pointers to dictionary senses into the corpus (in the PDIC model, as in SEMCOR) or (in the PCID model) to put pointers to corpus instances of words into the dictionary. The direct approach has a number of drawbacks. The primary practical one is fragility. If the corpus (PCID model) or the dictionary (PDIC model) is edited or changed in any way, maintenance of the links is a headache. (This has been an ongoing issue for SEMCOR, as new versions of WordNet call for re-building SEMCOR in ways which cannot in general be 1 Notes available at 135

149 ASIALEX 2005 fully and accurately automated; see Daudé et al (2000, 2001)). The theoretical one concerns levels of abstraction. A dictionary is an abstract representation of the language, in which we express differences of meaning but are not engaged with specifics of differences of form. The corpus is at the other end of the scale: the differences of form are immediately present but differences of meaning can only be inferred. What is needed is an intermediate level which links both to the meaning-differences in the dictionary and to the specific instances in the corpus. Our candidate for this role is the grammatical-relation triple, comprising <grammatical-relation, word1, word2> (examples are <object, drink, beer> and <modifier, giant, friendly>). Triples such as these 1 have, of late, been very widely used in NLP, as focal objects for parsing and parse evaluation (eg Briscoe and Carroll 1998, Lin 1998), thesaurus-building (eg Grefenstette 1992) and for building multilingual resources from comparable corpora (eg Lű and Zhou 2004). Approaching from the other end, they are increasingly seen as core facts to be represented in dictionaries by lexicographers, who usually call them just collocations (COBUILD 1987, Oxford Collocations Dictionary 2003). In our own work, we compile sets of all the salient collocations for a word into word sketches, which then serve as a way of representing corpus evidence to the lexicographer (Kilgarriff and Tugwell 2001, Kilgarriff and Rundell 2002, Kilgarriff et al 2004). 1.2 Few would dispute that collocations which incorporate grammatical information (and are thereby triples like <object, drink, beer> ) are a more satisfactory form of lexical object than raw collocates those words occurring within a window of 3 or 5 words to the right of, or to the left of, or on either side of the node word. Windowing approaches operate as a proxy for grammatical information and are appropriate only where there is no parser available, or it is too slow, or too often wrong. Historically, these factors have often applied and most older work uses windowing rather than grammar. As we are able to work with grammar, we do. We are repeatedly struck by how much cleaner results we get. We also find it preferable to work with lemmas rather than raw word forms, where a lemmatiser is available for a language. 1.3 Grammatical relation triples being unwieldy, I shall call these objects simply collocations, or say that the one word is the other s collocate. Strictly, the items in the triples are lemmas which include a word class label (noun, verb, adj etc) as well as the base form; in examples, the word class labels will be omitted for readability. Naturally, some grammatical relations are duals (object, object-of) or symmetrical (and/or); for a full treatment see Kilgarriff and Tugwell (2001). 2 Aside: Parsing and lemmatizing Terminology The collocation database In the proposed model, between the dictionary and the corpus sits a database. For each dictionary headword, there is a set of records in this database comprising 1) a collocate (including grammatical relation) 2) a pointer to the dictionary sense 1 Naturally, details vary between authors. Briscoe and Carroll do not in fact use triples, but tuples with further slots for particles and further grammatical specification. 136

150 WORDS IN ASIAN CULTURAL CONTEXTS 3) a set of pointers to corpus instances of the headword in this collocation The database is, in the first instance, generated from the corpus, so the corpus links are immediately available. To start with, the dictionary pointers are not filled in for polysemous words. (For monosemous words, the links can immediately be inserted.) A word sketch (see Fig 1) is an example of such a database. The corpus links are present, implemented as hyperlinked URLs: for on-line readers, clicking on a number opens up a concordance window for the collocation (go to to self-register for an account). 2.1 Limitations and potential extensions The word sketch model is dependent on Yarowsky s One sense per collocation (Yarowsky 1993). To the extent that this does not hold, the model will be inadequate and we will need to make the structure of the database record richer. The triples formalism does not readily express some kinds of information which are known to be relevant to WSD. An intermediate database to link dictionary to corpus should have a place for all relevant facts. Two kinds of fact which do not obviously fit the triples model are grammatical constructions, and domain preferences. Many, possibly all, grammatical constructions can be viewed as grammatical relations (with the other word field null). Thus a verb like found, when in the passive, means set up ( the company was founded in 1787 ). In this case we associate the triple <passive, found, _> with the set up meaning. We have already implemented a range of unary relations within the Sketch Engine, and believe this approach will support the description of all grammatical constructions. As much recent work makes clear, domains are central to sense identification (eg Agirre et al 2001, Buitelaar and Sacaleanu 2001, McCarthy et al 2004). However it is far from clear how domain information should be expressed: hand-developed inventories of domains have many shortcomings, but data-driven approaches to domain induction are not yet mature and suffer from the arbitrariness of the corpora they use. The incorporation of domain information into the database model remains further work. goal bnc freq = change options and/or objective try goal penalty target value conversion aim mission priority strategy object_of score achieve concede disallow pursue attain net kick grab reach set subject_of come give win help adj_subject_of important a_modifier ultimate away winning compact stated late dropped organisational long-term common headed

151 ASIALEX 2005 point notch organizational n_modifier drop penalty league consolation opening second-half first-half minute half policy relationship development modifies scorer difference scoring ace drought post kick keeper weight lead average setting pp_after-p minute particle back down up possessor England pp_from-p attempt Fig 1. Word sketch for goal (reduced to fit in article) Whereas a collocate tends to be associated with one and only one sense, so can be used to generate a Boolean rule of the form collocation X implies sense Y, both grammatical constructions and domains provide preferences. Royalty (singular) usually means kings and queens, whereas royalties (plural) usually means payments to authors. However a rule singular implies kings-and-queens should not be Boolean: we often talk about, eg, royalty payments which are payments to authors, not to (or from ) kings and queens. The facts are preferences, or probabilistic, rather than categorical. Our current model does not incorporate preferences or probabilities, and they raise theoretical problems: are the probabilities not as arbitrary as the corpora they were drawn from? This, again, is further work. The formalism will allow Boolean combinations of triples and of senses, so it is possible to say, eg, triple X AND triple Y imply NOT sense Z. We envisage that unary relations (eg, grammatical constructions) will often be used to rule out senses, or in conjunction with collocates. Once solutions to the domains issue are found, we will be able to view the database connecting corpus to dictionary as a database of collocations, constructions and domains: a CoCoDo database. 2.2 Linking collocations to senses There are a number of ways in which the pointers to dictionary senses m ight be added. Over the last forty years the WSD community has developed a host of strategies for assigning collocates to dictionary senses (Ide and Véronis 1998, Kilgarriff and Palmer 2000, SENSEVAL , SENSEVAL ). Many of them can be applied (depending, obviously, on the nature of the dictionary and the information it provides). We have specified the problem as the disambiguation of collocates rather than corpus instances. In practice, collocates (more widely or narrowly construed) are the 138

152 WORDS IN ASIAN CULTURAL CONTEXTS workhorse of almost all WSD. The core is of identifying a large set of collocates (or, more broadly, sentence patterns or text features) which are associated with just one of the word s senses, which then may be found in a sentence or text to be disambiguated. The task of assigning collocates is a large part of the task of assigning instances. Other differences between the task as specified here and the traditional WSD task are as follows. 1) Dictionary structure: We can link to any substructure of the dictionary entry; if the entry has subsenses, or multiwords embedded within senses or vice versa, we can link to the appropriate element, so need not make invidious choices about whether to use fine-grained or coarse-grained senses. 2) Other dictionary information: Since the larger goal is enrichment of lexical resources, where a resource is already rich, the information it contains is given. It can be used in WSD, and does not need to be duplicated. One resource we have looked closely at, the database version of Oxford Dictionary of English (McCracken 2003), contains particula rly full information on domain, taxonomy, multiwords, grammatical and phonologic al patterning etc., all sense-specific. This is all immediately av ailable, both for disambiguation and, obviously, in the output resource. 3) Precision-recall tradeoff: There is no commitment to disambiguating all corpus instances (or all collocates). Like many NLP tasks, WSD exhibits a precision-recall tradeoff. If a system need not commit itself when the evidence is not clear, it can achieve high accuracy for those cases where it does present a verdict. WSD has usually been conceptualised as a task where a choice must be made for every instance (so precision=recall) and in the PDIC model this seems appropriate. But in the PCID model it is not necessary. What we would like is some corpus-based information about all dictionary senses, and it is immaterial if there are some corpus instances which do not contribute to any lexical entry. Once we view the WSD task in this light, we welcome high-precision, low-recall strategies (for example Magnini et al 2001, which achieved precision 5% higher than the next highest-precision system in the SENSEVAL-2 English all-words task, with 35% recall). We can do WSD without the shadow of an apparent 60% precision ceiling (SENSEVAL ) hanging over us. 4) Mixed-initiative methods: Once WSD is seen as a step towards the enrichment of lexical resources, it becomes valid to ask how humans may be involved in the process. Kilgarriff and Tugwell (2001), and Kilgarriff, Koeling, Tugwell and Evans (2003) present a system in which a lexicographer assigns collocates to senses, and this then feeds Yarowsky (1995) s decision-list learning algorithm. In general, in the proposed architecture, both people and systems can identify collocate-to-sense mappings, with each potentially learning from evidence provided by the other and correcting the other s errors or omissions. (There will be a set of issues around permissions: which agents (human or computer) can add or edit which mappings.) Ideally, the process of identifying the mappings for a word is a mixed-initiative dialogue in which the lexicographer refines their analysis of the word s space of meaning in tandem with the system refining, in real time, the WSD program which allocates instances to senses and thereby provides the lexicographer with evidence. 139

153 ASIALEX Dictionary-free methods While most WSD work to date has been based on a sense inventory from an existing resource, some (eg Schűtze 1998) has used unsupervised methods to arrive, bottom -up, at its own senses inventory. If the PCID model is being used to create a brand new dictionary, or if a fresh analysis of a word s meaning into senses is required, or if some dictionary-independent processing is required as a preliminary or complement to a dictionary-specific process, then dictionary-free methods are suitable. Methods such as Schűtze s are based on clustering instances of words. Our strategy will be to cluster collocates. One method we have already implemented uses the thesaurus we have created from the same parsed corpus as was used to create the word sketches. Looking at the verbs that goal is object of, in Figure 1, we see a number of verbs with closely related meanings, and we would ideally like to form them into two clusters, one for sporting goals and one for life goals (these being the two main meanings of goal). In the thesaurus entry for disallow, we find, within the top ten items, concede and net, thus providing evidence that these three items cluster together. Another method we shall be implementing shortly depends on the observation that a single instance of a word may exemplify more than one collocation. The instance score a drop goal exemplifies both <object, score, goal> and <modifier, goal, drop> so provides evidence that these two collocations should be mapped to the same sense. Collocate-clustering is best seen as a partial process, marking collocates as sharing the same sense only when there is strong evidence to do so and remaining silent elsewhere. It then provides good evidence for other processes, dictionary-based or manual, to build on. 3 The dispensable corpus As mentioned in the opening paragraph, a corpus is an arbitrary sample. A person s mental lexicon, while developed from a set of language samples, has learnt from them and moved on. 1 The corpus is dispensable. In a PDIC approach, this clearly does not apply: if the corpus is thrown away, all the evidence linking dictionary to corpus is lost too. Likewise for a PCID approach with direct corpus-dictionary linking. But in the model presented here, if the corpus is thrown away, the collocate-to-sense mappings are rich, free-standing lexical data in their own right (and could readily be used to find new corpus examples for each collocate or sense). 4 WordNet proposal The paper is largely programmatic. We have, as indicated above, starting exploring a number of the ideas, using the Sketch Engine ( platform and its predecessor, the WASPbench. We now want to develop it further, and are considering which dictionary (if any) to develop it with. (The Sketch Engine identifies all items collocations, triples, word instances- as URLs, thereby supporting distributed development, open access, and connectivity with other resources.) Dictionary-free development is attractive and under discussion, but, to develop a rich and accurate resource, a large investment will be required. It is unlikely the 1 The success of the memory-based learning paradigm, in both NLP and psycholinguistics, may be seen as casting doubt on this claim. 140

154 WORDS IN ASIAN CULTURAL CONTEXTS resulting resource would be in the public domain. Collaborations with dictionary publishers, to enrich their existing dictionaries, are under d iscussion. They too would not give rise to public-domain resources. For the development of the idea within the academic community, a public domain resource is wanted. The obvious candidate is WordNet. The proposal is then to develop a collocations database with links to WordNet senses, on the one hand, and colloca tes found statistically in a large corpus on the other. The WordNet links would be identified using the whole range of disambiguation strategies which have been develop ed for WordNet (including, potentially, the multilingual and web-based ones). We believe this could be a resource that takes forward our understanding of words and language and which supports a wide range of NLP applications. References Agirre E., Ansa O., Martínez D., Hovy E. Enriching WordNet concepts with topic signatures. In Proceedings of the SIGLEX workshop on WordNet and Other Lexical Resources: Applications, Extensions and Customizations. NAACL, Pittsburgh. Briscoe E. J. and J.Carroll Robust accurate statistical annotation of general text. In Proc. LREC Buitelaar P. and B. Sacaleanu Ranking and selecting synsets by domain relevance. In Proceedings of the SIGLEX workshop on WordNet and Other Lexical Resources: Applications, Extensions and Customizations, NAACL 200, Pittsburgh. COBUILD Collins COBUILD English Dictionary. Ed. J. Sinclair. Daudé J., Padró L. and Rigau G Mapping WordNets Using Structural Information. In Proc. ACL. Hong Kong. Daudé J., Padró L. and Rigau G A Complete WN1.5 to WN1.6 Mapping. In Proc. NAACL Workshop WordNet and Other Lexical Resources: Applications, Extensions and Customizations. Pittsburg, PA. Grefenstette, G Sextant: exploring unexplored contexts for semantic extraction from syntactic analysis. In Proc. ACL, Newark, Delaware: Ide, N. and J. Véronis, Editors Special issue on word sense disambiguation: The state of the art. Computational Linguistics, 24(1). Kilgarriff, A. 1997a. I don t believe in word senses. Computers and the Humanities 31: Kilgarriff, A. 1997b. What is Word Sense Disambiguation Good For?, In Proc. NLPRS: Phuket, Thailand. Kilgarriff, A., R. Koeling, D. Tugwell, R. Evans An evaluation of a lexicographer s workbench: Building lexicons for machine translation, Workshop on MT tools, European ACL, Budapest. Kilgarriff, A., P. Rychly, P. Smrz and D. Tugwell The Sketch Engine. In Proc. Euralex. Lorient, France, July: Kilgarriff A. and M. Rundell Lexical profiling software and its lexicographic applications - a case study, In Proc. EURALEX, Copenhagen, August: Kilgarriff A. and D. Tugwell WASP-Bench: an MT Lexicographers Workstation Supporting State-of-the-art Lexical Disambiguation. In Proc. MT Summit VIII, Santiago de Compostela, Spain: Kilgarriff, A. and M. Palmer Editors, Special Issue on SENSEVAL. Computers and the Humanities 34 (1-2). Lin, D A Dependency-based Method for Evaluating Broad-Coverage Parsers. Journal of Natural Language Engineering. Lü, Y. and Zhou, M Collocation Translation Acquisition Using Monolingual Corpora. In Proc. ACL 2004, Barcelona: McCarthy, D., Koeling, R., Weeds, J. and Carroll, J Finding predominant senses in untagged text. In Proceedings of the 42nd Annual Meeting of the Association for Computational Linguistics. Barclona, Spain. pp

155 ASIALEX 2005 McCracken J. and A. Kilgarriff Oxford Dictionary of English - current developments. In Proc. EACL. Magnini, B., Strapparava, C., Pezzulo, G. and Gliozzo, A Using Domain Information for Word Sense Disambiguation. In Proc. SENSEVAL-2: Oxford Collocations Dictionary for Students of English Ed. Lea. OUP. Schűtze, H Automatic Word Sense Discrimination in Ide and Véronis 1998, op cit. SENSEVAL-2 (2001) See SENSEVAL-3 (2004) See Yarowsky, D One sense per collocation. In Proceedings of the ARPA Human Language Technology Workshop, Morgan Kaufmann, pp Yarowsky, D Unsupervised Word Sense Disambiguation Rivaling Supervised Methods. In Proc. ACL: Adam Kilgarriff Lexical Computing Ltd adam@lexmasterclass.com Chinese word sketches Chu-Ren Huang Academia Sinica, Taiwan churen@gate.sinica.edu.tw Pavel Rychly Masaryk University, Czech Republic pary@textforge.cz Simon Smith Ming Chuan University, Taiwan ssmith@mcu.edu.tw Abstract Word sketc hes are one-page automatic, corpus-based summaries of a word s grammatical and collocational behaviour. They were first used in the production of the Macmilla n English Dictionary (2002). At that point, they only existed for English. Now, we have developed the Sketch Engine, a corpus tool which takes as input a corpus of any language and corresponding grammar patterns and which generates word sketches for the words of that language. It also automatically generates a thesaurus and sketch difference s, which specify similarities and differences between near-synonyms. We have recently prepared word sketches for Chinese, based on the Chinese Gigaword corpus. The corpus was tokenized and part-of-speech tagged in collaboration with Academia Sinica. We shall demonstrate the Chinese word sketches, and show how they can be used in lexi cography, to enrich and improve existing definitions, to quickly add further informatio n on collocates, and for selecting typical illustrative examples, as well as for more ambit ious projects involving electronic and web dictionaries and translation aids. We shall also describe the software and principles behind it, differentiating it from other corpus que ry systems such as Wordsmith and the Stuttgart tools. We shall outline the issues arisin g in the preparation of the Chinese data, and plans for exploiting bilingual word ske tches for Chinese and English. For more details see sketchengine.co.uk Corpus creation for lexicography Adam Kilgarriff Michael Rundell Elaine Uí Dhonnchadha 142

156 WORDS IN ASIAN CULTURAL CONTEXTS Lexicography MasterClass UK Lexicography MasterClass Institiúid Teangeolaíochta UK Éireann Abstract In a 12-month project we have developed a new, register-diverse, 55-million-word bilingual corpus the New Corpus for Ireland (NCI) to support the creation of a new English-to-Irish dic tionary. The paper describes the strategies we employed, and the solutions to problems enco untered. We believe we have a good model for corpus creation for lexicography, a nd others may find it useful as a blueprint. The corpus has two parts, one Irish, the other Hiberno-English (English as spoken in Ireland). We describe its design, collection and encoding. 1. Introduction In this paper we describe the development of the New Corpus for Ireland (NCI) a substantial lexicographic corpus in two-parts, one being Irish (the Celtic language of Ireland), the other Hiberno-English ( the variety of English that is spoken in Ireland). The three main sections describe its design, collection, and encoding. 1 The NCI was developed as part of the set-up phase of a project for a new English-to-Irish Dictionary (NEID).2, intended for use by scholars, school and university students, translators, people working in the media, and the general public. Irish is one of the two official languages of Ireland, the other being English. 62,000 speakers use Irish as their main everyday language, and almost 340,000 speakers use Irish on a daily basis 3. There are three main dialects: Connacht, Munster, and Ulster. The language has an important place in Irish culture and identity and is widely taught in schools 4. English, however, has been the dominant language in Ireland for well over 100 years. 2. Design In the first instance, a detailed corpus-design document was pre pared, and the target sizes for the two major components were a greed as 30 million words for Irish, and 25 million words for Hiberno-English. Target proportions were set for different text types. These were based, in the first instance, on the design principles developed for the BNC (see Atkins, Clear and Ostler 1992), but then modified in response to local factors. The factors that led us to adjust the BNC model included: 1 For a fuller account, see the journal paper: Kilgarriff Rundell and Uí Dhonnchadha (forthcoming). 2 The project is under the direction of Foras na Gaeilge, the government-funded body responsible for the promotion of the Irish language throughout the island of Ireland, whose statutory functions include the development of new dictionaries ( Full details of the NEID project can be found at The main contractor for setting up the project, including corpus preparation, is Lexicography MasterClass Ltd ( 3 Figures from the 2002 Irish Census. 4 Irish is taught throughout the school system, and about 30,000 students are educated in Irish-medium schools. 143

157 ASIALEX Collection the social and cultural salience, in Ireland, of certain genres and domains which had played a less central role in the BNC, for example reminiscences, rural folklore and the Catholic religion the impossibility, given time and budget constraints, of developing new spoken corpus data, in light of which it was decided that the only transcribed speech would be taken from already-existing spoken corpora the desirability of sourcing significant amounts of the Irish data from native-speakers of Irish (a minority of Irish users) the importance of including texts representing the three main dialects of Irish (Connacht, Munster, and Ulster) the plan, agreed at the outset, to include the new category of data from the web Three corpus collection strategies were used: incorporating existing corpora contacting publishers, authors, newspaper companies etc. to request permission to use their texts collecting data from the web. 3.1 Existing resources Irish was one of the languages of the EU PAROLE project, and as part of that project, an 8-million-word corpus of Irish had been developed at ITÉ (Institiúid Teangeolaíochta Éireann, the Linguistic Institute of Ireland). ITÉ had continued its data collection programme after the end of the PAROLE project and had several million further words of Irish text in its archive. This formed the core of the Irish corpus. For English, we learned that there were two corpora of transcribed Hiberno-English speech already in existence: the 1-million-word Limerick Corpus of Irish-English and the 400,000-word Northern Ireland Corpus of Transcribed Speech (NICTS) from Queen s University Belfast. Both were incorporated. 3.2 Contacting publishers, authors, newspaper companies Numerous potential text-donors were contacted, and were given a short document explaining (for a mainly non-corpus-aware audience) how donated text would actually be used in the dictionary-making process. They were asked to contribute to the project by sending electronic copies of texts and signing permissions letters which allowed the texts to be used as part of a lexicographic corpus. Our experience with the BNC and other corpora had prepared us for some of the difficulties inherent in this approach: the publishing business is based on the sale of copyright material, so it is not surprising that the default response from the publisher, when asked to give texts for free, is no. High levels of charm and persistence are required, and it was a large task. Having said that, the response was in the main very positive, with most copyright-owners pleased to be associated with the project. Once we had agreement-in-principle, we needed to actually acquire the text. Sometimes it was sent on CD or other media, sometimes it was received by . As expected, text arrived in a wide range of formats, including proprietary forms such as Quark, so the first step was to reduce everything to the same plain-text format. Further 144

158 WORDS IN ASIAN CULTURAL CONTEXTS steps are covered in section 4, below. 3.3 Collecting data from the web The web offers enormous possibilities for corpus development, for language of all varieties (Kilgarriff and Grefenstette 2003) and for smaller languages in particular (Jones and Ghani 1999). We worked with Infogistics Ltd., a company with expertise in computational linguistics and web crawling. They found Irish data on the web through (1) going to known-irish websites, and (2) entering a set of Irish-only words into Google and harvesting the pages Google found, and then checked the language of pages found with a high-accuracy language classifier. We believe they found a large proportion of the Irish that there is on the web. They found Hiberno-English data on the same websites where Irish data was found: if the language was English, it was likely to be English written by Irish people. They delivered the data in three iterations, and at each turn, we inspected it and reported back on any problems we encountered, which they addressed prior to the next iteration. Recurring issues in using web data included: multiple input formats: We collected documents in.txt, html, pdf, rtf, MS-Word, and postscript form. formatting: The corpus collector s default model is continuous uninterrupted text, but on the web, frames and pages are often used to split up a text, and text is often split across different, short web pages. Documents which are split must be either rebuilt to reconstruct the correct linguistic structure, or rejected. navigational material: text like click here next page further details is specific to web genres, and will distort the statistics if left in a lexicographic corpus. Common navigational phrases and constructions were identified and removed, for both Irish and English. lists: the web contains many lists: price lists, product lists, the players in a sports team, local councillors, and so on. While it is not always obvious where lists should be included in a corpus, for our (lexicographic) purposes, the rule of thumb was that we most wanted language when it occurred in sentences, and lists which displayed no sentence-like characteristics were rejected. mixed language: many web pages were part Irish, part English; policies were developed for where to delete material and where to retain part-pages. duplication: this is pervasive on the web, and the level of duplication in the Irish web was particularly high. It raises theoretical questions: what is the textual unit for identification of duplicates? If the unit is set too large, lots of duplicates will remain, but if the unit is set too small, then common sentences like How do you do? may be rejected as duplicates. The algorithm developed by Infogistics proved highly effective, and no unwanted duplication has been encountered The questions, what types of text are there on the web, and in what proportions? are large, hard, and under-researched (Kilgarriff and Grefenstette 2003). To give an idea of the range and variety of texts gathered for Irish in this project, we list in Table 1 a dozen websites from which we took substantial quantities of text, along with the types of document found in each. 145

159 ASIALEX 2005 Name Organization type Document types include: FUTA FATA Magazine Reviews of, and extracts from Irish novels, books of poetry Galway County Council County Council Policy statements, application forms University College Galway Universi ty Policy statements, statements of objectives, reports Department of Community, Government Department Speeches and press-releases from Rural and Gaeltacht Affairs the Minister, news reports Údarás na Gaeltachta Regional Development Announcements, forms, policy Agency statements, grant schemes Ógras Irish-language Youth Activities, competitions Organisation Sinn Féin Political party History, policy, events Gaelport/Comhdháil Umbrella Irish-language Electronic newsletter Náisiúnta na Gaeilge organisation Rondomondo Magazine Arts, music, drama Irish Army/Navy Armed forces missions, career descriptions Raidió na Gaeltachta Radio station Notices, news Aran Mór College College Advertising, programmes, activities Table 1: Sample of websites and text types for Irish web corpus collection 3.4 Corpus composition, compared with targets The table below shows the composition of the final corpus, compared with our original targets. Text category Irish Hiberno-English Words: Words: target Words: actual Words: target actual Books-imaginative 7,600,000 9,000,000 6,000,000 7,500,000 Books-informative 8,400,000 6,000,000 7,000,000 5,000,000 Newspapers 4,500,000 4,500,000 5,300,000 3,750,000 Periodicals 2,600,000 2,500, ,000 2,250,000 Official/Government 1,200,000 1,500,000 1,000,000 1,000,000 Broadcast 400,000 1,000, ,000 Websites 5,500,000 5,500,000 5,000,000 4,750,000 TOTALS 30,200,000 30,000,000 25,000,000 25,000,000 Table 2: New Corpus for Ireland: target figures and actuals For the most part, our a priori targets could be met. The biggest disparity is in the Books category, where, it transpired, imaginative texts were harder to find (for both languages) than originally anticipated. Almost half of the text in the Books category of the Irish corpus can be reliably attributed to Irish native-speaker authors and around 80% belongs to one of the three major dialects. 4. Encoding Once a set of documents has been collected, it must be prepared so that it is in an optimal state for use by linguists and lexicographers. We call this stage encoding. 146

160 WORDS IN ASIAN CULTURAL CONTEXTS A key issue here is delivery format. For longevity, and as an interchange format, it was clearly appropriate that the corpus be delivered in XML, and in a standard corpus-encoding formalism (in this case the XML Corpus Encoding Standard, XCES: see However, for the corpus to be usable, it also had to be loaded into a corpus-querying system (CQS). Any particular CQS will have encoding conventions more specific than those imposed by XCES. The tool adopted for this project was the Word Sketch Engine (Kilgarriff, Rychly, Smrz and Tugwell 2004; The project included the delivery of a version of the corpus loaded into the Word Sketch Engine, in a set-up in which the type of queries a lexicographer would regularly need to make could be made quickly and efficiently. 4.1 Linguistic processing Some of these queries involve grammar, and most involve lemmas (give (v)) rather than word forms (give, gives, giving, gave, given). To this end the corpus was to be lemmatized and part-of-speech tagged. While software for lemmatizing and part-of-speech tagging is widely available for English, the situation for Irish is less advanced. As part of the project, we developed tools for Irish. The lemmatizer uses Xerox tools (Beesley and Karttunen 2003) and builds on work described in Uí Dhonnchadha, Nic Pháidín and van Genabith (forthcoming). The part-of-speech tagger was developed from scratch, working with the output of the morphological analyzer and using the Constraint Grammar formalism and parser (Tapanainen 1996). 4.2 Document IDs and filenames Corpus development involves very large numbers of documents. It is easy for documents to get lost. In other corpus projects, we had witnessed all too much effort expended on looking for lost files, so it was a priority to set up a system which minimized the risk. Our strategy was to assign to each document, at the earliest possible stage in the process, a document identifier and a specification for where in the file system the document was stored: the structure of the file system would map directly onto the document identifier. We enforced a rigid one document per file, one file per document 1 convention. Identifiers needed to be: unique: different teams would be collecting different parts of the corpus, so it was essential to preclude the possibility of different teams assigning the same ID to different documents short and informative not subject to change at any later stage. 4.3 Text cleaning, paragraph markup Once a document had had its ID assigned and had been saved as raw text with matching filename in the appropriate place in the file system, we examined the text in an editor, deleting parts of the text which were not suitable for a lexicographic corpus. The unsuitable parts included: title pages, tables of content and other tables, figures and diagrams, footnotes and endnotes, indexes, page headers and footers including running titles, 1 Different versions of the same text could be stored with the same base filename, though with different extensions, eg..txt,.xml. 147

161 ASIALEX 2005 crosswords, TV listings, isolated names and addresses dates from advertisements, racing results, lists of team members etc. To our surprise, the cleaning removed an average of a third of the words in a text. The text was then lemmatized, part-of-speech tagged, assembled with its header information, and rendered in XML and loaded into the Word Sketch Engine for lexicographic use. 5 Conclusion The project gathered a high-quality corpus of substantial size from a wide range of sources, in just over one year with modest resources. The corpus was designed primarily to meet the lexicographic requirements of an English-to-Irish dictionary, but with an eye to the resource being used more widely, by scholars of Irish and Hiberno-English. We have shown how the encoding of the corpus feeds into lexicography. Lexicographers are best supported by a linguistically-aware corpus query tool, and that requires a linguistically annotated corpus. The tools are readily available for English, but were not, at the outset of the project, for Irish, so we developed and extended tools for the morphological analysis and part-of-speech tagging of Irish within the project: we would encourage others, when working with a language where tools are currently limited in scope or non-existent, to do likewise. We anticipate that many of the procedures outlined here could be applied in order to rapidly and inexpensively gather corpora for other smaller languages. References Atkins, B. T. S., Clear, J. H., and Ostler, N. (1992). Corpus design criteria. Journal of Literary and Linguistic Computing: Beesley K. and Karttunen L. (2003). Finite State Morphology. CSLI Publications: California. Jones, R. and R. Ghani. (2000). Automatically building a corpus for a minority language from the web. 38th Meeting of the ACL, Proceedings of the Student Research Workshop. Hong Kong. Pp K ilgarriff, A. and Grefenstette, G. (2003). Web as Corpus: Introduction to the Special Issue. Computational Linguistics 29 (3) Kilgarriff, A., Rundell, M. and Uí Dhonnchadha, E. (forthcoming). Efficient corpus development for lexicography: building the new corpus for Ireland. Languages Resources and Evaluation Journal. K ilgarriff, A., Rychly, P., Smrz, P., and Tugwell, D. (2004). The Sketch Engine, in Williams and Vessier (Eds.) Proceedings of the Eleventh Euralex Congress, UBS Lorient, France: Tapanainen, P. (1996). The Constrai nt Grammar Parser CG-2. Publication No. 27, University of Helsinki. Uí Dhonnchadha, E., Nic Pháidín, C. and Van Genabith, J. (forthcoming). Design, Implementation and Evaluation of an Inflectional Morphology Finite-State Transducer for Irish. Machine Translation Journal, Special Issue on Finite State Language Resources and Language Processing. 148

162 WORDS IN ASIAN CULTURAL CONTEXTS Grammatical class, tags and lemmas: a corpus-based study of the Malay lexicon Abstract Gerry Knowles and Zuraidah Mohd Don Lancaster University g.knowles@lancaster.ac.uk Grammatical class in Malay is superficially straight forward, but does not pattern in real texts as one might expect. This paper reports on work undertaken on the MALEX project, which takes a systematic approach to the structure of Malay, using corpus data so far amounting to some 2.5M words. The first stage was to design a tagset for Malay, and it quickly emerged that Malay grammatical classes do not operate like English parts of speech. The complexity of Malay syntax makes it inappropriate to talk of ambiguity in tagging, and an alternative approach had to be devised. Work on the tagset required an analysis of the morphology, which in turn raised the question of the status of the Malay lemma. Whereas English lemmas are by definition tied to a single part of speech, Malay lemmas are structured in a different way and contain words of several grammatical classes. The classification of Malay lemmas raises fundamental theoretical issues different from those involved in tagging individual items. As far as we are aware, these issues have not hitherto been addressed in corpus linguistics. However, they have major practical consequences for Malay lexicography, since assigning a class to dictionary headwords is not the same as assigning a class to words grouped under the headword. At the time of writing, the database contains some 30,000 lemmatised lexical items extracted from the corpus, and work is underway on the classification of lemmas. Theoretical aspects of the compilation of the Macmillan Essential Dictionary Workbook Abstract Yuri Komuro Chuo University yurikomuro@jcom.home.ne.jp Although the necessity and importance of dictionary instruction has been recognized and emphasized, it has not been widely or systematically introduced in the classroom, and this is partly because the burden falls on teachers, who are already overloaded with class preparation and lots of other work. The Macmillan Essential Dictionary Workbook (MEDW), whose exercises can be immediately taken into the classroom activity, aims to help in such a situation. In my paper presentation, the theoretical aspects of the compilation of the MEDW will be explained. The MEDW is intended to help intermediate learners, including those who first try to use a monolingual learner s dictionary, to learn what kinds of information the Macmillan Essential Dictionary (MED) provides, where to find it, and how to make the best use of it not only for decoding tasks, but also for encoding tasks. The exercises are designed to help learners 149

163 ASIALEX 2005 to learn mainly: (1) dictionary conventions such as abbreviations and codes; (2) definition patterns; (3) grammatical and collocational patterns; and (4) spoken phrases. The paper also discusses some difficulty in designing such exercises in order to develop more practical and effective exercises. And the tiger cries, animal metaphorical expressions in the Thai language Abstract Montatip Krishnamra University Of Michigan Whether it is mama-huhu in Chinese, barking dog never bites in English, or hoeru inu wa kamanai in Japanese, using animals as metaphors represents an integral part of any language, including Thai. Metaphoric expressions are a significant part of language acquisition, not only in the linguistic aspect, but also in the cultural aspect. Learners of foreign language, in particular, will not be able to be fully immersed in the language unless they understand the meaning of the metaphors in the target language. Some metaphors across cultures may be identical in form but are opposite in meaning. A very good example is represented in the use of the metaphor white elephant in the English language which conveys a negative connotation, whereas white elephant in Thai metaphor suggests a very positive one. As a result, one cannot assume that because the words are identical, the meanings are the same. This paper will explore different facets of animal metaphorical expressions in everyday Thai language, including how the metaphors are originated, how they are used, and what socio-cultural impacts they have on the Thai society. Similarities and differences between western and Thai perceptions of animal characteristics which are adopted as metaphors will also be covered. 150

164 WORDS IN ASIAN CULTURAL CONTEXTS Lay and lie Where should they go? A practical hands-on workshop on lemmatizing English words Abstract Kam-Mei Jacqueline Lam, Sue Chang, Gregory James and Bronson So Hong Kong University of Science and Technology lcjacqui@ust.hk A team of researchers at the Language Centre of the Hong Kong University of Science and Technology, working on the assumption that tertiary-level students need to know the academic senses of core English words in order to cope with their university work (cf. Laufer, 1989; Nation, 1997), has been developing an annotated list of everyday general and semi-technical vocabulary based on a 5-million word corpus, the HKUST Wordlist. This list constitutes an advance on Nation & Coxhead s Academic Word List (AWL) of 570 word families, developed at Victoria University, Wellington (Coxhead, 1998, 2000). Following the AWL, the HKUST Wordlist has been lemmatized by using an extended version of Someya s on-line program, E-lemma.txt, Ver. 1, 1998, so as to reflect the occurrence of base words and their families across disciplines. In the process of lemmatizing the words manually, our team learned that there is no unique set of rules for lemmatizing English words. Rather, the list should reflect the students needs. In this workshop, through hands-on activities, we invite participants to address questions such as: How do we establish headwords? How should we categorize spelling variants, suffixes, prefixes and compound words? What is the best way to deal with hyphenated words, abbreviations, key proper names and, above all, polysemy and words with inflectional and derivational meanings? A three-way dictionary bringing Cantonese, Putonghua and English together Jacqueline Lam Hong Kong University of Science & Technology lcjacqui@ust.hk Lan Li Hong Kong Polytechnic University eglilan@polyu.edu.hk Tom McArthur Editor, English Today Cambridge University Press scotsway@aol.com Abstract It is more than a decade since the Hong Kong government advocated implementing a biliterate and trilingual language (BTL) policy: that is, written Chinese and English and spoken Cantonese, Putonghua and English. However, any viable means of realizing the government s policy of a biliterate and trilingual workforce needs to bridge the gap between the Chinese- and English-medium dimensions in the secondary school systems, in effect making the languages complementary. However, given that English and Chinese are grammatically and phonologically very different, learning both is far from easy. To implement a BTL policy, learners of two writing systems and three spoken languages 151

165 ASIALEX 2005 need to be alerted regarding how their mother tongue (for the vast majority, Cantonese) works within a broader comparative linguistic framework. In this paper, we argue that Hong Kong s role as a world city cannot easily be sustained and strengthened without a strong BTL policy with appropriate dictionaries. Although currently a limited project, our proposed Bi-Tri dictionary represents a three-ways-in contribution towards implementing the government s master plan, both for local students and other interested persons. Introduction: Sociolinguistic development since 2002 This is the third report on the Bi-Tri project since The first report, What kind of dictionary do Chinese students need in Hong Kong? A biliterate and trilingual dictionary, was presented at the third AsiaLex Biennial International Conference Dictionary and Language Learning: How can Dictionaries Help Human and Machine Learning?, in Tokyo, Japan in August The second report, The Chinese languages: A new lexicographical perspective from Hong Kong, was presented at the Congress of the World s Major Languages, organized by the Dewan Bahasa dan Pustaka (Institute of Language and Literature) in Kuala Lumpur, Malaysia, in October, In the meantime, we have made progress in compiling the BiTri Basic, the first of two biliterate and trilingual dictionaries, having to date completed 2,000 entries. In 2002, Jacqueline Lam presented a paper at the International Conference Translation and Bilingual Dictionaries at the Chinese University of Hong Kong. It was co-authored with Tom McArthur and entitled Could there be a dictionary tailor-made for Hong Kong: Both biliterate and trilingual? In that paper, we mentioned that the then Chief Executive Tung Chee-hwa announced Hong Kong s new biliterate and trilingual language policy in his first speech to the Provisional Legislative Council in In his second term in 2003, he stressed again the importance of a BTL policy in his first Policy Address. Further research has however indicated that such a language policy was in fact proposed well before the handover and some measures had been taken to realize it. The first move was in In response to public pressure, the British colonial government enacted the Official Languages Ordinance, which declared that English and Chinese were the official languages of Hong Kong for the purposes of communication between the government or any public officer on the one side and members of the public on the other (cf. So, 1996; Poon, 2004). In 1995, the Government began to develop a biliterate and trilingual civil service, and in the Report of the Working Group on the Use of Chinese in the Civil Service, it foreshadowed a fundamental change in the language orientation of the Government (Lau, 1995). In 1996, a biliterate and trilingual policy was first proposed as a language-in-education policy in the Education Commission Report, although without any description of a framework or plan for its implementation. Since 2002, the determination of the government to implement a BTL policy has intensified, and by and large the community, where and when it thinks about such things, has appeared to share the sentiment. This interest is for example explicitly and constantly demonstrated by calls for bi-tri skills in job advertisements in the South China Morning Post. In June 2003, the Standing Committee on Language Education and Research 152

166 WORDS IN ASIAN CULTURAL CONTEXTS (SCOLAR 1 ) released its Action Plan to Raise Language Standards in Hong Kong, which was in effect an implementation plan to carry out the BTL policy. Unlike former language policies, most of which focused only on the English language, the Action Plan pays heed to Chinese, notably emphasizing the use of simplified Chinese characters in writing and of both Putonghua and Cantonese in speech. In addition, departing from previous language policy, the plan adopted the perspective of employers and sought to motivate the business world to join hands with the government to reform the Hong Kong education system 2 (cf. SCOLAR, 2003 pp. 45-6; Poon, 2004). In December 2004, the Education and Manpower Bureau (EMB) issued its third progress report on education reform, in Section 2, Language Education, again making it clear that the language policy of the Government is to enable students and the working population to be biliterate (in written Chinese and English) and trilingual (in Cantonese, Putonghua and English). In February 2005, Chris Wardlaw, Deputy Secretary for Education and Manpower, stated that there are rising expectations regarding language skills in Hong Kong, noting in his paper Supporting Language Learning in Hong Kong: Are we doing enough? That: whatever past language standards have been, current community needs and expectations place a premium on higher levels of trilingualism (Cantonese, Putonghua and English) and biliteracy (Chinese and English) to support the rapid restructuring of Hong Kong, its rich culture and global inter-connectedness.. Together the government, education and business sectors and the wider community are investing heavily in the aspirational goal of a trilingual and biliterate Hong Kong. In March 2005, SCOLAR issued its final report regarding its Second Action Plan to Raise Language Standards in Hong Kong, which called for a two-montconsultation in January. The report again reminded Hong Kongers that being period of public biliterate and trilingual is to our competitive advantage because such language ability helps bridge the gap between an English-speaking global business community and Chinese-speaking merchants and traders in Hong Kong and on the mainland of China, a factor contributing to Hong Kong s success as an international city. The report further emphasizes that proficiency in Cantonese, English and Putonghua is essential for life-long learning and the communication of knowledge, ideas, values, attitudes and experience (p.7). Internationality and language reality Hong Kong has long been variously described as an international city, Asian s world city, a cosmopolitan city, a cosmopolitan trading hub and an international cultural metropolis (cf. Lau 1995; Tung 2003 & 2005; SCOLAR, 2005). From a demographic perspective, Hong Kong has always been more than simply a Chinese city, because its people are varied. Although its population is mostly Chinese (94.9% of its 6.5 million people in the 2001 Population Census), it includes people whose forefathers came there, or who themselves came there from other places. Such people are as much part of the community as the Chinese (Ching, 2005). From a majority perspective, 1 SCOLAR was set up in early 2001 by the Hong Kong Government to review language education in both schools and the wider community. Its aim is to develop a set of recommendations on ways of raising language standards in Hong Kong. 2 A coalition on Education in the Business Sector was formed by the Federation of Hong Kong Industry and 10 Chambers of Commerce to examine Hong Kong education system and propose ways to improve it from the perspective of the business sector (Ming Pao Daily, 3 September 1999). 153

167 ASIALEX 2005 however, one can assert that Hong Kong is in essence a Chinese city, in which Cantonese has always served (and is likely to continue to serve) as either a mother tongue or a lingua franca for Chinese people of all backgrounds. We must note, however, that pressure to speak Putonghua has been steadily increasing. At the same time, although English continues to be widely used in such public domains as government, law, and education, it often interfaces with Chinese through code-switching and code-mixing, both at the spoken and the written level. If Hong Kong is to li ve up to its wish to be international, cosmopolitan and Asia s world city, its people will by and large need to become proficient biliterates and trilinguals, and it is with this in mind that our idea of BiTri dictionaries has been evolving. In creating such products, the language base has necessarily been English, because more work has been done on the teaching, learning, and lexicography of English a s a second or foreign language (ESL, EFL) than on any other language in the world. A s a result, it has a stronger tradition of learners dicti onaries than any other language. In addition, since Putonghua and its romanization (Hanyu Pinyin) have received considerable linguistic and lexicographical attention, we do not anticipate major problems in compiling the component that focuses on mainland China. The most innovative and demanding element, however, has been the foregrounding and in effect upward equa lization of Cantonese, on the basis that it is the key linguistic medium of Hong Kong, Macau, Shenzhen and Guangdong province. As the mother tongue of the vast majority of Hong Kong people, it is spoken by over seven million people and should therefore serve as the lexicographical anchor point for the majority of users. However, the sad fact is that, although Cantonese is hardly likely to die out, and although the two SCOLAR Action Plans recommend that teachers should master and teach students standard Cantonese pronunciation (2003), a majority of Chinese (including the Cantonese at large and Cantonese scholars in particular) thinks that Cantonese is no more than a dialect and that its distinctive written form is of doubtful status (Harrison & So, 1996). As a result, Cantonese gets little objective attention in its own community. In Europe, however, one time-honoured way of turning a dialect into a language has been to give it a dictionary, and Cantonese certainly has dictionaries. The first was published in 1877 by Ernest John Eitel, a German missionary, with distinct characters, such as for zo 2, hai 6, lai 4, ge 3, go 3, 喥 dou 6, 嘢 ya 5, 嘛 ma 3, 佢 keoi 5, that are not part of the traditional set of Chinese characters. In this graphic sense, Cantonese is a language, with more than twenty often very different systems of romanization. In addition, its widely varied and non-standardized systems for reflecting its pronunciation in writing have unfortunately made the language difficult to learn. The BiTri Basic aims to address this issue, among others. Firs t, we have adopted a mildly modified form of the Jyutping system of romanizatio n developed and promoted in recent years by the Linguistic Society of Hong Kong. The Society s Jyutping Pronunciation Guide 粵語拼音字表 (jyut 6 jyu 5 ping 3 yam 1 zi 6 biu 2 ) has been developed by a group of academics working over the past ten years in five universities in Hong Kong, and recently the Hong Kong Government adopted it as the default Cantonese romanization system for its civil service. This is the system we use in the BiTri Basic. Moreover, our research to date shows that such a modified Jyutping system brings its learners closer to the native Cantonese pronunciation than any other romanization systems currently at work. Second, if Hong Kong wishes to live up to its claim of being Asia s world city, its citizens will need to be able to communicate with both non-cantonese users of 154

168 WORDS IN ASIAN CULTURAL CONTEXTS English worldwide and with Mainlanders in Putonghua who do not know Cantonese. Any language education system with such a goal in mind should start early, and because of this the target users of the BiTri Basic are lower-secondary students (Secondary 1 to 3, aged 13 to 16), almost all of whom are mother-tongue speakers of Cantonese who need to build up a sound basic vocabulary in English and in Putonghua. To fulfil the city s claim that Hong Kong is an international cultural metropolis, they also need recognition and reading skills in the roman alphabet for English, in Hanyu Pinyin for Putonghua, and, importantly, we would argue, in a Jyutping-based romanization for Cantonese so that they will be able to help non-cantonese with Hong Kong s key language. Third, it is tim e to unify the romanization of Cantonese, an aim that can be achieved by supplementing the BiTri Basic with sets of practice materials that turn it into a workbook for teaching children one system of Cantonese romanization after they have learned to read and write in the traditional way. This is in effect an attempt to match the basic language-teaching approach in the mainland, where Hanyu Pinyin for Putonghua is taught to students as early as their kindergarten years and serves as a base for learning to read and write characters. Aligning the languages Aligning such very different languages as English, Cantonese, and Putonghua/Mandarin is hardly a simple business, but it is in our view one of the most interesting and useful l exicographical activities one might engage in at the beginning of the twenty-first centur y. During the twentieth century, a great deal of lexicographical attention was lavished on the English language, in the United Kingdom, the United States and indeed worldwide, and in both the native area of dictionary-making and the booming dimension of ELT lexicography, which was not only a commercially attractive undertaking but also an area of considerably innovation, most particularly with regard to handling the divide between UK and US usage (in terms particularly of differences in pronunciation, spelling and the use of particular words). As a result, in creating the base list of English words for the BiTri project we first of all needed to establish what, for our purposes, the key words of English might be, both internationally and in terms of an English/Chinese interface. One advantage in this regard is that Tom McArthur has for many years been interested in the nature and history of English word lists developed for graded readers to be used with both native-speaking children and foreign learners, but most particularly as defining vocabulary for dictionaries of English as a second or foreign language. Even if such word lists have never actually represented the minimum essential words among children growing up in the world s various native English-speaking milieux, there is a century-old tradition of making, using, and researching such lists for educational purposes, as a result of which it probably represents the firmest basis for such work as interests us here than any other basis in any language. If not in this area, where could one look for a place in which to start? Selecting the words In brief, our starting point has been a synthesis of the key lexis in four firmly grounded dictionary-related word lists: first, Ogden s Basic English and Michael West s General Service List, the outcome of work begun in the first half of the twentieth century, and the defining vocabularies of the Longman, Macmillan, and Cambridge 155

169 ASIALEX 2005 dictionaries, in the second half of the century, resulting in a list of some 3,500 words. The Bi-Tri Basic therefore draws on a successful and long-established English-language tradition. While continuing and extending this tradition, the BiTri Basic word list responds both to the perceived needs and expectations of Hong Kong users and word lists that have emerged for the two Chinese languages. Therefore, while based primarily on, but by no means restricted to, The Oxford 3000 word list (cf. OALD, 7th edition, 2005), the words in the BiTri Basic English list are common and used in a wide range of contexts and texts, pragmatically including among them words that do not have a high frequency but are important and familiar to users of English worldwide. These words cover such everyday themes as parts of the body and household contents and appliances. At the moment, the overall list can only be conceived in terms of English, but in due course there will be three parallel and mutually-re-enforcing learners lists, the words in each serving to define the words in the others. A second source of English material, specific to Hong Kong and its people, will include such items as bird s nest, Canto-pop, da paidong and laishi. This local material will be culled from the South China Morning Post, the major English newspaper in Hong Kong, and other reliable and typical sources, and will tie in with a set of appended themes that includes aspects of local geography, travel, transport, food, restaurants, and education. There are however some aspects of language that we want to emphasize, so as to provide the most practical service possible. For example, we want to ensure that learners become acquainted early on with key English phrasal verbs and some of their idioms. As a result, we have incorporated them under such headwords as: Adjustments be in be in, be out, be up, be down, be off call in call in, call out, call up, call down, call off get in get in, get out, get up, get down, get on, get off And so on through lay and lie to see, set, and take. In the course of compilation, adjustments will need to be made, in order to refine our approach, especially as a consequence of the interaction of material in and across English, Cantonese, and Putonghua/Mandarin, especially so as to adapt any inadequate preconceptions we started out with, as regards the basic vocabulary of each language and adjusting it to the realities of putting everything together in two delivery systems : one, with printed English as the entry point, the other with character-based Chinese as the entry point: plus any supplementary back-of-the-book lists that may help users get as quickly and painlessly as possible from what they know to what they want to know. In effect, such a procedure (as in any work of reference) is the systematization and stylization of three fundamentals: easy access, adequate presentation, and (with luck) acquisition and assimilation. Such work means constant cross-checking and adjusting among the representatives of the two writing systems and three spoken languag es. In addition, we will need procedures for adding further words that present themselves for any reason as essential at the level we are aiming at: that is, the basic operating vocabularies of the three languages. Inevitably, further words will present themselves, but we currently assume that a point will arrive where we feel that the circle of interacting basic 156

170 WORDS IN ASIAN CULTURAL CONTEXTS vocabulary is as complete as we need it to be. A triangular balance of entries The principle we have kept in mind is that the three languages are equal in microstructural terms. Each entry is therefore organized according to the following scheme: Headword Grammatical category English British and American variants (if any) Equivalent in traditional Chinese characters (in Mainland terms: Complex Chinese) Chinese Equivalent in Simplified Chinese characters Hanyu Pinyin Colloquial Putonghua characters & Hanyu Pinyin (if any: shaded) Putonghua Cantonese romanization (a variant the JyutPing system) Colloquial Cantonese characters and its romanization (if any: Cantonese shaded) Examples Derivatives Table 1. Micro-structure of an entry Many of the basic words of English are well known for their multiple meanings. Thus, in the Collins English Dictionary (CED, Millennium Edition), the noun case has eighteen senses relating to instance and ten to container. The Macmillan English Dictionary for Advanced Learners (MED) does not separate the two, providing seven senses that include phrases relating to case. In its turn, the Oxford Advanced Learner s English Chinese Dictionary (OALECD, 2004 edition) has ten senses of case, starting from situations in sense 1 to container in sense 8. This indicates that even with a carefully selected wordlist, the decision regarding the senses of any entry is challenging. However, the BiTri Basic, with secondary students as its key end users, draws on a long-established tradition of taking words in their commoner senses and uses. We take any generic and concrete sense or senses of a word before any specialized and abstract sense. The current entry for the word case is shown below: case n 1 箱子, 箱 P. xiāngzi, xiāng C. soeng 1 zi 2, soeng 1 pencil case筆 ( 筆 ) 盒 P. bǐhé C. bat 1 hap 2 suitcase 旅行箱 P. lǔxíngxiāng C. leoi 5 haang 4 soeng 1 ; 旅行 leoi 5 haang 4 gip 情況, 事件 P. qíngkuàng, shìjiàn C. cin fong, si gin in any case 無論 ( 无论 ) 如何 P. wúlùn rúhé C. mou 4 leon 6 yu 4 ho 4 in that case 既然那樣 ( 样 ) P. jìrán nàyang C. gei 3 yin 4 naa 5 yoeng 6 Figure 1. Sample entry of the word case In dictionary compilation, frequency and usage are always the first selection criteria. The frequency information (of a word in corpora), however, is not difficult to obtain in this hi-tech era with various software of natural language processing. Leading 157

171 ASIALEX 2005 dict ionaries nowadays, whether produced in the UK or in USA all claim corpora of contemporary English are a well of lexical information, and a few have included different marks to indicate frequency of entry words, such as Collions and Macmillan. A kind of information commonly available to lexicographers is word frequency, but unfortunately sense frequency, which is more useful in dictionary compilation, has not (yet) been well-documented. Great effort has been made by computational linguists in semantic annotation and classification of word senses by employing algorithm into a computer to disambiguate a word with its gloss. Senses of justice Sense 1 justice, justness => righteousness => morality => quality => attribute => abstraction Sense 2 virtue => natural virtue => cardinal virtue => virtue => good, goodness => morality => quality => attribute => abstraction (from WordNet) Figure 2. A sample algorithm of the word justice The results of sense frequency can be obtained automatically by computing such an algorithm, but tagging a corpus is by no means an easy task. More extensive experiments, which include POS-tagging the gloss of all WordNet nouns, are under way (Rosso 2003). It has also been noticed that there is always ambiguity that needs to be resolved from the context. The literature has not shown a satisfactory picture of sense discrimination and sense frequency. Such practice in dictionary compilation is probably still based on manual analysis when one is looking at concordances of a word, using personal experience, intuition and decision to judge the meaning and order of different senses. We checked with frequency-based lists from different resources and then compare them to find the dominant ones. For example: degree n 1 程度 P. chéngdù C.cing 4 dou 6 degree of accuracy 準確 ( 準確 ) 程度 P. zhǔnquè chéngdù C. zeon 2 kok 3 cing 4 dou 6 2 溫 ( 溫 ) 度 P. wēndù C. wan 1 dou 6 five degrees Celsius 攝氏五度 P. shèshì wǔdù C. sip 3 si 6 ng 5 dou 6 3 角度 P. jiǎodù C. gok 3 dou 6 an angle of 40 degrees 四十度角 P. sìshídù jiǎo C. sei 3 sap 6 dou 6 dik 1 gok 3 ; 四十度嘅角 sei 3 sap 6 dou 6 ge 3 gok 3 Figure 3. Sample entry of the word degree 158

172 WORDS IN ASIAN CULTURAL CONTEXTS Bearing in mind that the Bi-Tri Basic is a localized dictionary, we have added elements spec ial to the interests of Hong Kong people, such as in the entry dollar. dollar n 元, 圓 ( 圆 ) P. yuán C. yun 4, yun American /US dollar 美元, 美金 P. měiyuán, měijīn C. mei yun Hong Kong dollar 港幣 ( 币 ) P. gǎngbì C. gong 2 bai 6, 港元 gong 2 yun 4 ; 港紙 gong 2 zi 2 Figure 4. Sample entry of the word dollar Although Standard Chinese used in Hong Kong is based on Putonghua, the northern language, there are variants where Cantonese people use different Chinese characters but not necessarily colloquial terms, so the entry structure has to be moderated. For instance, the words paint and plastic: paint n P. 油漆, 颜料 yóuqī, yánliào C. 漆油, 顏料 cat 1 yau 2, ngaan 4 liu 2 plastic n P. 塑料 sùliào C. 塑膠 sok 3 /sou 3 gaau 1 plasti c bag P. 塑料袋 sùliàodài C. 膠袋 gaau doi 1 2 Figure 5. Translation difference between Cantonese and Putonghua The BiTri Basic does not include English phonetic pronunciations so as to avoid using more symbols and contrasts than users can comfortably handle. Hanyu Pinyin is used for Putonghua and a modified form of Jyutping for Cantonese. Hanyu Pinyin has been standarised by the government for nearly fifty years in mainland China, so phoneticising the characters is straight forward. As for the Cantonese, there are over thirty inventions for marking its sound system. The Jyutping we have adopted, is a scholarly developed system developed and promoted by a group of Hong Kong academics in the past ten years. Although Hanyu Pinyin came into use in 1958 under the guidance of an institution of central government, confusions, especially in segmentation, still exist when it is used in texts. For words with one or two characters, it is not difficult to handle. When a word has more than three syllables, or when a Chinese idiom is brought into the text, the compiler must follow certain rules to keep consistent. For instance, careless adj 不小心的, 粗心大意的 P. bù xiǎoxīn de, cūxīndàyì de C. bat 1 siu 2 sam 1 dik 1, cou 1 sam 1 daai 6 yi 3 dik 3 ; 唔小心嘅 ng 4 siu 2 sam 1 ge 3 partly adv 部份地, 在一定程度上 P. bùfèn de, zài yīdìng chéngdù shàng C. bou 6 fan 6 dei 6, zoi 6 yat 1 ding 6 cing 4 dou 6 soeng Figure 6. Segmentation of Hanyu Pinyin and Jyutping When Pinyin is added after the character group 不小心的, we have to decide 159

173 ASIALEX 2005 whe ther it is in the form bùxiǎoxīnde, or bùxiǎoxīn de, or bù xiǎoxīn de; for 粗心大意的 cūxīn dàyì de, or cūxīndàyì de, or cūxīndàyìde. The standard we follow is the Chinese Pinyin Glossary published by the Committee for Language Reform of China (CLRC) in 1965, which includes 59,100 words with 2,100 single syllabic words, 35,000 double syllabic words, 15,000 triple ones and 6,000 phrases and idioms with four or more syllables. We have also followed the Basic Rules for Hanyu Pinyin Orthography provided by the CLRC in 1996, which advocates such an arrangement as 分詞連寫, fēnc í liánxiě, which puts multi-syllabic words together and separates parts of speech. For example, the word broad in English is one word, but its equivalent in Putonghua/Mandarin is 廣闊的 guǎngkùo de (N + particle), which separates off the particle de. It has been noticed that various Cantonese romanisation systems, mainly appearing in glossaries, do not have segmentations when they phoneticise sentences. Such a long string of letters without separation must however be hard work for learners. We have therefore adopted Pinyin rules to Jyutping, so that the segmentation alignment of Putonghua and Cantonese are identical. Pinyin not only serves as a pronounciation system but also promotes the romanisation of Chinese. As such, it has been used at sentence level with capitalisation and punctuation in children s reading alongside Chinese characters. When we began compiling the BiTri Basic, we used Pinyin and Jyutping only at the word level, to indicate the sound of a word. When more content and examples were added, however, we integrated the rules of the Pinyin and English writing systems, for example capitalizing the first letter of a sentence. In all probability the neatest ultimate solution to the problem of pronunciation for all three languages would be to have both the print and spoken versions available in electronic form, a level of sophistication we may not reach for some time. Bringing Cantonese, English and Putonghua together in the way we envisage is hardly simple, but we believe that it can be done, and that it does respond to the Hong Kong gov ernment s policy regarding language and education. At the very least it sets the three systems side by side, and allows comparisons to be made and parallel learning to take place. At its best, if the idea catches on, it could serve to promote a new way of learning together languages that Hong Kong citizens have to use together. References Chan S. W. (Forthcoming). Lexicography in Hong Kong: In The Hong Kong Linguist, No. 25. Ching, F Make every Hongkonger feel at home, South China Morning Post, 22 March. Education and Manpower Bureau Progress Report on the Education Reform 3, December. Retrieved 28 March 2005 at < progress%20eng.pdf>. Guangdong Education Bureau Basic Rules for Chinese Language and Orthography. Guangzhou: Jinan University Press. Harrison, G. and So, Lydia K. H The background to language change in Hong Kong. In Current Issues in language and Society, 3(2). Lam, J. K. M. & McArthur, T Could there be a dictionary tailor-made for Hong Kong: Both biliterate and trilingual? In S.W. Chan (Ed.) Translation and BilingualDictionaries. (Lexcographica 160

174 WORDS IN ASIAN CULTURAL CONTEXTS Series Maior 119). Tübingen: M. Niemeyer, Lam, J. K. M Chinese languages: A new lexicographical perspectives from Hong Kong. In two related India-based online journals: Language in World, Vol. 1. at < languageinworld. com> and Language in India, Vol. 4 (12 December 2004) at < Lam, J. K. M. (Forthcoming). The type of dictionary Chinese students need in Hong Kong: A biliterate and trilingual dictionary. In The Hong Kong Linguist, No. 25. Lau, C. K Language of the future. South China Morning Post, 18 September. Poon, A. Y. K Language policy of Hong Kong: Its impact on language education and language use in post-handover Hong Kong. Journal of Taiwan Normal University, 49(1), Rosso, P et.al Word sense disambiguation combining conceptual distance, frequency and gloss. In NL-PKE2003 Proceedings. < NL-PKE2003/pdf/A05-01.PDF>. SCOLAR Action plan to raise language standards in Hong Kong. Hong Kong: SCOLAR, Education and Manpower Bureau. So, D. W. C Language Policy. In P. Dickson & A. Cumming, eds, National Profiles of Language Education in 25 Countries. Berkshire, England: National Foundation for Educational Research. Tung, C. W In the Foreword of the programme for the Ballet Nacional de España, 33rd Hong Kong Arts Festival. Hong Kong. Appendix: A chronology of language dictionaries published in Hong Kong (culled from the reference list of Chan S. W Lexicography in Hong Kong: In The Hong Kong Linguist, no. 25). Cantonese dictionaries, Eitel, Ernest John, comp. A Chinese Dictionary in the Cantonese Dialect, 3 vols. London: Trübner and Co.; Hong Kong: Lane, Crawford & Co Cowles, Roy T., comp A Pocket Dictionary of Cantonese. Hong Kong: Hong Kong University Eitel, Ernest John, comp. A Chinese Dictionary in the Cantonese Dialect, 3 vols., revised and enlarged by Immanuel Gottlieb Genähr. Hong Kong: Kelly & Walsh Well, Herbert Richmond, ed. An English-Cantonese Dictionary. Hong Kong: Kelly and Walsh Meyer, Bernard F. and Theodore F. Wempe, comps. (1935). The Student s Cantonese-English Dictionary. Hong Kong: Catholic Truth Society. Bilingual (English-Chinese) dictionaries published in Hong Kong, Zhang, Shiliu 張世鎏, comp. 求解作文兩用英漢模範字典 (Model English-Chinese Dictionary), 20th edition. Hong Kong: The Commercial Press 商務印書館 Lin, Fuhua 林輔華 and Xia Mingru 夏明如, trs. and eds. 四福音大辭典 (A Dictionary of Christ and the Gospels), 2 vols. Shanghai: Society for the Dissemination of Knowledge 廣學會 Wang, Yuming et al. 汪余鳴等, comps. 自然科學辭典 (A Dictionary of Natural Science). Hong Kong: The World Press 世界出版社 Lin, Fuhua 林輔華 and Xia Mingru 夏明如, trs. and eds. 四福音大辭典 (A Dictionary of Christ and the Gospels). Hong Kong: Bible Society 聖書公會 Zhu, Jixuan 朱積暄, comp. 應用化學辭典 (An English-Chinese Dictionary of Applied Chemistry). Hong Kong: The Commercal Press 商務印書館 Chan, Yan 陳印, comp. 外國郵票地名譯文手冊 (A Handbook of Translations of Place Names on Foreign Stamps). Hong Kong: The Commercial Press 商務印書館 Sun, Jue 孫覺, tr. (1959). 地理學小辭典 (A Dictionary of Geography). Hong Kong: Oriental Geographical Society 東方輿地學社 Weng, Liang 翁良 and Yang Shixi 楊士熙, eds. 英文成語大詞典: 英華雙解 (A Complete Dictionary of English Phrases with Bilingual Explanations). Hong Kong: Hong Kong Zhiming Bookstore 香港知明書店 Zheng, Yili 鄭易里, comp. 袖珍英漢辭典 (A Pocket English-Chinese Dictionary), paperback, revised edition, traditional characters. Hong Kong: The Commercial Press 商務印書館. 161

175 ASIALEX Zheng, Yili 鄭易里, ed. 最新詳解英華大辭典 (A New English-Chinese Dictionary). Hong Kong: Guangtai Bookstore 廣泰書局 Zhang, Qichun 張其春 and Cai Wenying 蔡文縈, comps. 簡明英漢詞典 (A Concise English-Chinese Dictionary). Hong Kong: The Commercial Press 商務印書館 Wu, Xizhen 吳奚真, ed. 現代高級英漢雙解辭典 (The Advanced Learner s Dictionary of Current English). Hong Kong: Oxford University Press Translation Sub-committee, the Industrial Training Advisory Committee 工業訓練諮詢委員會繙譯小組委員會, comp. 技術及工業訓練詞匯手冊 (Manual of Technical and Industrial Training Terms). Hong Kong: The Government Printer 政府印務局 Editing Group, A New English-Chinese Dictionary 新英漢詞典 編寫組, comp. 新英漢詞典 (A New English-Chinese dictionary). Hong Kong: Joint Publishing Co. (HK) Ltd. 三聯書店 ( 香港 ) 有限公司 Transport Department 運輸署, comp. 運輸署常用詞匯 (Commonly-used Terms in Transport Department). Hong Kong: The Government Printer 政府印務局 Committee on Translation, Hong Kong Vocational Training Council 香港訓練局翻譯事務委員會, comp. 英漢工業常用詞匯 (An English-Chinese Glossary of Common Technical Terms). Hong Kong: Hong Kong Government Publication Bureau 香港政府印務局 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第一輯 : 教育 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 1: Education). Hong Kong: Government Printer, 政府印務局 Loyang Tractor Institu te 洛陽拖拉機研究所, com p. 英漢汽車詞典 (English-Chinese Dictionary of Automobile). Hong Kong: The Commercial Press Zhang, Xuegong 張學工 comp. 最新英漢電工辭典 (New Technical Dictionary of Electrical Engineering: English-Chinese). Hong Kong: Wan Li Book Company 萬里書店 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第二輯 : 房屋與地政 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 2: Housing and Land). Hong Kong: Government Printer 政府印務局 Editorial Committee, An English-Chinese Glossary of Psychology, Psychology and Social Research Centre 中譯心理學詞匯 編譯組與社會研究中心, e d. and tr. 中譯心理學詞匯 (An English-Chinese Glossary of Psychology). Hong Kong: The Chinese University Press Ho, Hsiu Hwang 何秀煌, comp. 中譯邏輯學詞匯 (An English-Chinese Glossary of Logical Terms). Hong Kong: The Chinese University Press Lü, Zhongxian 呂仲賢, comp. 英漢電腦技術詞匯 (English-Chinese Computing Technique Dictionary), revised and enlarged edition. Hong Kong: The Commercial Press Chang, Jonathan 張王楊, comp. 英漢貿易商業詞典 (English-Chinese Dictionary of Trade and Commerce). Hong Kong: The Commercial Press 商務印書館 Computer Society of the Chinese Association of Electronics, The 中國電子學會電子計算機學會, comp. 新英漢電腦辭典 (A New English-Chinese Computer Dictionary). Hong Kong: Joint Publishing Co. (HK) Ltd. 三聯書店 ( 香港 ) 有限公司 Xie, Detong 薛德炯, tr. 英漢化學辭典 (Anglo-Chinese Chemical Dictionary). Hong Kong: The Commercial Press 商務印書館 Zheng, Yili 鄭易里, ed. 英華大詞典 (A New English-Chinese Dictionary), second revised edition. Hong Kong: The Commercial Press 商務印書館 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第四輯 : 銓敘 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 4: Civil Service). Hong Kong: Government Printer 政府印務局 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第五輯 : 交通 (An English-Chinese Glossary of Terms Commmonly Used in Government Depa rtments, Vo lume 5: Transport). Hong Kong: Government Printer 政府印務局 Lou, Er-ying 婁爾行 and John B. Farrell 約翰.B. 法雷爾, eds. 英漢, 漢英中美會計詞匯 (English-Chinese, Chinese-English Accounting Terminology in Use in the PRC and the USA). Hong 162

176 WORDS IN ASIAN CULTURAL CONTEXTS Kong: Joint Publishing Co. (HK) Ltd. 三聯書店 ( 香港 ) 有限公司. 1986a. Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會, comp. 中學電腦科常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Computer Studies in Secondary Schools). Hong Kong: Curriculum Development Committee Education Department Hong Kong 香港教育署課程發展委員會. 1986b. Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會, comp. 中學社會教育科常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Social Studies in Secondary Schools). Hong Kong: Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會 Zhou, Jilian 周紀廉 and Jiang Xihe 江西河, eds. 當代英華詳解詞典 (A Contemporary English-Chinese Dictionary). Hong Kong: The Commercial Press 商務印書館 Chen, Tiejun 陳鐵君, comp. 英漢圖解汽車工程詞典 (Dictionary of Automobile Technology). Hong Kong: Wan Li Book Co. Ltd. 萬里書店 Department of Accounting and Finance, The Chinese University of Hong Kong 香港中文大學會計與財務學系, comp. and ed. (1987). 中譯會計與財務詞匯 (An English-Chinese Glossary of Accounting and Finance), second edition. Hong Kong: The Chinese University Press Department of Economics, The Chinese University of Hong Kong 香港中文大學經濟學系, tr. and comp. (1987). 中譯經濟學詞匯 (An English-Chinese Glossary of Economic Terms). Hong Kong: The Chinese University Press a. Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會 comp. 中學工業科目常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Technical Subjects in Secondary Schools). Hong Kong Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會. 1988b. Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會 comp. 中學經濟科常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Economics in Secondary Schools). Hong Kong: Curriculum Development Committee Education Department Hong Kong 香港教育署課程發展委員會. 1988c. Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會, comp. 中學科學科常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Science in Secondary Schools). Hong Kong: Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會. 1988d. Curriculum Development Committee, Education Department Hong Kong 香港教育課程發展委員會, comp. 中學人類生物科常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Human Biology in Secondary Schools). Hong Kong: Curriclum Development Committee, Education Depa rtment Hong Kong 香港教育課程發展委員會. 1988e. Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會, comp. 中學商業科目常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Commercial Subjects in Secondary Schools). Hong Kong: Curriculum Development Committee, Education Department, Hong Kong 香港教育署課程發展委員會. 1988f. Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會, comp. 中學音樂科常用英漢辭匯 (An English-Chinese Glos sary of Terms Commonly Used in the Teaching of Music in Secondary Schools). Hong Kong: Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會. 1988g. Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會, comp. 中學宗教教育科( 基督教 ; 天主教 ) 常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Religious Education (Christian) in Secondary Schools). Hong Kong: Curriculum Development Committee, Education Department Hong Kong 香港教育署課程發展委員會. 1988a. Curriculum Development Council 香港課程發展會. 中學地理科常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Geography in Secondary Schools). Hong Kong: Government Printer 政府印務局. 1988b. Curriculum Development Council 香港課程發展會. 中學家政科常用英漢辭匯 (An 163

177 ASIALEX 2005 English-Chinese Glossary of Terms Commonly Used in the Teaching of Home Economics in Secondary Schools). Hong Kong: Government Printer 政府印務局. 1988c. Curriculum Development Council 香港課程發展會. 中學歷史科常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of History in Secondary Schools). Hong Kong: Government Printer 政府印務局. 1988d. Curriculum Development Council 香港課程發展會 comp. 中學美術與設計科常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Art and Design in Secondary Schools). Hong Kong: Government Printer 政府印務局. 1988e. Curriculum Development Council 香港課程發展會 comp. 中學生物科常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Biology in Secondary Schools). Hong Kong: Government Printer 政府印務局 f. Curriculum Development Council 香港課程發展議會 comp. 中學數學科常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Mathematics in Secondary Schools), revised and amalgamated. Hong Kong: The Government Printer 政府印務. 1989a. Curriculum Development Council 香港課程發展議會, comp. 中學中六會計學原理科常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Sixth Form Principles of Accounts in Secondary Schools). Hong Kong: Government Printer 政府印務局. 1989b. Curriculum Development Council 香港課程發展議會, comp. 中學中六企業概論科常用英漢辭匯 (An English-Chinese Glossary of Terms commonly Used in the Teaching of Sixth Form Business Studies in Secondary Schools). Hong Kong: Education Department 香港教育署 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第六輯 : 勞工 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 6: Labour). Hong Kong: Government Printer 政府印務局 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第七輯 : 入境事務 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 7: Immigration). Hong Kong: Government Printer 政府印務 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第八輯 : 社會福利 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 8: Social Welfare). Hong Kong: Government Printer 政府印務局 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第九輯 : 醫療與衛生 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, V olume 9: Medical and Health). Hong Kong: Government Printer 政府印務局 Chinese Languag e Division, Hong Kong Gov ernment 香港政府中文公事管理局, comp. 政府部門英漢辭匯第十輯 : 公安 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 10: Public Order). Hong Kong: Government Printer 政府印務局 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第三輯 : 財經 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 3: Finance). Hong Kong: Government Printer 政府印務局 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第十一輯 : 環境保護 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 11: Environmental Protection). Hong Kong: Government Printer 政府印務局 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第十二輯 : 工業貿易 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 12: Trade and Industry). Hong Kong: Government Printer 政府印務局 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第十三輯 : 憲制與選舉事務 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 13: Constitutional and Electoral Affairs). Hong Kong: Government Printer 政府印務局 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部 164

178 WORDS IN ASIAN CULTURAL CONTEXTS 門英漢辭匯第十四輯 : 文康體育 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 14: Culture, Recreation and Sport). Hong Kong: Government Printer 政府印務局 Department of Social Work, The Chinese University of Hong Kong 香港中文大學社會工作學系, comp. and tr. 中譯社會工作學詞匯 (An English-Chinese Glossary of Social Work). Hong Kong: The Chinese University Press Wang, Rui Tian 王銳添, comp. 管理學英漢詞匯 (An English-Chinese Glossary of Management). Hong Kong: The Commercial Press (H.K.) Co. Ltd Zhou, Guozhen 周國珍, tr. 朗文進階英漢雙解詞典 (Longman Active Study English-Chinese Dictionary [edited by Della Summers]). Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第十五輯 : 公眾衛生 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 15: Public Health). Hong Kong: Government Printer 政府印務局 Curriculum Development Council 香港課程發展議會, comp. 中學政府及公共事務科常用英漢辭匯 (An English-Chinese Glossary of Terms Commonly Used in the Teaching of Govermnemt and Public Affairs in Secondary Schools). Hong Kong: Education Department 香港教育署 Department of Sociology, The Chinese University of Hong Kong 香港中文大學社會學系 tr. and ed. 中譯社會學詞匯 (A Glossary of Sociological Terms). Hong Kong: The Chinese University Press Chen, Jiqin 陳繼勤 and Yan Pengfei 顏鵬飛, trs. and comps. 朗文英漢科學圖解詞典 (Longman English-Chinese Illustrated Diction ary of Science [edited by Arthur Godman]), second edition. Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司 Chen, Jiqin 陳繼勤, Zhang Hongde 張宏達 and Ren Zhan-xiang 任善相, trs. and eds. 朗文英漢植物圖解詞典 (Longman English-Chinese Illustrated Dictionary of Botany [edited by Andrew Sugden). Hong Kong: Longman Grou p (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司 Longman Group (Far East) L td. 朗文出版 ( 遠東 ) 有限公司, comp. 朗文袖珍英漢雙解精選辭典 (Longman Handy Learner s English-Chinese Dictionary). H ong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司 N i u, Boshen 紐伯燊 and Li Zuoming 李作明, tr. and ed. 朗文英漢地質圖解詞典 (Longman English-Chinese Illustrated Dictionary of Geology [edited by Alec Watt]). Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司 Chen, Jiqin 陳繼勤, tr. and ed. 朗文英漢化學圖解詞典 (Longman English-Chinese Illustrated Dictionary of Chemistry [edited by Arthur Godman]). Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司 Zhang, Hongyou 章鴻猷, comp. 朗文 - 清華英漢電腦詞匯 (Longman-Qinghua English-Chinese Dictionary of Computer Terms). Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司 Zhang, Xinwei 張信威, tr. 朗文英漢雙解多功能分類辭典 (Longman Lexicon of Contemporary English [compiled by Tom McArthur]). Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司 Chinese Language Division, Hong Kong Government 香港政府中文公事管理局, comp. 政府部門英漢辭匯第十六輯 : 氣象 (An English-Chinese Glossary of Terms Commmonly Used in Government Departments, Volume 16: Meteorology). Hong Kong: Government Printer 政府印務局 Shi, Xiaoshu 石孝殊, tr. (1993). 朗文英漢雙解片語動詞辭典 (Longman English-Chinese Dictionary of Phrasal Verbs [compiled by Rosemary Courtney]). Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司 Lu, David, tr. 朗文實用英語慣用語詞典 (Longman Learner s Dictionary of Idiomatic English [edited by Clare Vickers]). Hong Kong: Longman Group (Asia) Ltd. 朗文亞洲有限公司 Dai, Bocheng 戴柏誠, tr. 朗文英漢物理圖解詞典 (Longman English-Chinese Illustrated Dictionary of Physics [edited by Percy Harrison]). Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司. 165

179 ASIALEX Hao, Yiming 郝以明, tr. 朗文實用醫學辭典: 英漢雙解 (Longman Concise Medical Dictionary: English-Chinese [compiled by Mikel A. Rothenberg and Charles F. Chapman]). Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 亞洲 ) 有限公司 Walker, Peter M. B. and Tsinghua University Press 清華大學出版社, ed. and tr. (1994). 朗文英漢雙解科技大辭典 (Longman English-Chinese Science and Technology Dictionary). Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司 Zhang, Hongda 張宏達, Lin Zhe-fu 林哲甫, and Zhang Wei-qin 張維欽, trs. 朗文英漢雙解生物學辭典 (Henderson s English-Chinese Dictionary of Biological Terms [edited by Lawrence Elearnor). Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司 Chiu, Aman 趙嘉文, ed. 朗文英漢雙解英語成語辭典 (Longman English-Chinese Dictionary of English Idioms). Hong Kong: Longman Asia Co. Ltd. 朗文出版亞洲有限公司 Wu, Yukui 吳毓騤, tr. 朗文英漢雙解商業英語辭典 (Longman English-Chinese Dictionary of Business English [compiled by J.H. Adam]). Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司. 1996b. Longman (Asia) Co. Ltd., ed. 朗文實用護理辭典( 英英.英漢雙解 ) (Longman Concise Nurses Dictionary (English-Chinese) [by Nancy Roper] ). Hong Kong: Longman (Asia) Co. Ltd. 朗文出版亞洲有限公司 Editorial Board, Comprehensive Dictionary of Computing (English-Chinese) 綜合電腦辭典( 英英 英漢雙解 ) 編輯委員會, comp. (1997). 朗文綜合電腦辭典: 英英.英漢雙解 (Comprehensive Dictionary of Computing (English-Chinese)). Hong Kong: Longman Asia Limited 朗文出版亞洲有限公司 and Beijing: Tsinghua University Press 淸華大學出版社 Longman (Asia) Co. Ltd., ed. 朗文國際貿易辭典( 英英.英漢雙解 ) (Longman Dictionary of International Trade (English-Chinese) [by Edward G. Hinkelman]). Hong Kong: Longman (Asia) Co. Ltd. 朗文出版亞洲有限公司 Lu, Guoqiang 陸國強, rev. 朗文當代高級辭典: 英英.英漢雙解 (Longman Dictionary of Contemporary English:English-Chinese). Hong Kong: Longman (Asia) Co. Ltd. 朗文出版亞洲有限公司 Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司, comp. 朗文當代英漢雙解詞典 (Longman English-Chinese Dictionary of Contemporary English). Hong Kong: Longman Group (Far East) Ltd. 朗文出版 ( 遠東 ) 有限公司 Au-yeung, Annie 歐陽因, comp. 朗文中國流行新詞語 (Longman Dictionary of Chinese Popular New Terms). Hong Kong: Addison Wesley Longman China Limited 艾迪生 維斯理 朗文出版社中國有限公司 Shen, Weixian 沈維賢, ed. 朗文十萬辭典 (Longman Dictionary of 100,000 Words). Hong Kong: Pearson Education North Asia Limited 培生敎育出版中國有限公司 Chan, Sin-wai 陳善偉, rev. 朗文進階英漢雙解詞典 (Longman Active Study English-Chinese Dictionary). Hong Kong: Pearson Education North Asia Limited 培生教育版北亞洲有限公司 Dictionary Research Centre, The Commercial Press 商務印書館辭書研究中心, rev. 英華大詞典 ( 最新修訂版 / 繁體字版 ) (A New English-Chinese Dictionary), third revised edition, traditional-character version. Hong Kong: The Commercial Press (HK) Co. Ltd. 商務印書館 ( 香港 ) 有限公司 Lu, Guoqiang 陸國強 and Chan Sin-wai 陳善偉, revs. 朗文當代大辭典( 英英.英漢雙解 ) (Longman Dictionary of English Language and Culture (English-Chinese)). Hong Kong: Pearson Education North Asia Limited 培生敎育出版亞洲有限公司 Wang, Rongpei 汪榕培, rev. 朗文當代高級辭典 ( 英英.英漢雙解 )(Longman Dictionary of Contemporary English (English-Chinese), third edition. Hong Kong: Pearson Education Asia Limited 培生教育出版亞洲有限公司. Bilingual (Chinese-English) dictionaries published in Hong Kong, Goodrich, Chauncey, comp. A Pocket Dictionary, Chinese-English, and Pekingese Syllabary. Shanghai: Kwang Hsueh Publishing House Goodrich, Chauncey, comp. A Pocket Dictionary, Chinese-English, and Pekingese Syllabary. New York: Colombia University Press. 166

180 WORDS IN ASIAN CULTURAL CONTEXTS Lee, S.T. 李仕德, comp. 漢英大辭典 (A New Complete Chinese-English Dictionary). Hong Kong: Chinese Books Publishing Company 中國圖書出版公司 Goodrich, Chauncey, comp. A Pocket Dictionary, Chinese-English, and Pekingese Syllabary. Hong Kong: Hong Kong University Press Zhang, Pengyun 張鵬雲, comp. 最新漢英大辭典 (A New Complete Chinese-English Dictionary). Hong Kong: Great China Book Co Liang, Shiqiu 梁實秋, ed. 最新實用漢英辭典 (A New Practical Chinese-English Dictionary). Hong Kong: The Far East Book Company 遠東圖書公司 Editorial Committee, Guangzhou Chinese Medical Institute, Chinese-English Glossary of Common Terms in Traditional Chinese Medicine 廣州中醫學院 漢英常用中醫詞匯 編寫組, comp. 漢英常用中醫詞匯 (Chinese-English Glossary of Common Terms in Traditional Chinese Medicine). Hong Kong: Joint Publishing Co. (HK) Ltd. 三聯書店 ( 香港 ) 有限公司 Ding, Guangxun 丁光訓, comp. 新漢英詞典 (A New Chinese-English Dictionary). Hong Kong: Joint Publishing Co. (HK) Ltd. 三聯書店 ( 香港 ) 有限公司 Mok, Wing-yin 莫詠賢 and Ngan Yuen-wan 顏婉雲, comps. 傳譯常用語編: 漢英.英漢 (A Glossary of Terms Commonly Used in Interpretation: Chinese-English, English-Chinese). Hong Kong: The Commercial Press (H.K.) Co. Ltd. 商務印書館 ( 香港 ) 有限公司 Duan, Shizhen 段世鎮, comp. 現代漢英詞典 (A Modern Chinese-English Dictionary). Hong Kong: Oxford University Press 牛津大學出版社 Chung Hwa Bookstore, ed. 通用漢英辭典 (General Chinese-English Dictionary). Hong Kong: Chung Hwa Bookstore 中華書局 ( 香港 ) 有限公司 Yao, Naiqiang 姚乃強, ed. Collins Cobuild 高階漢英詞典 (Collins Cobuild Learner s Dictionary: Chinese and English). Hong Kong: The Commercial Press (HK) Co. Ltd. 商務印書館 ( 香港 ) 有限公司. Bilingual (Other language-chinese) dictionaries published in Hong Kong, Brooks, Robert Jones (1961). 最新標準馬英大辭典 (Newest Standard Malay-Chinese-English Dictionary). Hong Kong: Educational Book Co. 教育書店. 1966c. Longman (Asia) Co. Ltd., ed. (1996c). 朗文英漢日彩圖辭典 (Longman English-Chinese -Japanese Photo Dictionary [by Marilyn S. Rosenthal and Daniel B. Freeman] ). Hong Kong: Longman (Asia) Co. Ltd. 朗文出版亞洲有限公司. Trilingual (English-Chinese-Other language) dictionaries published in Hong Kong, Zhong, Nanshan 鍾南山, comp. 英中日圖解醫學辭典 (English-Chinese-Japanese Pictorial Dictionary of Medicine). Hong Kong: Wan Li Book Co. Ltd. 萬里書店 Zhang, Lingzhu 張玲珠 and Zhan Anren 詹安仁, comps. 朗文機電工程詞匯 (Longman Glossary of Mechanical and Electrical Engineering Terms: English-German-Chinese. Chinese-German-English). Hong Kong: Longman Group (Far East) Ltd. 朗文遠東有限公司. Trilingual (Chinese-English-Other language) dictionaries published in Hong Kong, Liu, Fulang 劉福朗, ed. 中英文生物學字典 (A Chinese-English Dictionary of Biology). Hong Kong: Swindon Co Lin, Yutang 林語堂, ed. 林語堂當代漢英詞典 (Lin Yutang s Chinese-English Dictionary of Modern Usage). Hong Kong: The Chinese University of Hong Kong Beijing Language Institute 北京語言學院, comp. 漢英法書報常用詞匯手冊 (A Chinese-English-French Handbook of Chinese Journal Terminology). Hong Kong: Joint Publishing Co. (HK) Ltd. 三聯書店 ( 香港 ) 有限公司 Tamura, Saburo 田村三郎 and Fumiko Shiratori 白鳥富美子, comps. 漢英日化學詞匯 (A Chinese-English-Japanese Glossary of Chemical Terms), revised and enlarged edition. Hong Kong: Joint Publishing Co. (HK) Ltd. 三聯書店 ( 香港 ) 有限公司 Lin, Jianping 林建平 and Zhu Yongqing 朱永慶, comps. 漢英建築工程常用詞匯 (A Chinese-English Common Vocabulary of Civil Engineering). Hong Kong: The Commercial Press (H.K.) Co. Ltd. 商務印書館 ( 香港 ) 有限公司. Monolingual (Chinese) dictionaries published in Hong Kong,

181 ASIALEX Commercial Press, The 商務印書館, comp. 辭源 (An Etymological Dictionary of Chinese), adapted edition, traditional characters. Hong Kong: The Commercial Press 商務印書館 Hsia, Ronald and Peter Penn, comps. Dictionary of Simplified Chinese. Hong Kong: Oriental Book Co Commercial Press, The 商務印書館, comp. Ci Yuan 辭源 (An Etymological Dictionary of Chinese), Vol.2, revised edition. Hong Kong: The Commercial Press 商務印書館 Editing Group, Xinhua Dictionary 新華字典 編纂組, ed. 新華字典 (Xinhua Dictionary). Hong Kong: The Commercial Press (H.K.) Co. Ltd. 商務印書館 ( 香港 ) 有限公司 Dictionary Editing Unit, Institute of Languages, Chinese Academy of Social Sciences 中國社科院語言研究所詞典編輯室, ed. 現代漢語辭典 ( 繁體版 ) (A Dictionary of Contemporary Chinese). Hong Kong: The Commercial Press (HK) Co. Ltd. 商務印書館 ( 香港 ) 有限公司 Editing Unit, A Dictionary of Chinese Characters 漢語大字典 編輯部, ed. 漢語大字典 CD-ROM 繁體單機 2.0 版 (A Dictionary of Chinese Characters in CD-ROM Traditional Character Stand-alone 2.0 Version). Hong Kong: The Commercial Press (HK) Co. Ltd. 商務印書館 ( 香港 ) 有限公司. Monolingual (English) dictionaries published in Hong Kong, a. Longman (Asia) Co. Ltd., ed. 朗文常用英文字根辭典 (Longman Dictionary of English Roots [by Carol Cavness Walker]). Hong Kong: Longman (Asia) Co. Ltd. 朗文亞洲有限公司. Micellaneous, Man, Ssu-ch ien, ed. English-Hakka Dictionary. Hong Kong: JiGuang chuban she. Towards a bilingualized electronic dictionary for second language production Batia Laufer University of Haifa batialau@research.haifa.ac.il Tamar Levitzky-Aviad University of Haifa lytami@hotmail.com Abstract Conventional bilingual L1-L2 dictionaries list translation options for L1 words without explaining the differences between them and without providing sufficient information on how to use each option. We investigated the usefulness of an electronic Hebrew-English-English (L1-L2-L2) mini-dictionary, which was designed for this study, for production. Each entry included L1 (Hebrew) words, their L2 (English) translation options, usage specifications, semantically related English words, and additional L1 meanings for each translation option. Seventy five students of four L2 proficiency levels were asked to translate thirty six sentences from Hebrew into English. Each sentence contained one target word, defined as simple, or complex for translation purposes. Each student used four dictionaries, one dictionary for nine sentences: a Hebrew-English-English electronic dictionary, a Hebrew-English-English paper dictionary, an English-English-Hebrew bilingualized dictionary, and a Hebrew-English bilingual dictionary. Learners were also asked to rate the four dictionaries in terms of usefulness and justify their ratings. In the electronic dictionary condition, all dictionary activity was recorded in log files. We compared the number of correct translation choices in all dictionary conditions, for simple and complex words, at each proficiency level, and analyzed participants dictionary ratings, their lookup preferences 168

182 WORDS IN ASIAN CULTURAL CONTEXTS and the effect of the information looked up on translation correctness. Results demonstrated the superiority of L1-L2-L2 dictionaries and some unique advantages of the electronic format. Introduction Surveys of dictionary use indicate that the majority of foreign language learners prefer bilingual L2-L1 dictionaries and use them mainly to find the meaning of unknown foreign (L2) words (Atkins 1985; Piotrowsky 1989). However, if learners writing in L2 need an L2 word designating a familiar L1 concept, they do not readily turn to an L1-L2 dictionary for help. The reason for this may lie in a serious limitation of most L1-L2 bilingual dictionaries. They rarely differentiate between the possible L2 translations of the L1 word, nor do they provide information regarding the use of each translation option. Because of these limitations, many teachers often discourage their students from using bilingual dictionaries for writing (Rundell 1999; Tomaszczyk 1983). However, if this kind of information, which is essential for using new words, were added, the L1-L2 bilingual dictionaries would enhance rather than hinder foreign language production. With this aim in mind, Laufer (1995) suggested to create a special L1-L2-L2 dictionary for production. An entry in such a dictionary would include three parts: 1. L1-L2 translations, 2. L2 information (definitions, examples, etc.) about each translation option, 3. Thesaurus-like information, i.e. semantically related words to each translation option and additional L1 meanings of the L2 translations. An electronic dictionary can fulfill the above requirements since it can combine the features of an L2-L1bilingual dictionary, an L1-L2 bilingual dictionary and an L2 monolingual dictionary (Nakamoto 1995; Tono 2000; Kernermann 2000; Nesi and Leech 1999). The advent of electronic dictionaries has already inspired research into thei r use and their usefulness as on-line helping tools and as contributors to incidental vocabulary learning. The built in log files can keep track of words looked up, type of dictionary information selected (definition, translation, example, etc.), the number of times each word was looked up, and the time spent on task completion. Some studies compared the effect of different types of glosses (paper, electronic textual, electronic pictorial, electronic and video) on reading comprehension, translation, the number of words looked up, time- on- task and satisfaction of dictionary users (Leffa 1992; Aust, Kelly and Roby 1993, Lomicka 1998, Nesi 1999). Others investigated incidental vocabulary learning via computer glosses (Chun and Plass 1996, Plass, Chun, Mayer and Leutner 1998, Lyman-Hager, Davis, Burnett and Chennault 1993, Laufer 2000, Laufer and Hill 2000, Hill and Laufer 2003, Tono 2003). Our study investigates the use and effectiveness of a special type of electronic dictionary L1-L2-L2 dictionary for production. The study Our research questions were as follows: a. How does an electronic dictionary for production compare with other dictionaries in its effectiveness for correct word usage? b. How does an electronic dictionary compare with other dictionaries regarding students perceptions of a useful production dictionary? c. Which parts of the electronic dictionary entries do students look up when they write and do students lookup patterns affect correct word usage? 169

183 ASIALEX 2005 Seventy five students, of four English proficiency levels participated in the study: 16 University English majors, 19 university students from other departments who were exempted from courses of English as a foreign language, 17 who were not and, at the time of the experiment, were enrolled in an EFL course, and finally, 23 high school students. All students had native or near native competence of Hebrew. All reported being familiar with the computer environment and knew how to use the mouse. Thirty six Hebrew words were chosen as the target words for the study. Twelve words were simple words (Hebrew words with one-to-one correspondence of meaning with their English translation). The other twenty four words were either semantically com plex words (Hebrew words with more than one meaning in English) and/or syntactically complex words (Hebrew words different from their English equivalents in the prepositions they combine with). Four dictionar y types were used in the study: 2 existing dictionaries (English-English-Hebrew bilingualized dictionary and Hebrew-English bilingual dictionary), and two specially designed new mini-dictionaries: a Hebrew-English-English paper dictionary and a Hebrew-English-English electronic dictionary. The latter two dictionaries consisted of specially written entries of the 36 target words. The electronic version was designed in such a way that the participants choices of different kinds of lexical information (definitions, examples, related words, etc) were saved on log files. The participants were asked to translate 36 sentences from Hebrew into English. Each sentence included one of the target words. The rest of the vocabulary in the sentences was simple enough so as not to present a translation problem. Each participant received four worksheets with nine different sentences on each worksheet. Each of the 4 worksheets was provided with a different dictionary to be used for that specific worksheet. The participants were also instructed to use the dictionaries even if they thought they knew the English equivalent of the Hebrew target word. Using a cross balanced design, we created 16 conditions whereby each word was looked up in each of the 4 dictionaries and each student worked with four dictionaries. Since all sessions were conducted with a small number of participants, it was possible to ensure that the participants followed our instructions. On completion of translation tasks, the participants were given a questionnaire asking them to rate the usefulness of the four dictionaries from the least useful to the most useful and to justify their rating. Results and Discussion Each worksheet, representing one dictionary condition, was scored in terms of translation correctness of the nine sentences in it. Thus, each participant received four total correctness scores, one for each condition. In addition, s/he was given one score for the translation of simple words and one for the translation of complex words for each worksheet. We also analyzed the participants subjective rating of the four dictionaries and rating justification. Finally, on the basis of computer log files, we analyzed the types of dictionary information the participants had looked up. All the analyses were performed for the entire sample and per proficiency level. Research question 1 addressed dictionary usefulness. The results of the study showed that in all cases, regardless of the proficiency level, the two new L1-L2-L2 dictionaries yielded better production scores (p<.01 or p<.05) than the other two dictionary types, for all words in general, as well as for the complex words. The significantly lower results obtained with the bilingual dictionary may have been caused 170

184 WORDS IN ASIAN CULTURAL CONTEXTS by the lack of lexical information in this dictionary. Since semantic incongruence quite often exists in any pair of two languages, a li st of L2 equivalents of an L1 word without further lexical information may not be very useful for choosing the correct L2 option. As for the L2-L2-L1 dictionary, when an L2 word is unknown, it cannot be looked up. No wonder, therefore, that this dictionary was not useful for production at all. In the case of simple words, on the other hand, a significant difference was found only between the L2-L2-L1 bilingualized dictionary and the other three, confirming the above statement regarding its ineffectiveness for production. When comparing the new L1-L2-L2 dictionaries with the bilingual dictionary, we found significantly better scores with the new dictionaries only in the high school group and in the entire sample. The lack of significant differences in the other groups may suggest that when a one-to-one correspondence of meaning exists between the L1 and the L2 word forms (as in the case of the simple words in our study), there may be no need for any semantic information differentiating between various L2 translation options and a traditional bilingual dictionary may, therefore, suffice. No significant differences were found between the two versions, the paper and the electronic, of the L1-L2-L2 dictionaries in their effect on correctness scores, suggesting that what mattered more was the type of information provided rather than its format. Research question 2 addressed the learners perceptions of dictionaries usefulness for production. The results of the study clearly showed that in all cases, the participants rated the L1-L2-L2 dictionaries higher than the other two dictionary types, and the bilingual dictionary higher than the bilingualized dictionary. The most commonly provided justification for the higher rating of the L1-L2-L2 dictionaries was that important information was provided (e.g. definitions, examples, prepositions, related words, other meanings). The next most common justification for the high rating of these dictionaries was positive subjective reasons (for example, the dictionary search is quick and saves time). The most commonly provided justification for the lower rating of the bilingual dictionary w as that it lacked important information. The most frequently provided justification for the low rating of the bilingualized dictionary was that its initial point of reference was English. These justifications demonstrate the participa nts awareness of both the importance of detailed lexical information before selectin g and using an appropriate English equivalent and of the importance of the initial point of reference in L1 which enables access to this information. In most cases no differences were found between the participants ratings of the two versions of the new dictionaries, with one exception. In the high school group, participants rated the computerized version as significantly more useful than the paper version, justifying this rating with positive subjective reasons. This justification seems to support previous findings which have demonstrated dictionary users preferences of computerized dictionaries as easier reference tools, enabling a relatively instantaneous access to the entry (Nesi 2000; Nesi and Leech 1999; Tono 2000). Research question 3 addressed the issue of the participants look up patterns, i.e. preferences of certain types of lexical information and the effect of these preferences on translation scores. The log files revealed that different participants chose to look up different combinations of lexical information. This suggests that a good dictionary should include various types of lexical information to cater to a variety of look up possibilities and personal preferences. In three out of four proficiency levels (English majors, EFL students, and high 171

185 ASIALEX 2005 school students) there were positive significant correlations with the translation correctness score when information regarding the use of L2 translation options was included in the search, specifically examples of use and thesaurus information. These correlations provide clear evidence of the benefits of lexical information for written production. This information, which is a special feature of the L1-L2-L2 dictionary, may help to clarify the looked up words by showing how similar or different they are from other semantically related words. Conclusion The study showed that the best dictionaries for L2 written production were the L1-L2-L2 dictionaries. In most of the comparisons we performed (of four types of dictionaries at four language proficiency levels), these dictionaries yielded the best results. Analysis of log files together with learners scores revealed that the reason for the dictionary effectiveness was the lexical information on each translation equivalent of the looked up L1 words. Even though the scores received with the paper version of the L1-L2-L2 dictionary were just as good, the electronic dictionary was viewed more favorably than the paper alternative by more learners. Hence, in terms of usefulness together with user preference, the electronic version fared best. We hope that these encouraging results will motivate lexicographers to produce dictionaries for L2 production, preferably in an electronic form. References Atkins, B.T.S. (1985), Monolingual and bilingual learners dictionaries: a comparison, in R. Ilson (ed.), Dictionaries, lexicography and language learning (Oxford: Pergamon Press and the British Council), pp (ELT Documents 120). Aust, R., Kelley, M.J. and W. Roby (1993), The use of hyper-reference and conventional dictionaries, Educational Technology Research and Development 41(4): Chun, D. M., and J. L Plass (1996), Effects of multimedia annotations on vocabulary acquisition, The Modern Language Journal 80: Hill, M. and B. Laufer (2003), Type of task, time-on-task and electronic dictionaries in incidental vocabulary acquisition, International Review of Applied Linguistics. Kernerman, L. (2000), Pedagogic lexicography: some recent advances and some questions about the future in Ulrich Heid, Stefan Evert, Egbert Lehmann & Christian Rohrer (eds.), Proceedings of the Tenth Euralex International Congress (Stuttgart: Germany), pp (Vol. II) Laufer, B. (1995), A case for semi-bilingual dictionary for production purposes, Kernerman dictionary news, 3. Laufer, B. (2000) Electronic dictionaries and incidental vocabulary acquisition: does technology make a difference? in Heid U, Evert S, Lehmann E, Rohrer, C (eds), EURALEX. (Stuttgart University, Germany), pp Laufer, B. and M. Hill (2000), What lexical information do L2 learners select in a call dictionary and how does it affect word retention?, Language Learning and Technology 3(2): Leffa, V. (1992), Making foreign language texts comprehensible for beginners: an experiment with an electronic glossary, System 21 (1): Lomicka, L. (1998), To gloss or not to gloss: an investigation of reading comprehension online, Language Learning & Technology 1: Lyman-Hager, M, Davis, J.N. Burnett, J and Chennault, R. (1993), Une vie de boy: interactive reading in French, in F.L. Borchardt & E.M.T. Johnson (eds.), Proceedings of the CALICO 1993 annual symposium on assessment (Durham, NC: Duke University), pp Nakamo to, K. (1995), Monolingual or bilingual, that is not the question: the bilingualised dictionary, Kernerman Dictionary News, 2. Nesi, H. (1999), A user s guide to electronic dictionaries for language learners, International Journal of 172

186 WORDS IN ASIAN CULTURAL CONTEXTS Lexicography 12(1): Nesi, H. (2000), Electronic dictionaries in second language vocabulary comprehension and acquisition: the state of the art, in U. Heid (ed.), Proceedings of the Ninth Euralex International Congress (Stuttgart: Germany), pp Nesi, H. and G. Leech (1999), Moving towards perfection: the learners (electronic) dictionary of the future, in H. Thomas & P. Kerstin (eds.), The perfect learners dictionary? (Tübingen: Max Niemeyer Verlag), pp Plass, J. L, Chun, D.M, Mayer, R.E, and D. Leutner (1998), Supporting visual and verbal learning references in a second language multimedia learning environment, Journal of Educational Psychology 90: Piotrowsky, T. (1989), Monolingual and bilingual dictionaries: fundamental differences, in M.L. Tickoo (ed.), Learners Dictionaries: State of the Art. SEAMEO RELC, Singapore. Rundell, M. (1999), Dictionary use in production, International Journal of Lexicography 12(1): Tomaszczyk J. (1983), On bilingual dictionaries: the case for bilingual dictionaries for foreign language Learners, in R.R.K. Hartmann (ed.), Lexicography: principles and practice (Academic Press, London), pp Tono, Y. (2000), On the effects of different types of electronic dictionary interfaces on L2 learners reference behaviour in productive/receptive tasks, in U. Heid, S. Evert, E. Lehmann & C. Rohrer (eds.), EURALEX 2000 Proceedings (Stuttgart, Germany), pp Tono, Y. (2003), The effects of information selection functions in electronic dictionaries and language tasks on L2 vocabulary acquisition, paper presented at the 13th World Congress of Applied Linguistics, Singapore. Teaching language use and communication skills: a bilingual approach Abstract Cynthia Lee Hong Kong Baptist University cfklee@hkbu.edu.hk This paper reports on the syllabus design and teaching effectiveness of a course on bilingualism and sociolinguistics designed for Chinese tertiary students who are not majoring in English in Hong Kong. The subject offers a new perspective to teaching and learning subjects related to language use and communication. The subject adopted a bilingual approach and was co-taught by two teachers (one English and one Chinese teacher), covering and exploring the application of four fundamental linguistic concepts for language use in a variety of daily communication with particular reference to the Hong Kong context: speech act theory, the cooperative principle, politeness principle and conversational implicature in both English and Chinese, which students were required to apply in analysing and composing English and Chinese messages. A myriad of audio-visual materials extracted from newspapers, students writer s projects, films, advertisements and so forth were used to illustrate the concepts, as well as similarities and differences between the two languages. The two teachers took turns to introduce the linguistic concepts and their applications in both languages. At the end of the study, the students were required to select a Chinese and an English audio or visual text and present the ways in which the texts conform to the linguistic concepts and linguistic features of the two languages. The students feedback and some students work will be shown at the end of the paper presentation. 173

187 ASIALEX 2005 Abstract An analysis of polysemous verbs: a cognitive view Keedong Lee Yonsei University kdlee@yonsei.ac.kr Frequently used verbs in English are polysemous. Most dictionaries present the different meanings as separate and unrelated. The purpose of this paper is to show that the meanings of polysemous verbs are related to each other, and the nature of the relation. This analysis will be carried out in the framework of cognitive grammar. One of the basic assumptions of the grammar is that semantic structures derive their value through the imposition of a profile on a base. This assumption will be applied and tested in the analysis of a polysemous verb shut. The verb has a conceptual base, with its participants and their interactions. Some of the major participants are: agent, patients, and participants affected by the shutting process. Profiling different participants in different combinations give us different but related meanings, as the following examples show: (1) a. He shut the door. b. He shut the house. c. He shut his finger. d. He shut the bird in/out. In (1) the verb is used transitively, but the nature of the object is different in each examples. In other words, the objects profiles different participants. In (2) the same verb is used intransitively, profiling only one participant. It is hoped that this kind of analysis will lead to a deeper and better understanding of polysemous verbs on the part of not only learners of English as a second/foreign language, but also linguists interested in dictionary making. Survey of pragmatic information in bilingual English-Chinese dictionaries Saihong Li Beijing Foreign Studies University, China saihongli@hotmail.com or saihong@hum.ku.dk Xiuying Wang Liaoning Police College, Dalian, China xwhz888@sohu.com or wang@hum.ku.dk Abstract: Pragmatics, a relatively new area of linguistics, describes how people use language and reflects the relationship between language and human life. Due to the great difference between the conventions and cultures of languages, learners dictionaries, especially bilingual ones, need to provide users with pragmatic information as to better understand the language. This paper starts with a statistical analysis of 20 randomly selected words from the most frequently used 2000 words in BNC corpus in order to examine how pragmatic information was provided in recently published three bilingualized dictionaries for advanced learners in China: OLAD(2004), LOCDE (2004), COBUILD (2004) and one bilingual dictionary: The New Century Multi-Functional English-Chinese Dictionary (2003) (hereafter NCMFD) Test and questionnaires are also used to coincide with the first part to investigate 1) What is 174

188 WORDS IN ASIAN CULTURAL CONTEXTS pragmatic information in learner s dictionaries? 2) Classification of pragmatic information 3) What types of pragmatic information should be included in these dictionaries? 4) What kinds of user groups need more pragmatic information? 5) How pragmatic information is provided in these bilingual dictionaries and 6) What kind of pragmatic information can we get from corpora? Recommendations of bilingual dictionaries making in respect to English-Chinese dictionaries, such descriptive concepts as pragmatic information provided, which incorporates multifarious techniques of meaning and usage, are given according to the results of the whole study. Introduction Most recent developments in semantics and pragmatics are widely used in lexicographical studies today (Bejoint, 1990; Stensen, 1993; Tomas, 1995; Cock, 2002; Pruvost, 2004; Cowie, 2004). However, John Sinclair claims that this development has to work to fit into the established lexicographic formats or into the whole conceptual background of lexicography. At the same time, the discipline of lexicography strengthens the linguistic theories, especially those concerning language use (Atkins, 1998; Hartmann, 1996). If the first aim of dictionary work is to exhibit the actual variety of usage (Murray1888: x) and capture meaning potentials rather than meaning (Patrick Hanks, 1994), then the integrated pragmatic information is one of the most important components of dictionaries. However, in the past, dictionary compilers have paid little attention to the pragmatic aspect of communication, merely giving unsystematic indications via Usage Labels. (Hartmann 1998: 111) Pragmatics, a relatively new area of linguistics, is meaning in interaction (Thomas 1995:22). It describes how people use language and the relationship between language and human life. Pragmatics foreshadows many of the issues which are of major importance to lexicography today. Due to the great difference between the conventions and cultures of languages, learners dictionaries, especially bilingual ones, need to provide users with pragmatic information so as to better understand the language. As every language teacher knows, some sentences may be grammatically correct, but pragmatically improper. A word has no meaning unless it is placed in context; the illustrative phrases and sentences bring the word to life (Cowie 1995:294). Therefore, the constructed textual examples, or to be specific, pragmatic information such as register, range, and subject field incorporated in meaning are necessary in order to illustrate meaning and the use of words in context (Allan, 2004). Above all, pragmatic information is necessary in learners dictionaries; bilingual ones that can actually be used for decoding purpose are a case in point. Based on Verchurueren s pragmatic perspective, theory of adaption and linguistic meaning theory, this study has the following presumptions: 1. The two levels of linguistic meaning: semantic and pragmatic meaning is integrated into learners dictionaries for advanced learner and bilingual dictionaries; 2. Lexicographers perspective meet users perspective in regard with pragmatic information. Learner s dictionaries Broadly speaking, dictionaries can be divided into general-purpose dictionaries and learners dictionaries. Learners dictionaries are designed specially for foreign 175

189 ASIALEX 2005 language learners. It is also called pedagogical dictionary (Dolezal ed. 1999; Hartmann and James 1998), which aims at non-native speakers of a language. The Advanced Learner s Dictionary of 1948 by A.S. Hornby is regarded as the first learner s dictionary. Ever since then, learner s dictionaries remained unrivalled for about three decades, OLAD 6th, LDOCE 3rd, COBUILD 3rd., CIDE 1995 and Macmillan 2003 have become the representative learner s dictionaries of the 21st century. As instruments for the acquisition of foreign words and of the foreign language in general (Bejoint 1994:74), learners dictionaries (1) normally integrate the learning process into different culture information; (2) use defined vocabularies around 2000 to 3000 words for definition (3) provides many detailed information about usage on how, when and where to use the words, which is irrelevant to native speakers; (4) grammatical specification to combine the merits of encoding and decoding purpose dictionaries; (5) International Phonetic Alphabet is used for pronunciation and (6) pictorial illustrations are used to ensure the definition understandable, thus, learners dictionaries, can serve as the mirror both of the conscious (or unconscious) social values and judgments of its compilers, and of the chief social and ideological concerns of the period of compilation. Here in this paper, both bilingual and bilingualized dictionaries are regarded as learners dictionaries, despite the controversial arguments about its functions in foreign language learning process (Hartmann 1994; Wiegand 1998; Atkin 1995). Pragmatic information in learners dictionaries Pragmatic perspective can serve as an umbrella for the microstructure and macrostructure of dictionaries study and dictionary research. Due to the different perspective of defining pragmatics, it is not surprising to know that pragmatic principles in learners dictionaries vary accordingly. Pragmatic perspectives on Pedagogical Lexicography at the macro-level in Monolingual Dictionary normally provide with extra intercultural information and societal cultural information. In bilingual and bilingualized dictionary, pragmatic equivalent, cross-cultural information, societal cultural information and language error are normally used as a supplement. The linguistic awareness and consciousness of pragmatics, such as Speech Act, Politeness cooperative principles and so on are getting more and more attention from language learners and lexicographers. It can be offensive and insulting to another person with inappropriate use of words. However, a thorough reading of present and past research on pragmatics and lexicography causes me to be cautious in making any claims. There is no broad agreement on what types of pragmatic information and how much and how pragmatic information to include in learners dictionaries. The departure of this paper is to investigate how the main scope of pragmatic studies, such as Speech Act, politeness principles, constructed context are realized in bilingualized dictionaries. According to Rundell (2002), the type of pragmatic information learners dictionaries provide can include Macro-pragmatic Information and Micro-pragmatic Information, such as Speech Act or language functions; Language and politeness; Vague language; Attitude and feelings; Cultural information; Metaphor meaning. In order to provide pragmatic information, the compilers have to describe the typical situation or situations in which a phrase may be used and explain what the phrase means in such situations. Bilingual dictionaries, rarely mentioned linguistic theories, such as pragmatic information, could no longer meet our need. Full equivalents may need mere listing, while partial and surrogate equivalents require further explanation 176

190 WORDS IN ASIAN CULTURAL CONTEXTS or exemplification to ensure sense identification and discrimination (Alan 2004). The fact that have been in English-Chinese dictionary prefaces demonstrates that most lexicographers have tended to confine themselves to collecting examples or more often than not borrowing them directly or indirectly from other dictionaries and classifying meanings on the basis of conventional methods with little or no background linguistic knowledge. A case study of pragmatic information in learner s dictionaries In this part, the author did a case study of four learner s dictionaries: LOCDE, OLAD, Cobuild and NCMD. I chose these four dictionaries because they stood for the present most representative learners dictionaries in Chinese market. In this preliminary case study, 20 randomly selected words from the most frequently used 2000 words in BNC corpus were used to find out where and how pragmatic information is provided in these learners dictionary. The following chart is the statistical results of the study. pragmatic information Definition Equivalents Study Pages or Language notes Examplification usage notes stylistic labels others OLAD COBUILD LDOCE NCMFD The case study showed that each of the contemporary dictionaries has its own strengths and weakness. COBUILD2 made a radical change in definition (contextualized definition). It is entirely based on COBUILD corpus. It integrates pragmatic information into definition in form of a whole sentence and with only geographical labels and style labels in a separate column as a complement. Both Cobuild and OLAD have general introduction on pragmatics. OLAD contains the least pragmatic information compared with the other three. Usage notes on word focus and word choice are the main focus of Pragmatic information. LOCDE and NCASD have similar separate pages on pragmatic information of politeness in: addressing people; apologies; invitations; offers; politeness; requests; tentativeness; thanks; criticism and praise. A preliminary test and questionnaires Experimental tests are one of the most successful methods of observing how large groups of language learners actually employ dictionaries or glossaries. Instead of simply asking people to report on their reference habits and opinions about dictionaries, this technique allows researchers to examine how well dictionaries may assist or mislead them. The questionnaire and interview provide information about what the respondents think they do, and the case study and protocol reveal what they do, but the results of individual dictionary reference acts can best be shown by experimental tests. Having reviewed the advantages and disadvantages of the methods used in dictionary research, I decided to use a test and questionnaires as a supplement for my paper. The experiment test was conducted with 20 overseas scholars and students. Among them, 10 were university teachers or translators from China and 5 were MA 177

191 ASIALEX 2005 degree students in Denmark, 5 were PhD students in Denmark. Each person took the same test, which consisted of 15 multiple choice questions and 10 culture words to test their satisfaction with pragmatic information provided in the above me ntioned four dictionaries. They were allotted 40 minutes for the test and 20 minutes for the questionnaires. The 20 persons were divided into four groups. Each group was given a copy of one of the above mentioned dictionaries. The test had to be conducted four times due to limited space and time. Each time was 5 persons were given the same test with different dictionari es. Short questionnaires were given after the test. Not surprisingly, most of testers had no ideas on pragmatic information in learners dictionaries. In order to do the test, I first explained what pragmatic information in the dictionaries was and where to find them. Data Analysis and Discussion of Results The test results showed that among the 20 users: Group 1 with LOCDE was satisfied with the usage information provided by the dictionary, but not really satisfied with the translation equivalents, such as deteriorate (shi e hua) etc. They couldn t find the right answers simply by reading the translation. Group 2 with COBUILD dictionary was content with the sentence definition, but also agreed that some of the sentences were not easy to understand. Group 3 with OLAD dictionary had no problem finding the right answers for the multiple choice questions, and thought the usage information provided was very useful. Group 4 with NCMFD were not happy with the translation equivalents. The followings are tests and questionnaires results in concerning pragmatic information in particular. Results of cultural information: Group 1 with LOCDE was able to find 80 per cent of the words from the dictionary given and satisfied with the information provided. However, there were 10 per cent of words which they were not sure about the answers from the dictionary and there were 10 per cent of words which they couldn t get any answer from the dictionary. Group 2 with COBUILD was able to find 70 per cent of the words from the dictionary. 30 percent of the words they couldn t find them. Among those words they could find from the dictionary, they were 20 per cent of the words they are not sure about the answer with information given by the dictionary. Group 3 with OLAD was able to find 80 per cent of the words from the dictionary and they thought the information given was clear. However, there were 20 per cent of words they couldn t find from the dictionary. Group 4 with NCMFD, there was 20 per cent of the words they couldn t find the answer. 50 percent of the words they couldn t figure out the meaning from the equivalents given by the dictionary. There was only 20 per cent of equivalents were acceptable. Results of pragmatic information in politeness: Questionnaire results for the pragmatic information in politeness principles by the 20 participants showed that apologies, criticism and praise don t have big difference with Chinese, thus, it was not that necessary for Chinese learners. However, everyone agreed that politeness principles in addressing people, invitations, offers, politeness, 178

192 WORDS IN ASIAN CULTURAL CONTEXTS requests, tentativeness and thanks were very useful, eg. information on Thank you, which is seldom used by Chinese for small things, is good to know. Results of usage notes and stylistic labels: Among these 20 participants, they were all satisfied with the usage notes in OLAD, LDOCE and COBUILD and agreed that it was very useful. Surprisingly enough, NCMFD had no separate column for usage notes. Regardless of the increasing debates on stylistic labels, 20 participants were happy with the stylistic labels and committed that it was very useful. 10 participants thought some of the labels were ambivalent. Results of construed context: Construed context such as collocations and examp les as a supplement to definition were widely used in learner s dictionaries. The results o f the test and questionnaires showed that Group 1 was able to find 60 per cent exemplification; Group 2 was able to read 80 per cent exemplification; Group 3 could consult 70 per cent of exemplification and G roup 4 with 85 per cent of exemplification. Conclusion The preliminary study of pra gmatic information in learner s dictionaries indicated that bilingual dictionaries and bilingualized dictionaries in concerning equivalents provided a re not satisfactory, especially in NCMFD. Therefore, more focus in the future dictionaries are recomm ended to put more emphasis on pragmatic equivalents. Due to the size of the samples, I would not say that the results of the study were very reliable. Further test and questionnaires in large sample will need to be administered. References: Alan, Kirkness (2004), The Handbook of Linguistics (Oxford: Blackwell Publishing) Atkins, B.T. Sue ed. (1998), Studies of Dictionary Use by Language Learners and Translators, in Lexicographica Series Maior 88, (Tübingen: M. Niemeyer) Bejoint, Henri (1990), Culture-specific Items in Bilingual Dictionaries of English, Dictionaries Cowie, P. Anthony (1995), The learner s Dictionary in a Changing Cultural perpective, in B.B. Kachru and H. Kahane (eds.), Cultures, Ideologies and the Dictionary, Lexicographica, series major 64 (Tübingen: M. Niemeyer) Cowie, P. Anthony (2004), Lexiculture and the EFL Dictionary, Kernerman Dictionary newsletter No. 12:6-8 Cock Sylvie De (2002), Pragmatic Prefabs in learner s Dictionaries, Eurolex conference Proceedings Dolezal, Fredric T. and McCreary, Don R. comps. (1999), Pedagogical Lexicography Today, Lexicographica Series Maior 96, (Tübingen: M. Niemeyer) Hartmann, K. K. R(1998), and James, Gregory comps. (1998/2001), Dictionary of Lexicography. (London & New York: Routledge/Taylor and Francis) Murray (1988), The Oxford English Dictionary, (Oxford: Oxford University Press) Patrick, Hanks(1994), `Linguistic Norms and Pragmatic Exploitations in Computational Lexicography: Complex 94: 183 Pruvost, Jean (2004), Some lexicographic Concepts Stemming from a French Training in Lexicology, Kernerman Dictionary newsletter No. 12 Rundell, Micheal (2002), Good old-fashioned Lexicography: Human Judgement and the Limits of Automation, in Correard, M. Lexicography and Natural Language Processing, Eurolex Sinclair, John (1985), lexicographic Evidence, in ELT Documents:

193 ASIALEX 2005 Stevenson, B. (1993), Practical Lexicography: Principles and Methods of Dictionary-making (Oxford: Oxford University Press) Sue & Rundell, Micheal (2004), Lexicography Basis: In Lexicography Master Class Course Notes Thomas, Jenny (1995), Meaning in Interaction: An Introduction to Pragmatics (Harlow: Longman Group Limited) Wiegand, Herbert E. (1998), Wörterbuchforschung, Untersuchungen zur Wörterbuchbenutzung, zur Theorie, Ge-schichte, Kritik und Automatisierung der LexikographieVol. 1, (Berlin: W. de Gruyter) Abstract Do triphthongs exist in words spoken by young educated Singaporeans? Ee Ling Low Siew Siew Lim National Insitute of Education, Nanyang Technological University This paper explores whether young Singaporeans use the five triphthongs found in the speech of British speakers of English. In addition, it will investigate whether morphological and/or orthographical factors play a role in influencing the pronunciation of words which are deemed to contain triphthongs in British English. The working hypothesis is that Singapore English speakers do not use triphthongs at all in their speech as they are likely to insert a glide consonant before the schwa in what would be realised as triphthongs in British English. Data used for investigation include read speech containing the lexical items of interest from 10 Singapore English and 5 British English speakers. Additionally, conversational data from 26 Singapore English speakers are also examined to search for samples from informal speech. Our results support our hypothesis. Both perceptual and acoustic evidence was found for the non-existence of at least 2 triphthongs in read speech and 4 in conversational speech. In addition, it was found that the triphthongs in British English are produced as two syllables instead, where a glide is inserted before a schwa in the second syllable in place of triphthongs. As these glide insertions were employed consistently across the data set, there was no evidence found to support the influence of either morphology or orthography in the pronunciation of these lexical items. Finally, a perceptual test is carried out on non-singaporeans in order to ascertain whether the absence of triphthongs compromises on intelligibility. 180

194 WORDS IN ASIAN CULTURAL CONTEXTS Exit Charlie Chan, enter Allen Choice, an Asian American detective for the 21st Century Calvin McMillin University of Hawai i at Manoa mcmillin@hawaii.edu Abstract Charlie Chan, the very name causes many Asian Americans to cringe. Unlike his predominantly white, mostly male peers in detective fiction, Charlie Chan has come to represent an entire community. Even today, the words Asian detective will likely inspire images of the rotund sleuth. Though not deliberately created with malicious intentions, Charlie Chan has come to embody an Asian American male stereotype, one that pervades even today. While the ethnic detective has boomed since the days of Charlie Chan, the number of Asian American sleuths has been tiny, and the number of those created by Asian Americans has been almost nil. But in 2001 came Leonard Chang s Over the Shoulder, the first in what would become a series of detective novels featuring Allen Choice, a Korean American private eye. This series represents a landmark moment in Asian American literature. While these novels fulfill the requirements of a typical mystery story, Chang straddles the line separating literary and commercial fiction. By crafting a character whose race isn t just a novelty, but a reality, Chang turns Charlie Chan on his head and utilizes the mystery genre to normalize his character in a way heretofore unseen. The author has a vested interest in Asian American issues, but what separates his protagonist from previous minority characters is that Allen Choice has a pronounced disinterest in all things ethnic, racial, or cultural. By crafting a character with such a mindset, Chang strikes the perfect balance with his hybridized amalgamation of both detective and minority fiction. Charlie Chan and His Legacy Ohio-born writer Earl Derr Biggers supposedly based Charlie Chan on real-life Honolulu police detective Chang Apana, a man known for his daring, often unconventional tactics. By placing Chan in a law enforcement role, Biggers hoped to counter the image perpetuated by Sax Rohmer s nefarious Chinese mastermind, Fu Manchu. In his 1982 book The Yellow Peril: Chinese Americans in American Fiction, , William F. Wu calls Chan an example of overcompensation in an author s attempt to break away from the Yellow Peril stereotype (174). But even in making Chan more appealing than Fu Manchu, Biggers created something no less insidious. Michael Cohen (1999) states, A sympathetic minority character does not mean the colonizing stops, if the character is also rendered passionless, unthreatening sexually, or any other way, or otherwise ghettoized (149). Perhaps Chan s problematic character can be best summed up by Josephine Lee (1997), who describes the character as a 181

195 ASIALEX 2005 self-effacing, polite, domesticated Asian who speaks in broken English despite his native-born status, spouts pseudo-confucianisms, and exemplifies loyal service to a white superior (11). Despite whatever positives the character was meant to embody, there is little doubt of its negative influence. But what of the Asian American detectives that came after Charlie Chan? In 1999, authors Gina and Andrew Macdonald drafted a list of ethnic detectives. Chan exempted, the Macdonalds found ten Asian American sleuths. The following table merges their findings with my own and adds occupation as a category: Figure 1. Asian American Detectives Written by Non-Asian Americans Name Ethnicity Occupation Author Lydia Chin Chinese American Private Eye S.J. Rozan Kit Chang Chinese American Student Intern Charles Goodrum Bob Nakamura Japanese American Unavailable John Ball Chip Nguyen Vietnamese American Private Eye Dean Koontz Masao Masuto Japanese American Beverly Hills P.D. Howard Fast Reiko Masada Japanese American Unavailable Ray Gilligan Sam Ohara Japanese American L.A.P.D. Nan Hamilton John Rain Japanese/Caucasian American Assassin Barry Eisler Rei Shimura Japanese American Antiques Dealer Sujata Massey Josef Tanaka Japanese/Caucasian American Forensic Pathologist Richard LaPlante Mark Shigeta Japanese American FBI Agent Anne Wingate Tina Taniko Japanese American L.A.P.D. Paul Bishop April Woo Chinese American NYPD Leslie Glass Relatively speaking, the list is somewhat impressive, but upon further investigation, there is a sense of murkiness here in terms of representation. Rather than embody a kind of rugged individualism, several of these characters are compromised, either saddled with a non-asian partner or given amateur sleuth status. On this finding alone, one wonders if this has anything to do with the fact that none of these characters were created by a person of Asian descent. That isn t to suggest that only people of a race can write about that culture, but rather an acknowledgement of the noticeable absence of Asian American-themed mysteries written by actual Asian Americans. Perhaps the earliest examples of North American-based detective stories with protagonists and creators of Asian descent would be Yellow Fever and Chan is Missing, both from R.A. Shiomi s comical play Yellow Fever puts an all-asian cast in a typical Bogie and Bacall-style noir plot while Wayne Wang s film Chan is Missing centers on two Chinatown cabbies as they search for the title character, who may have absconded with their hard-earned savings. In literature, Chang-Rae Lee s 1995 novel Native Speaker is often cited as an ethnic detective story. But while Lee s protagonist is a Korean American spy, the novel is not a commercial thriller, but instead, utilizes the language of espionage as a metaphor for Asian American identity. In 1996 came Dale Furutani s novel Death in Little Tokyo, which features Ken Tanaka, an unemployed computer operator turned sleuth, who finds himself entangled in an international smuggling scheme. Followed by one sequel, 1997 s Toyotomi Blades, the novel went on to win several awards, making Furutani the first Asian American to win a major mystery prize. And finally, China Boy author Gus Lee penned two legal thrillers, 1996 s Tiger s Tail and 1998 s No Physical Evidence. The former centers on Jackson Kan, a Chinese 182

196 WORDS IN ASIAN CULTURAL CONTEXTS American JAG lawyer sent to the DMZ to look for a missing friend. No Physical Evidence is set in Sacramento and revolves around Joshua Jin, a district attorney assigned to prosecute a rape case he seemingly has no hope of winning. Although none of these novels depict a detective in the traditional sense, the latter three were marketed as mystery thrillers, a fact which illuminates the kinds of commercial work Asian Americans were writing prior to Leonard Chang s foray into detective fiction. The Literary Detective Novel The idea of a literary detective novel isn t a new one, but the possibility that it could exist didn t seem to occur to anyone until the mid-twentieth century. In The Guilty Vicarage, W.H. Auden (1954) wrote of his love for detective stories, but added the caveat that they did not qualify as art. However, he viewed the work of Big Sleep author Raymond Chandler, as serious studies of a criminal milieu, the Great Wrong Place, and his powerful but extremely depressing books should be read and judged, not as escape literature, but as works of art (151). Although Auden was stuck on a semantic poin t, here one can see the beginnings of a dramatic shift in thought about detective fiction. In truth, Chandler had grown sick of the puzzle box nature of mysteries. As documented in the posthumously-released Raymond Chandler Speaking, the author admits, I didn t care whether the mystery was fairly obvious, but I cared about the people, about this strange corrupt world we live in and how any man who tries to be hon est looks in the end either sentimental or plain foolish (Gardiner 1984: 233). Post-Raymond Chandler, the detective story has become a frame authors use to explore and comment on the world around them. As Leonard Chang told the L.A. Daily News: There s a reevaluation going on: those pulp writers are suddenly recognized as artists in their own right, working miracles within narrow parameters (Murphy 2001). In 1944 s The Simple Art of Murder, Raymond Chandler writes of the detective, He is the hero, he is everything. He must be a complete man and a common man and yet an unusual man ( ). Thus, the literary detective becomes much more than just a human decoder ring meant only to solve puzzles, a paradigm shift which allows issues like race and ethnicity to be more than just cultural novelties used to spice up an otherwise formulaic plot. In other words, the stage has been set for Allen Choice s arrival. Allen Choice, Detective Allen Choice is an Asian American hero for a new generation. His unique surname serves as a direct reflection of the American immigrant experience. In Over the Shoulder, Allen reveals its origin: Once immigrating here, my father Americanized his last name, Choi, by going through his English dictionary and looking for the word with the closest spelling (Chang 2001: 22). The name evokes any number of interpretations: the freedom to choose a name embodied in the name itself and the difficult choice immigrants must make between cultures are just a few examples. But the idea of choice in this context is most closely associated with notions about the American dream, the idea of remaking (choosing) one s own identity in a land of opportunity: the Horatio Alger story, the rags to riches tale, the small town boy makes good myth, and perhaps most famously, F. Scott Fitzgerald s The Great Gatsby. As Chang explained to the L.A. Daily News, the Allen Choice stories aren t designed to solely represent Korean Americans, but are instead the experience of anyone who is different by color, size, 183

197 ASIALEX 2005 speech anything (Murphy 2001). At the time of Over the Shoulder, Allen works in security as a kind of bodyguard. In the sequel, he s waiting to the take the P.I. exam, and by the third novel he s a full-fledged private eye. Pegged as the strong, silent type, Allen s physicality is hinted at through his nickname The Block. As he explains in 2003 s Underkill, The reason why I was called The Block in high school was because of my build, which some kids thought looked like a square building block, and I played fullback on our soccer team, successfully defending and blocking our end. (Chang 29) Although a minor detail, the fact that Allen played competitive sports in school undermines stereotypes about nerdy Asian American males. Choice also has no problem attracting women. However, his sexual escapades are never of the bed-hopping, James Bond variety. He sleeps with three women in the series Sonia Baumgartner, Linda Maldonado, and Serena Yew two of whom become relatively long-term girlfriends. In making Allen Choice a sexual being, Chang again defies the emasculation stereotypes that have dogged Asian American males in popular culture. And in making the relationships both sexual and serious, Allen is granted the kind of personal life that Charlie Chan simply wasn t allowed to have. And unlike his predecessor, Allen falls more into the hardboiled detective mold of Sam Spade and Philip Marlowe than into the role of the traditional puzzle-solving sleuth. And here the distinction matters. In Maxine Hong Kingston s Tripmaster Monkey, protagonist Wittman Ah Sing remarks that Charlie Chan never caught a criminal by fistfighting him. And he doesn t grab his client-in-distress and kiss her hard, pressing her boobs against his gun (1989: 320). It s a comically astute observation, one that highlights Chan s weak, ineffectual image. As William F. Wu explains, Chan maintains an apologetic manner, and at no time in this first novel does Biggers allow any hint of aggression, assertiveness, or temper to show in his detective (1982: 176). Allen, however, is the kind of detective who plays hunches, who gets his hands dirty, and yes, even gets into fistfights much in the spirit of the best private eye characters of the last century. Aligned firmly with the hardboiled tradition, Allen Choice is a detective who must negotiate his way in a version of Chandler s Great Wrong Place, a man prepared to walk those mean streets that are dark with something more than night. The Race Card Early American multicultural detectives have been predominantly faux exotic, a kind of stereotyped exploitation of the exotic rather than serious explorations of the Other (Macdonald 1999: 60). Charlie Chan was noted for spouting supposed Chinese aphorisms, and as William F. Wu explains, The sayings emphasize the foreign origin of the character, not his trials and other experiences as a member of an American minority (1982: 179). Even with the emergence of more politically sensitive writers, there is still a tendency to make the minority protagonist little more than an old school detective in ethnic drag. But Chang avoids what Dean Koontz spoofs in his novel TickTock. Koontz s Vietnamese American private eye, Chip Nguyen is said to parody authors who believe that ethnic names and a few references to special food, language, or customs really make a detective multicultural (Macdonald 1999: 69). What sets Allen Choice apart is his lack of knowledge about culture. He cannot speak Korean and knows very little about his heritage. In Allen Choice, we have a character more prone to quote Kierkegaard than Confucius. Certainly, race is an issue in the Allen Choice novels, but 184

198 WORDS IN ASIAN CULTURAL CONTEXTS refreshingly and rather realistically it is not the only issue. In Korean Quarterly, Chang states: One of the reasons I m doing a mystery series is because it s a way to write about a Korean American character for whom race is not the sole function of his life (Vickery 2003). Unlike many other Asian American protagonists, whose sense of Asian-ness Luke later apologizes, Allen asks him seems to dominate their concept of self from page one, Allen Choice is not preoccupied with issues of racial identity. It would be inaccurate to say Allen is Americanized because that would imply that he wasn t an American from the beginning. Furthermore, Allen s lack of connection with his Korean heritage isn t due to an embrace of the dominant American culture he s alienated from both. This feeling relates to his ideas about relationships in general, since Allen believes that the bonds of family are illusory, that just because you are related to someone doesn t mean you are really connected (Chang 2004: 71). Through his protagonist, Leonard Chang questions the assumptions implicit in the words culture or heritage and seems to be proposing a more existential view towards racial identity. To borrow the text s phrasing: just because you are the same race as someone doesn t mean you are really connected. A further example of this disconnect is evident in Allen s inability to speak Korean. In Underkill, he meets Linda s stepfather, Luke Sherwin, who greets him in Korean, causing Allen much discomfort. When about his surname. Luke reveals the name is Polish in origin, but shortened to be more American-sounding. Allen soon notices an inconsistency in Luke s assumptions, but chooses not to point it out: He had expected me to speak Korean, whereas neither he nor his father spoke Polish (Chang 2003: 33). For Allen, Korean heritage should no more indicate fluency in Korean than Polish heritage should indicate fluency in Polish. Chang highlights the double standard in contemporary American culture in which an Asian American who doesn t know his or her native language is made to feel like an ethnic dunce, whereas the English-only Caucasian is given an unspoken free pass. But it s not just mere racial misunderstandings or stereotypical assumptions that Allen must face. In several instances, he must also confront racism head on. In Over the Shoulder, Allen s uncomfortable encounter with a racist precipitates his partner s murder. In the same book, he s referred to as a gook and Bruce Lee. Things don t get much better in the later novels, where he s derogatorily referred to as Papa-San and even called a chink. While Charlie Chan, too, was often the victim of racism, in most cases, he would patiently deal with the situation by winning over his critics. As William F. Wu writes, Charlie Chan s calm, apologetic, and passive tolerance of racial insults and harassment is an obvious sop to those who would be threatened by an Asian American detective with normal assertiveness and temper (1982: 181). Allen Choice, however, neither tolerates racists, nor attempts to win their respect. And although he does use these racist incidents as a time for reflection, Allen doesn t dwell on these experiences for long. There is, after all, a mystery to be solved. Case closed? Eighty years have passed since the publication of Earl Derr Biggers House Without a Key and yet Charlie Chan still hasn t gone away, his figure looming large over the Asian American pop cultural landscape. But with Leonard Chang s emergence onto the detective fiction scene, one is encouraged by the prospect that the ghost of Charlie Chan may finally be exorcized. The very existence of Chang s work has set the stage for an emerging genre, as evidenced in the wave of Asian American mysteries in the last few years: Don Lee s Country of Origin, Suki Kim s The Interpreter, and Naomi 185

199 ASIALEX 2005 Hirahara s The Summer of the Big Bachi and its forthcoming sequel, Gasa-Gasa Girl being the prime examples. Whether future Asian Americans writers will find the mystery genre to be a compelling avenue of expression for their own concerns is impossible to predict. But if Leonard Chang s novels serve as any indication, the choice is clear. References Auden, W.H. (1956), The Guilty Vicarage, The Dyer s Hand and Other Essays. (New York: Random House, pp Chandler, Raymond (1950), Introduction to The Simple Art of Murder, in Frank MacShane (ed), Raymond Chandler: Later Novels and Other Writings (New York: Library of America,), pp (1997) Raymond Chandler Speaking. Ed. Dorothy Gardiner, Paul Skenazy, and Katherine Sorley Walker. (Berkeley: University of California Press) (1944) The Simple Art of Murder, in Frank MacShane (ed), Raymond Chandler: Later Novels and Other Writings (New York: Library of America), pp Chang, Leonard (2001), Over the Shoulder, (New York: Ecco/HarperCollins) (2003) Underkill, (New York: St. Martin s Minotaur/Dunne) (2003), Why I Love Crime Fiction, The Thrilling Detective, available online, (accessed: 9 November 2004) (2004), Fade to Clear, (New York: St. Martin s Minotaur/Dunne). Cohen, Michael (1999), The Detective as Other: The Detective Versus the Other, in Kathleen Gregory Klein (ed), Diversity and Detective Fiction. (Ohio: Bowling Green State University Popular Press). Lee, Josephine, (1997) Performing Asian America: Race and Ethnicity on the Contemporary Stage. (Philadelphia: Temple University Press). MacDonald, Gina and Andrew (1999). Ethnic Detectives in Popular Fiction: New Directions for an American Genre, in Kathleen Gregory Klein (ed), Diversity and Detective Fiction. (Ohio: Bowling Green State University Popular Press). Murphy, Bernadette (20 February 2001), Chang s Latest: Not What Readers Expect? Los Angeles Daily News, available online, (accessed: 8 November 2004). Vickery, M artha. (2003), The Inner Detective: A Portrait of Mystery Author Leonard Chang. Korean Quarterly, available online, (accessed: 7 Novemb er 2004). Wu, Willi am F. (1982), The Yellow Peril: Chinese Americans in American Fiction (Connecticut: Archon Books). Compiling Russian English lexicon: from terms to culture Alexei Medvedev University of Adelaide medvedev@student.adelaide.edu.au Abstract English as a worldwide language permanently borrows lexical items from other languages. The aim here is to encompass a borrowed from Russian lexicon which is presumably already used - or not! - in everyday (Australian) English, and to enhance 186

200 WORDS IN ASIAN CULTURAL CONTEXTS the awareness of English speakers in Russian culture by additional explication. For data collection, different sources are used: media, Internet, literature, dictionaries, personal conversations, etc. The selection criteria are Russian origin, usage without simultaneous interpreting and translation, frequency, comprehension by English speakers, cultural and informative representation, and personal intuition. It confines three dozen words, such as sputnik, troika, vodka, KGB, etc. Etymological analysis provides their primordial meanings. It allows defining their historical semantic shifts and thus some main trends in Russian society. Reconstructing a particular socio-cultural context can be relevant to the whole lexicon because such a pragmatic approach reveals additional conceptional links between the terms and establishes a unifying image of Russia at a certain historical stage and then throughout her history. Surprisingly, the relatively small lexicon, even sporadically collected, appears to be representative enough for expressing the Russian idea, the idea of ultimate sacrifice by the individual for the common sake. Thus, an initially lexicographical inquiry transforms into a multifaceted cultural endeavour. Though once accepted, these Russian words have not yet integrated into English. They still are quite odd for ordinary English speakers. Nevertheless, they play an important communicative role, while filling out the gaps in the English semantic continuum in terms of Russian cultural and social peculiarities. Aim The aim of the web-site is to give clarification to the meaning of some Russian words, which are used in English speech often without translation, and illuminate for readers the integrated image of Russian culture through these fresh for English Russian borrowings. Commitment Given that we are living in a rapidly changing environment, natural and social, we are obliged to seek a better understanding of ourselves and the others in society and nature. We all are responsible for mutual misunderstanding and disorder in the world, and have to try to overcome it, even in something little. Act locally but think globally! should be everybody s commitment and responsibility. My personal commitment is to assist English speaking people in better understanding of Russia and her people, meanwhile, trying altogether to find any solution to Russians numerous problems. Communication The people s world has become globally intertwined, and people are intensely communicating with each other. Usually, it is understood as exchange of some information only. Ironically, etymological analysis shows that communication originally designates common responsibilities (from the Latin root -mun- which means responsibility), and when communicating to each other, we actually should share them. 187

201 ASIALEX 2005 Sincereness is a highly appreciated quality and virtue amongst Russian people, and it makes communication genuine in a philosophical sense. Cultural gravity Having set up at the beginning an apparently pure linguistic and informatics goals, I then unexpectedly realised that the adequate clarification of meaning needs some cultural and historical context to be characterised. By the end, cultural aspects have taken the first place in the inquiry. Having been compiled as a lexicon, the words, like a mosaic to some extent represent the whole image of the bearing them culture, and every word is a carrier of the meaning of a part of the culture. Having been borrowed by other culture, the words unite both cultures to some extent and accomplish communication. To strengthen the cultural context, I have introduced into the lexicon some personal names ( Alyosha and Katya ), some historical toponyms ( Kremlin, Leningrad, Stalingrad ), and some notorious concepts ( Russian roulette and Russian mafia ), because they convey the Russian culture, while becoming her symbols. Data Here the primary data is the Russian words, which are used in English speech, often without translation. I have been memorising such words for about six years, in , since I arrived in Australia. The words have sporadically appeared in media, mostly in documentaries, on Australian TV channels ABC and SBS. Some of them occurred only once, for example, Alyosha was used in the documentary about a men s chorus from Northern Norway. Discourse The content of the web-site may be considered a personal discourse, because I expose my personal understanding of the words, and my own judgment on the recent Russian history. I tell stories about Russian people, events, and customs. Focus The inquiry is focused on some Russian cultural and political trends, and their explanation to English speaking people. It also can be useful for students, who study English at an advanced level. Impartiality To reach impartiality seems to be impossible in such a case as the turbulent and controversial Russian history. The researcher, at least implicitly, takes one side of a conflict, according to their political views. Hence, I openly take the side of ordinary Soviet people, Russians and of other ethnicities, which have been sharing during the 20th century a lot of common suffering due to big and small wars, revolutions, and inappropriate policies of the ruling classes. The clearly articulated subjectivity can only enhance the objectivity of the inquiry. Moreover, I provide URL, where other points of view are represented. It should facilitate a multidimensional pluralistic understanding of the matter. 188

202 WORDS IN ASIAN CULTURAL CONTEXTS Involvement It is the insight of a person, who is familiar with the described phenomena, who was an insider of the explicated political system, and who has always been concerned with these issues through his life. I have seemed to be a non-man rather than a yes -man under both socialism and capitalism. I would prefer something in the middle. Starting the project, I considered it boring, but later I was reminded of and rethought my past experience and skills, and I shall probably set my own web-site in hope that it would be helpful and interesting to some. Humour People could not have survived without humour, especially, under such complicated living conditions as it was in Russia. Therefore, I try to transmit this flavour to the mix of sadness and joy, scepticism and hope, the gap between the great expectations and reality. Language and Lexicon This is initially a linguistic investigation of a particular segment of everyday language in the form of a dictionary. It does not pretend in any way to be inclusive and comprehensive. It is a small lexicon rather then a full vocabulary. Having once been accepted, these Russian words have not become yet completely integrated into English; they are still quite unusual for ordinary English speakers. Nevertheless, they play a significant communicative role, while filling out the gaps in the English semantic continuum in relation to Russian cultural and political peculiarities. Only time can define by use what words will survive in English in the future. The forthcoming items: Agitprom, cosmonaut, Cosmos, Mir, pogrom, samovar, refusnik, Stakhanovism, steppe. Methodology Alyosha Kalashnikov perestroika apparatchik Kalinka Russian mafia bistro Katyusha Russian personal names Bolshevik KGB Russian roulette Bolshoi kolkhoz Samizdat borsch Kremlin Soviet dacha kulak Soyuz glasnost Leningrad sputnik Gorbi Molotov cocktail Stalingrad GULAG NKVD troika intelligentsia nomenclatura tsar Ivan otkaznik vodka For compilation of the lexicon, some particular methods have been used at different stages: 189

203 ASIALEX 2005 (1) Selection of words on the basis of their Russian origin from everyday English in the media. (2) Etymological analysis of the words to define their original intention, or primordial meaning. (3) Discovery of the contemporary meanings of the words to reveal possible semantic and sometimes phonetic shifts. (4) Exploration of the socio-cultural context of the words use, and getting new eventual meanings of socio-cultural significance. (5) Disclosure of lexico-semantic relationships (links) between the different words and concepts. (6) Unification of some terms into more abstract concepts. (7) Synthesis of the integrated image of Russia and her culture, and formulation of the Russian idea, the basic trend in her socio-cultural development. Surprisingly, the words, having been collected sporadically and often independently from each other, can define the multidimensional image of Russia. Stages 1-3 may be considered socio-linguistic analysis. Stages 5-7 actually become socio-cultural synthesis. Meanwhile, stage 4 serves as a transition from analysis to synthesis through socio-cultural context. (8) Rethinking and correction, if it is necessary, of the previously gained meanings of the words, taking into account their particular role in the Russian semantic continuum. Stage 8 is final and it may be concluded that all the suggested methodology that define one total philosophical method, thereby revealing progress of the meaning from the concrete through the abstract to the concrete again, moving forward like by a spiral. Starting as linguistic analysis of particular words through working out of the lexicon of Russian words in modern English the inquiry became a kind of everyday culturological investigation. For a better understanding of the meaning of selected words, methods of sensual associations are applied combined with visual, and acoustic (sound) means that enhances the acquisition of terms. Therefore, a deep investigation of language gives great opportunities as a whole for socio-cultural analysis in addition, and vice versa. Novelty I do not pretend to possess any particular recognition of novelty in this inquiry. It is merely a compilation of already existing concepts, and formulation some eventual principles, which could have underpinned them. My role is only to reveal them. Probably, the invasion of Russian female personal names is quite a new phenomenon. It would be interesting to look at it, for example, from feminist perspective and to compare women s and men s attitudes to femininity of different cultures. I also raise the question of how a personal name does gain particular cultural meaning, often implicitly and subconsciously. It requires further investigation. Sources The Internet provides the main source for illustration of the collected words, and to create as much as possible their cultural and historical context. Very often, it is 190

204 WORDS IN ASIAN CULTURAL CONTEXTS impossible to indicate precisely most of the sources, because the words are mostly well-known to Russians and belong to their everyday language. Nevertheless, some URL are provided to the audience for access to more information if it is necessary. Structure The project is made in Microsoft Front Page format and has thirty eight pages, with hyperlinks between the pages and relevant Internet sides. The main (Home) page is connected with all other pages allowing to reach each of them through it. This consists of the table of INDEX, every cell of which contains a particular name, and by clicking on it, the user can get a required page with a relevant description. An alternative method is passing through the package, page by page, from Home page, with INDEX, up to COMMENTARY page, which is supplementary to the INDEX and the other thirty six pages with the words. Technology The invention of modern information technologies opens new possibilities for an integrated, systematic approach to language studies. The Internet facilities, hypertext linking, and some multimedia elements (pictures, charts, sound tracks, and video files) have been applied here. It allows connecting the audience with various files, within the package and out of it, on the Internet, enhancing significantly teaching facilities. Test and Evaluation The process of testing and evaluation has almost been permanent and ongoing. First, it was found that I understood the Insert operation as copying and cut off the original files from the project folder, and lost most of the images, to repair the damage required then a lot of time and additional efforts. While working out the content of the project, I found that the dimension of the task was constantly increasing involving more and more persons, events, and facts to explain all nuances of meaning. On the other hand, I found that the words through the same socio-cultural context are more intertwined than I had expected before. Therefore, the need for external growth could partly be compensated for by more intensive insight to the matter. During the preparation of the lexicon, my computer skills were growing too, and it acted upon my direction: the more multi-media means I applied the better I could solve this problem of extension. My pages became more complicated, and sophisticated, the number of links was increasing, and meanwhile, I started to gain a more integrated and unified perception of Russia as a whole. The search for relevant sites on the Internet shows that there is not enough well-elaborated programs for Russian educational purposes on politics, sociology, philosophy, except for, maybe, literature and history. Many sites are simply poor, without multi-media facilities, often with bare texts. The modern multi-media computer technologies are obviously not properly used, and that urgently requires much effort from the teaching staff. 191

205 ASIALEX 2005 At the beginning, the tutorial task seemed to me to be unstimulating, routine, and even boring. But through accomplishing it, I have realized that it has become enormously fascinating, creative, and actually unlimited for self-expression, and there is a real opportunity to convert it into a personal web-site. Web as Corpora Nowadays, the web of the Internet has actually become the biggest collection of the hugest corpora of information. None of the reproduced texts, chats, pictures, videos and sounds originate from me. Thus, I express my deep feelings of appreciation and gratitude to all the authors of these visual, textual, and acoustic items. To indicate all of them seems to be impossible, but I try to give their relevant web-sites on every page. English learners dictionaries: An undervalued resource Julia Miller Flinders University, Adelaide, South Australia Julia.Miller@flinders.edu.au Abstract Many students are unaware of the value of learners dictionaries and feel that bilingual dictionaries offer a quick and easy translation of unknown vocabulary. Nevertheless, they frequently find that bilingual dictionaries do not provide enough information to allow them to use new words productively. By undertaking a vocabulary exercise using either bilingual dictionaries or monolingual learners dictionaries, ESL students at an Australian university were able to compare the two resources and evaluate the contribution each made to their use of unfamiliar English vocabulary and the potential contribution that learners dictionaries could make to their use of academic English. Introduction Most ESL students will assert that the prerequisites for tertiary study in Australia are an IELTS certificate, a theoretical knowledge of the English language and... a bilingual electronic dictionary. In practice, however, the IELTS test is not always an accurate indicator of academic success (Feast 2002), English knowledge is often incomplete, and recourse to a grammar book or dictionary is frequently necessary. Most students therefore rely on their bilingual dictionaries, which provide a quick and easy translation, often suitable for what Béjoint terms decoding, but inadequate for the intricacies of encoding (Béjoint 1981, p. 210). If students were aware of the range of information provided by a monolingual English learners dictionary, they might be able to use the dictionary to improve their vocabulary, grammar and cultural knowledge, and thereby enhance their academic writing skills. This pilot study was undertaken in order to determine if students in an Australian university context were sufficiently aware of the contents of learners dictionaries and whether, being aware, they would make 192

206 WORDS IN ASIAN CULTURAL CONTEXTS more use of such dictionaries in their future searches. Literature review Previous studies suggest that learners do not fully understand the range of resources offered by their dictionaries. Lemmens (1996, p. 75) indicates that of the 112 translation students in his study, 46.5% did not know what grammatical information could be found in the bilingual dictionaries which they used regularly. Indeed, many students assume that they already know how to use dictionaries and do not take the trouble to investigate further (Bensoussan, Sim & Weiss 1984, p. 271). There is also a general reluctance among students to use learners dictionaries rather than bilingual dictionaries. Atkins and Varantola (1998, pp. 38 & 111) and Bensoussan, Sim and Weiss (1984, p. 267) found that, even when shown the advantages of using a learners dictionary, the majority of students still preferred to use their bilingual dictionary. In some cases this was for pragmatic reasons, the student s bilingual electronic dictionary proving easier to transport than a more cumbersome hard copy. Even training in dictionary use, however, did not increase students desire to use a learners dictionary (Atkins and Varantola 1998, p. 36). Bogaards (1995, in Hulstijn & Atkins 1998, p. 10) suggests that this reluctance on the part of the students may be due to ignorance of correct and effective dictionary skills. One problem outlined by Cowie (1980, in Béjoint 1981, p. 208) is that learners dictionaries demand reference skills beyond the capabilities of many of their users. The coding systems used by dictionaries vary in complexity and depth, making it difficult for students to use them effectively and to switch easily between dictionaries. B éjoint s study revealed that 89% of those surveyed had not studied their dictionary s in troductory matter, and 55% did not use the coded information on word usage (Béjoint 1981, p. 216). Béjoint s conclusion, like that of Atkins and Varantola (1998, p. 115), was that students should be guided in their use of monolingual learners dictionaries. The type of learners dictionary used (whether Oxford, Cambridge, Longman or Collins Cobuild) seems to have little effect on users accuracy, according to Nesi (1998, p. 159). Students are more often hindered due to their own poor knowledge of g rammar. In Nesi and Haill s study (2002), students often failed to identify the word class of the look-up word (p. 282), confusing nouns, verbs and adjectives (p. 283). Bensoussan, Sim and Weiss (1984) found that most students ignored the part of speech of a given word (p. 270) and felt that they merely [needed] to look up words in order to understand the text (p. 269), leading to the conclusion that teachers need to make students more grammatically aware when using dictionaries (p. 271). The above studies were conducted in European and Israeli universities, with predominantly European or Israeli students. In Australia, the majority of international university students come from Asian countries. The study which follows aimed to discover whether these students knew what was in an English learners dictionary, and how this knowledge affected their view of dictionaries in general. It was confirmed that students in Australia, like their counterparts in Europe, do need to be made aware of the learners dictionary s contents and be trained in effective dictionary use. Participants Forty-two students participated in the study. They were all members of an Australian university ESL class and were in their first or second year of tertiary study. Forty of the students were undergraduates; the remaining two were postgraduates. All 193

207 ASIALEX 2005 had a minimum global IELTS score of 6 and had been in the ESL class for three months, receiving instruction in academic writing skills, vocabulary and grammar. There was an almost equal number of male and female students (M=23, F=19), most of them aged in their early twenties. Thirty-one of the participants spoke Chinese as a first language. The other participants were L1 speakers of Amharic (1), Arabic (2), Japanese (3), Khmer (1), Portuguese (2), Swedish (1) and Tagalog (1). Method The ESL students were taught in three separate groups. In this study, after a brief discussion of the different kinds of dictionaries available (bilingual, electronic, monolingual and learners ), the students were introduced to a selection of learners dictionaries. These were limited to the hard copies available in the Student Learning Centre and comprised: Collins COBUILD (1987, 1995 and 1996 editions); Cambridge Advanced Learner s Dictionary (2003); Cambridge Learner s Dictionary (2nd edition) (2004); and Longman Dictionary of Contemporary English (2003). The coding information was briefly explained. The students were then asked to use either a learners dictionary or a bilingual dictionary (their own or that of a classmate) and complete a gapfill exercise, using ten words to complete ten sentences. Students were reminded to fit the words into the sentences grammatically. Each word was to be used once only. The words used were: accentuate, discrepancy, elaborate, elucidate, epitome, euphemistically, idiosyncratically, myriad, panacea and paucity. These words were chosen from a list devised by Nesi and Haill (2002) of the vocabulary most commonly used by English-speaking lecturers, and were felt to be those that students might not know but which would be useful to them in their studies. As they completed the gapfill exercise, students were also required to fill in a chart which listed the ten words and asked for details of word class (adjective/adverb/noun/verb); context (formal/informal/scientific); collocation (eg prepositions); and pronunciation. After finishing the exercise, the participants were invited to complete a questionnaire on dictionary usage (based on Nesi and Haill 2002) designed to investigate the students prior knowledge of dictionaries, the ease with which they had completed the exercise and the attitude which they now had to learners dictionaries. Findings The questionnaire revealed that each student owned on average two dictionaries. It is not clear, however, whether the students interpreted own as possess or have on the desk before me now, since several who claimed to own a learners dictionary said they had never used one before the exercise. (This may perhaps reflect the fact that they had a dictionary but did not use it). Nineteen students specified that they had at least one electronic dictionary (usually Chinese/English, but three of these incorporated a le arners dictionary). Twenty-two students said they had used learners dictionaries before. Of these, eleven were Oxford learners dictionaries (the exact name and edition were usually not specified), two were Longman s, two were Collins Cobuild and four were Cambridge. Nineteen students used an electronic bilingual dictionary to do the exercise. Twenty used learners dictionaries (one of which was electronic). Three students did not use a dictionary. Twenty-three students used their own electronic or paper dictionary to do the exercise. The remainder used an unfamiliar dictionary. Students were asked whether the dictionary they had used in the exercise explained all the words correctly and, if not, which words were not explained and what 194

208 WORDS IN ASIAN CULTURAL CONTEXTS was not clear in the explanation. Since students had been given a chart to fill in, described in the Method section above, it is likely that most students used the chart s criteria to determine whether the inform ation they had gained was sufficiently comprehensive. The students were asked to fill in one column of the chart only, answering questions on either bilingual or learners dictionaries, but sixteen students filled in both columns. Eighteen users of bilingual dictionaries said that all the words used in the exercise were suf ficiently explained in their diction ary. Twelve felt that inadequate explanations were given. (Some students thus felt that their dictionaries were paradoxically both adequate and inadequate.) The words which caused most problems were idiosyncratically (4 instances), euphemistically (2 instances), panacea (1 instance) and elucidate (1 instance). Three commented that no indication was given of whether the word was formal or informal. Another three said that there was no indication of usage or context. Two said that the countability status of a noun was not given. Twenty-six who commented on learners dictionaries said that the information given was adequate. Only one felt that not enough information was given. Paucity caused problems for two students, elaborate for one and euphemistically for another. Of the comments made, four students said that the formal/informal status of the word was not given, and two found the phonetic symbols used for pronunciation too complicated. At the end of the questionnaire, participants were invited to write their comments on whether the exercise had changed their view of dictionaries. Sixteen replied affirmatively and eleven replied negatively. Of the negative comments, one student had already been using a learners dictionary for five years and had found it helpful. Two students commented that they would continue to use the dictionary they already had, since they were more comfortable with it. Only one student was disappointed, commenting that: I expected that a learners dictionary was more complete, with the same details in all words, not just in some words. Four students felt that a learners dictionary could enhance their English skills: Learner s dictionary is a powerful tool during English study in that it force you to forget your native language and go through it totally in English background. I need a learner s dictionary. It is useful not only in look up words but also helps your reading. Learner s dictionary is very helpful if I write a report or essay. I think I should spend more time on paper dictionary and I find it is very useful for writing in the right way. Students thus related the dictionary to the core skills of decoding and encoding, and realised the importance of working totally in their target language. Six students thought that the learners dictionary could give them more information than their bilingual dictionary: The dictionary shows the context besides word class and pronunciation. I think have a learning dictionary can help you to understand the words, to know the usage more thoroughly while you can only learn the meaning of the words in elecernic [sic] one which is suit for quick search. Yes, it tells me using the dictionary is not only just know what the meaning is, but also we should know how to use and explain them in English. Furthermore, the example sentence could help me to understand the different use. Yes, I might need to have a learner s dictionary if it helps me more than the electronic dictionary. Yes, more details and information, better than my electronic dictionary. Personally, I used to use electronic dictionary. It s faster, but I find the learner s dictionary can 195

209 ASIALEX 2005 help me a lot by providing details about the use of words. These students had moved on from seeing the dictionary purely as a decoding device and had realised the importance of cultural information and the usefulness of providing example sentences and collocations. Five students commented that the exercise had changed their view of dictionaries generally: Yes, it tells me using the dictionary is not only just know what the meaning is, but also we should know how to use and explain them in English. Furthermore, the example sentence could help me to understand the different use. Yes. I can t really tell what is a learner s dictionary before. Yes, dictionary is more useful than I think about it before. Yes, before I did not take note of the context of the word (formal, informal, scientific etc.). However, after doing this exercise I am made aware of this. As some dictionaries provide more information as well as the meaning of a particular word, it is probably better for people who are not only seeking the meaning of a word but also seeking the use of the word. It changed my view of dictionaries as before all the dictionaries are same. These students now saw that the dictionary was a wider resource than they had thought. In an overall comparison of the two types of dictionary, learners dictionaries scored more highly in every category (see table 1). The dictionary allowed me to: Bilingual dictionary (average response) Strongly agree (1) Strongly disagree (7) Learners dictionary (average response) Strongly agree (1) Strongly disagree (7) Understand items in context Use the words in speech or writing Know which words to use before or after (eg prepositions, participles) Know contexts for words (eg scientific, formal, informal) Know how the word is pronounced Table 1: comparison of responses to dictionary types These results indicate that the students realised that a learners dictionary could contribute not just to the decoding of unfamiliar words, but to their productive use in encoding and their correct use in terms of register and context. Limitations The main limitations to this study are the small sample size (n=42) and the fact that many students did not fill in the questionnaire correctly, completing answers for both bilingual and monolingual dictionaries, rather than for the dictionary used for the exercise. In addition, most students were not familiar with English learners dictionaries, and there was inadequate time during the lesson for them to learn how the dictionaries and their coding systems worked. 196

210 WORDS IN ASIAN CULTURAL CONTEXTS Recommendations It is recommended that in a future study more time be given to allow students to familiarise themselves with the new dictionaries, and longer explanations be provided as to the types of information to be found. Larger numbers of students could participate and a different questionnaire be administered for each dictionary type. Conclusion The students in this exercise realised that learners dictionaries have great potential to improve their academic writing skills by providing accurate grammatical and contextual information. However, even when students possess a learners dictionary, their own lack of grammatical knowledge may frustrate their search. It then becomes a chicken and egg situation. Can dictionary skills help students to improve their grammar, or is a good knowledge of grammar needed before dictionary skills can be gained? The two are probably complementary. The fact that four users of learners dictionaries failed to find the obvious information on a word s formal/informal status confirms the conclusion of Béjoint (1981) and Atkins and Varantola (1998) that students need training in the use of learners dictionaries. Students also need to be shown or reminded of how the forms of a word can be built. In this way they will be able to separate nouns, adjectives and adverbs, so that words such as euphemism, euphemistic and euphemistically can be more easily recognised and found. They could also be made more aware of the importance of, for example, countability, in determining article use, or the irregular forms of English verbs, and look for these details in their dictionary searches. Two vital questions remain. Will students use a heavy learners dictionary when a bilingual electronic dictionary is readily available? And, if the learners dictionary is added to their electronic dictionary, will they use it in preference to the bilingual dictionary? These are questions for further research. It remains clear, however, that no matter which country they are studying in, learners dictionaries can play a vital role in enhancing ESL students academic writing skills. References Atkins, B. T. Sue and K. Varantola (1998), Monitoring dictionary use in B. T. Sue Atkins (ed), Using Dictionaries: Studies of Dictionary Use by Language Learners and Translators (Tübingen: Max Niemeyer Verlag), pp Béjoint, Henri (1981), The foreign student s use of monolingual English dictionaries: a study of language needs and reference skills, Applied Linguistics 2(3): Bensoussan, Marsha, D. Sim & R. Weiss (1984), The effect of dictionary usage on EFL test performance compared with student and teacher attitudes and expectations, Reading in a Foreign Language, 2(2): Feast, Vicki (2002), The impact of IELTS scores on performance at university, International Education Journal 3(4): Hulstijn, Jan H. & B. T. Sue Atkins (1998), Empirical research on dictionary use in foreign-language learning: survey and discussion in B. T. Sue Atkins (ed), Using Dictionaries: Studies of Dictionary Use by Language Learners and Translators (Tübingen: Max Niemeyer Verlag), pp Lemmens, Marcel (1996), La grammaire dans les dictionnaires bilingues in H. Béjoint and P. Thoiron (eds), Les Dictionnaires Bilingues, (Louvain-la-Neuve, Belgium: Duculot s.a.), pp Nesi, Hilary (1998), Defining a shoehorn: the success of learners dictionary entries for concrete nouns in B. T. Sue Atkins (ed), Using Dictionaries: Studies of Dictionary Use by Language Learners and Translators (Tübingen: Max Niemeyer Verlag), pp Nesi, Hilary and R. Haill (2002), A study of dictionary use by international students at a British 197

211 ASIALEX 2005 university, International Journal of Lexicography 15 (4): Asian magazine advertising: database compilation and avenues for future research Andrew Moody University of Macau Azirah Hashim University of Malaya Abstract Print media advertising represents a significant economic and cultural development in Asia, where developed and developing consumer markets have experienced rapid and sustained growth. Advertising in Asia fulfils a number of functions as diverse as encoding national and ethnic identities, symbolising cultural values, and modelling innovative language. While the language of advertising in various Asian countries has received some attention, characterisitics and trends across Asia are rarely discussed. This paper will examine some of the previous work on advertising with special attention to Asian advertising and outline a number of approaches that have been taken to study the language of advertising. While previous studies have clearly demonstrated that Asian advertising may be of interest to sociolinguistics, discourse analysis, and critical approaches to language, there is an urgent need for a tool that will allow analyses of advertising across Asian cultures. The paper will also report on the on-going compilation of a database of Asian magazine advertising and present preliminary results from the research project. Because the project to compile the database emphasises random sampling of texts from specific comparable magazine genres in Asia, it offers a chance to engage in comparative studies of Asian advertising with the assurance that results will be both comparable and reliably representative of the publishing cultures from which they were sampled. Adjectives in other wor(l)ds: the uses and functions of adjectives in sociocultural context Elizabeth Morales-Nuncio University of the Philippines, Diliman bethmorales@yahoo.com Abstract The use of adjectives within cultural contexts represents the dynamic significance of language in accentuating and describing the cultural norms and practices of a given community or society. Adjectives ostensibly and ontologically ground subject-nouns 198

212 WORDS IN ASIAN CULTURAL CONTEXTS understood in a given context. In Philippine Tagalog society lexical meanings of words are derived primarily by adjoining adjectival words or phrases to illustrate verbally or textually the descriptive nature, degree, quantity and quality of subject-nouns being described. To situate this claim, the cultural and sociolinguistic functions of adjective words would be analyzed concerning the norms and practices governing (1) talismanism (2) euphemism (3) body-face sociolingual naming and (4) ethos of small-group communication. This study is an extension of early scholarship on Philippine language, society and culture and aims to disseminate to Asian audiences a working glossary of adjectival words in sociocultural context. It presents and lists down commonly used words being described or used to describe the norms and practices mentioned above. For example under talismanism these words are common: karunungang itim (black magic), Sto. Niñong hubad (nude Sto. Niño), magandang loob (virtuous self) and many others. The study presents how these norms and practices of communication and culture are embedded by or constructed in words that animate and connect people with their lifeworld. Introduction Filipinos possess primary traits and beliefs, which serve as strong foundations in the way they adapt to certain changes in the society. These changes may be brought by human intervention or by natural causes. Melba Padilla Maggay called these traits survival values in her book titled, Pagbabalik-Loob (1993), because these traits signify the abilities of Filipinos to suit themselves to changes happening around them. These abilities include the communicative process/practice to unite and engage with other Filipinos in attaining desired goals. In this process, the interlocutors designated/located within the same culture can distinguish those who share the same traits and ethnic behaviors. In view of this cultural reality, Filipinos use language in its verbosity, rigidity, artistry and rhetoric which animate and connect Filipinos with their lifeworld. McCarthy explains Habermas (1984: xxvii) definition of the lifeword as constitutive structural components of culture, society and personality rooted in the structural components of speech acts. The uses and functions of adjective-words are significantly manifested in both practical and solemn events and in both cognitive and affective ends. In the rules governing grammatical and semantic usage of adjectives together with the cultural norms and practices of Filipinos, there is a conceptual model that amplifies a discourse on identity and ethnicity. Grammar is the structure of language that deals with the correctness and appropriateness of usage in formal writing and speaking. Added to this precept is the purpose of language to communicate verbal and non-verbal messages by expressing and understanding human thoughts, emotions and actions. More so, social and cultural ingredients, or frameworks, ground communication and understanding of people. That is why rules may change and usage may vary based on specific sociocultural contexts. Nouns are used as terms to delineate proper or common usage in conversation, discourse or dialogue; while, adjectives, on the other hand, form part in the essential communicative intentionality and meanings in sentences. Therefore, this structure points to the performativity of Filipino language as it transcends the words with lexical meanings and rhetorical impressions and thereby circumscribes a world, a life-world that is embedded and constructed within the ethos and lifeways of Filipinos. In such case, the paper intends to discuss four lifeworld practices (aside from the linguistic 199

213 ASIALEX 2005 issue), which are delineated in the model below: Adjectives Talisman Nouns Locating someone Euphemism Legend: Language use of adjectives Indirect/euphemistic Direct and dialogic Person/participant Animals Body-face socio-lingual naming Objects 2: Talismanism Filipino society & culture Figure 1. Model of adjective use and functions in sociocultural context 1: Structure of the Sentence In linguistic context, the adjective is associated with the function of the noun and other parts of speech like the pronoun, verb and adverb. In short, the adjective is an extension of the noun. The essential use of the adjective would only be complete if this is related and associated with the noun and other parts of speech. This is the first significance of the adjective that focuses on its relation with the various parts of the sentence. As stated by Santiago and Tiangco (1991), the adjective states the characteristics or kind of person, animal, object, place etc. that stand as noun or pronoun in the sentence. The intrinsic meaning and structure of the sentence maintains this type of relationship between the noun or pronoun and the adjective (aside from other parts of speech). Therefore, this type of linguistic structure gives unity to the intrinsic nature and meaning of the sentence. But we must also acknowledge that the kind of significance given by the adjective evolves on the characteristics, comparativeness, levels, appropriateness an d nature/culture of the entity or phenomenon and people being described as words in a sentence. Studying the cultural framework of the adjective, this paper will attempt to elaborate the use of language outside the traditional boundaries (of language structure). Instead, it will focus more on the conspicuous usage of adjectives within the cultural and social realities of Filipinos. Magic and talisman are always connected with the body and character of the person. In this respect, adjective usage becomes prominent in grounding the discourse 200

214 WORDS IN ASIAN CULTURAL CONTEXTS and practice of talismanism. In Covar s essay Potensiya, Bisa at Anting-Anting (Potency and power of Talisman) the talisman becomes powerful only if the one possessing it has a benevolent heart (malinis na puso) and prays deeply, sincerely (2000: 70). Benevolent and sincere these are the characters of a virtuous self (magandang loob). The model shows that the power of talisman comes from within the self. It is through one s virtuousness that magic and potency spring eternal. He/she becomes then powerful (makapangyarihan). He/she becomes strong, is kind, generous, upright, wise, remarkable, and humble. Talisman works for people who are not egocentric or selfish. Bearing it incites selflessness and someone who has it unites with others as brothers and fellows in good spirit (kapwa). Through the eyes of others, what they see are the incorruptible body and the kind-hearted soul of someone bearing the talisman. The talisman is used for intersubjective relationship (pakikipagkapwa) like healing the sick and rejuvenating the soul of the distress. It is used also to communicate with the spirits of nature. Apart from this, the cultural discourse on talismanism bordering on urban/folk legend talks about the power of talisman like protecting the bearer against bullets, having disappearing spells, healing and driving away bad spirits. 3: Ethos of small-group communication: the case of locating someone in the community Zeus Salazar stated in his essay, Various Bases for a Filipino Psychology (Ilang Batayan para sa Isang Sikolohiyang Pilipino), that social traditions are dependent on language (Pe-Pua, 1982: 53). Part of this Filipino tradition is the process of familiarizing and engaging with neighbors or with community members. For instance, in this situation, (see diagram) two or more persons (they either come from the inside or the outside of the community) may acquaint themselves with one another, starting off with queries on someone in the community. This instant acquaintance between/among familiar strangers is forged by curiosity and the desire by someone to break the cultural barrier through assimilating him/herself in the community. If, say, someone needs to locate Mang Isko in Barrio Masilong, certain information needs to be asked among community members who know Mang Isko personally. Community members may have the following descriptions about Mang Isko: Si Isko yung payat, na maitim, nakasalamin at medyo kuba ba ika mo? (Did you say Mang Isko, the one who s slim, dark-skinned, bespectacled and short?) Another would say: Hindi siya gaanong matangkad pero matipuno ang katawan, dyan siya malapit sa gulod nakatira. (He s not quite tall but he s a husky fellow; he lives near the hilltop.) Sometimes, the familiar stranger may get as much information as he/she wants. If he/she were to chance upon Mang Isko s inner circle, those who know him well (relatives or immediate neighbors) might give more than what is asked for. Si Petra yung asawa niya, labandera, at may tatlo silang anak dalawang lalaki t isang babae. (His wife is Petra, a laundry woman. They have three children three boys and a girl. Eventually the familiar stranger, who might be just living in a nearby barrio, would know from his/her acquaintances in the community the life and character of the person he/she is looking for. 4: Euphemism According to Maggay (2002: 187), euphemism (ligoy) refers to a kind of human thought driven by inward impulse (and manifested through outward affective speech), which is most of the time mistaken as circumlocutory form of expressive communication. The direct opposite of euphemism is candor, which Filipinos on some 201

215 ASIALEX 2005 occasions try to supplant with bloated euphemism verbose, elaborate and flowery in view of having a s mooth interpersonal relationship (SIR) (Lynch 1973: 10). Euphemism is an affective language, a kind of which its intentionality is grasped on a very particular occasion that commands tact and decisive choice of words. Relatively, this happens only inside a closed group of immediate family members, friends and intimate acquaintances. Outside this group, euphemism entails a different meaning, which is generally expressed as a defense mechanism to shy away from conflict and misunderstanding. T hus, courteous language is demanded. But far from a mere avoidance of conflict in favor of maintaining smooth relationships, adjectives are essentially euphemistic for Filipinos. Thus, Filipino expressive utterances in light of this assumption relate euphemism and communication as both rational and affective. Hence, if a child asks for additional allo wance from his mother, the child then chooses to convey the message using euphemistic verbosity. Ang ganda talaga ng nanay ko, masipag na at masarap pang magluto! (My mom is really pretty, diligent and an excellent cook!) This is followed by stating the real motive. In two ways, the communicative power of euphemism in this example extends towards the affective admiration of a child to his mother and a rational pursuit in achieving his desired goal: additional allowance. It is in this respect that the order of statements recognizes the precedence of emotive/affective/adjectival words to reinforce the rational end of communication. Also, Filipinos like poets use euphemism in expressing their feelings of love and passion. In some instances, though, friends may opt to seek approval regarding certain tastes in their choices of material things or the appearance they project. If one asks, How s my dress?, appraisal points to detailed characteristics instead of wholesale approval or disapproval. Maganda ang butones, parang gintong beads. Yung haba ng damit ok lang. ( The buttons are beautiful like golden beads. Well, the length is just fine.) Or in some cases, a mere utterance of OK displaces concrete value of choice. The burden goes straight back to the one making the choice. Verbosity cuts short the option back to the one making the decision. This happens latently and sometimes unintentionally, leaving the one who asks unaware of the burden to make the choice for him/herself. Or perhaps in other instances, if one is aware, the communicative event takes place to fish out compliments from others. 5: Body-face Sociolingual Naming The power of adjectives to describe the face and parts of the body can be understood through the models explained above in euphemism and ethos of small-group commu nication. Using euphemism, the undesirable and direct assault on one s face, when a person attempts to describe it, is shifted rhetorically to other positive traits. Thus, instead of saying, He has an ugly face, one can possibly point to his muscular body or her beautiful hair. Kung pango ang ilong tukuyin ang kanyang magandang mata, kung mahaba ang baba tukuyin ang mapupulang labi. (If she has a flat nose then point out her tantalizing eyes; if she has a long chin, point out her red luscious lips.) This is how asking somebody about someone in the community operates. It dwells too in highlighting the positive characteristics and tries not to divulge any negative traits especially to someone who is a stranger in that community. Covar (1998) mentions the ways in which body-face is described and talked about culturally in the Filipino context. Ang malagkit ang tingin, may pag-ibig na ipinaparating. (His enticing look signifies love.) Maduming bibig, masamang magsalita. ( Impure mouth means badmouthing.) Malapad ang noo, matalino. (Wide forehead 202

216 WORDS IN ASIAN CULTURAL CONTEXTS signifies intelligence.) Maitim ang budhi, masamang tao. (A callous self means an evil person.) Busilak ang kanyang kalooban, mabuti siyang nilalang (A good heart means a virtuous self.) Conclusion This paper has attempted to lay down the basics for a critical interrogation that expound the relationship between culture and the Filipino identity through language in general and through the use of adjectives in particular. This is the linguistic construct of Filipinos, shaped and molded by social and cultural imperatives. Such construct can be derived from the syncretic make-up of local and foreign influences in which past colonialism and today s postcolonialism are interposed and examined culturally and discursively. It is important also that within these influences we should identify and critique the central culture, as mentioned by Maggay (2002), since societal change means shaking inevitably the grounds of statism and traditionalism. The cultural and linguistic dynamics of adjectives in the Filipino context defines and illustrates the richness of words circumscribing the Filipino lifeworld. Words in other worlds (lifeworld) could build up a corpus of cultural dynamics where language plays a critical role in shaping and reconstructing it. Hence, as shown in this project, the criticality and discursivity of cultural dynamics rest on signifying the deeper meanings of what a true Filipino identity is all about. References Covar, Prospero (2000), Potensiya, bisa at anting-anting, in Benilda Santos (ed), Likha (Quezon City: Kagawaran ng Filipino, Ateneo de Manila University), pp (Ch. 3). (1998), Larangan (Manila: National Commission for Culture and the Arts). Habermas, Jurgen (1984), The theory of communicative action, translated by Thomas McCarthy (Boston: Beacon Press). Lynch, Frank (1973), Social acceptance reconsidered, in Frank Lynch and Alfonso de Guzman II (eds), Four readings on Philippine values (Quezon City: Ateneo de Manila University), pp Maggay, Melba Padilla (2002), Pahiwatig: kagawiang pangkomunikasyon ng Filipino (Quezon City: Ateneo de Manila University Press). (1993), Pagbabalik-loob: moral recovery and cultural reaffirmation (Quezon City: ISACC). Salazar, Zeus A. (1982), Ilang batayan para sa isang Sikolohiyang Pilipino, in Rogelia Pe-Pua (ed), Sikolohiyang Filipino: teorya, metodo, at gamit (Quezon City: Surian ng Sikolohiyang Pilipino), pp (Ch. 2). Santiago, Alfonso and N. G. Tiangco (1991), Makabagong balarilang Filipino (Manila: Rex Bookstore). Abstract Students attitude toward dictionary use Minoru Moriguchi Department of Social Environment Hiroshima International University m-morigu@it.hirokoku-u.ac.jp When compiling a dictionary for students or teaching students how to use a dictionary, 203

217 ASIALEX 2005 we need to know how they are using dictionaries in their daily lives. This paper studies the attitude of university students toward dictionary use, based on a questionnaire survey. The survey questions include: (1) the students interest in language, either mother tongue or foreign language; (2) the number of dictionaries they own; (3) the kind of circumstances under which they consult dictionaries; and (4) how they are using electronic dictionaries. The conclusion reveals several factors concerning dictionary use. Some suggestions will also be made for dictionary instruction and the development of electronic dictionaries. The Mātāpuna dictionary writing system Dave Moskovitz Thinktank Consulting Limited Wellington, Aotearoa / New Zealand dave@thinktank.co.nz Abstract Mātāpuna is an Open Source, easy to use, multiuser, multilingual dictionary writing system. It is currently being used to compile Te Taura Whiri i te Reo Māori / The Māori Language Commission s monolingual dictionary of the Māori language. This paper provides a brief overview of the system, and invites interested parties to use the system and join the development community. Background Te Taura Whiri i te Reo Māori / The Māori Language Commission (TTW) is currently compiling a monolingual dictionary of the Māori language of approximately 20,000 entries. In the initial stages of the project, TTW examined existing software products on the market for writing dictionaries, and found none of them suitable for their project requirements. Requirements Specifically, they needed a system with the following features: Easy to use most of the people employed on the project were hired for their language skills, not their IT skills Multi-user dictionary staff collects information from all over the country, and they needed to be able to use the system to help them collaborate from remote locations Workflow the system should have the ability to pass entries through various stages, from writing through editing and quality assurance, and support the interaction of staff. 204

218 WORDS IN ASIAN CULTURAL CONTEXTS Open Source Self-checking the system should perform basic checking on information as it is entered, to maintain the quality of data Provide administrative support head office staff wanted to be able to monitor staff performance, as well as the progress of individual entries and the work as a whole Flexible the system should be easy to modify to support changing requirements through the life of the project Inexpensive the system should not require costly server hardware to run on, and should be easy to support on a wide variety of client machines running in very remote locations. None of the publicly available systems available at the time of software selection in late 2002 came close to fitting these criteria. TTW decided to commission the development of a custom-built system. In order to mitigate risk, TTW embraced the Open Source development model using the GNU Public License. The Open Source model provided the following benefits: Prevent vendor capture as the source code would always be available to anyone who wanted it, TTW could at any time use any software vendors whom they felt suited the project. The dictionary data would not be locked up in a proprietary system. Immortality since the software would not be owned by anyone, it would be impossible for any stakeholder entity to kill the project and prevent it from continuing in some form. Anyone wanting to continue the project would have the ability use the software. Encourage international collaboration Aotearoa / New Zealand is a small country with limited resources, and the project itself is being completed on a limited budget. TTW knew that there were potentially great benefits from encouraging people working on dictionary projects for other languages to continue to develop the system to the mutual benefit of everyone using the software. Low total cost of ownership there would be no licensing costs for the system itself or any of the underlying software components, no matter how many users. System Description The system is a classic LAMP system, based around Linux, Apache, mod_perl, and Postgres. The production system currently runs on a Pentium-600 server that was built in 1995, and supports 20 users scattered around the country. In the last two years of operation it has never crashed, other than for power failure. All user functions are accomplished through a standard web browser. Client systems run Windows, Mac, and Linux. Currently, the user interface is available in two languages: Māori and English. Translation to other languages would require little effort. All text is stored as Unicode, and the system can be used for any language that has a UTF-8 encoding. Unicode text is exchanged with the browser as HTML entities 205

219 ASIALEX 2005 to ensure maximum browser compatibility. The browser interface is very lightweight and entirely text based, and works well over dialup-grade bandwidth. Mātāpuna stores all data in a relational database, using ANSI SQL with no stored procedures. The current system uses Postgresql, but in principle any database with a Perl DBI/DBD interface (eg MySQL, Oracle, SQL Server) could be used. This differs from other dictionary writing systems, many of which store their data as XML. In the designer s opinion, the benefits of using a relational database (e.g. manageability, maintaining referential integrity, product stability) outweigh the convenience of using XML as a data store. Mātāpuna can use XML as an import or export format. The dictionary entries could be browsed online by the public, but this feature is not currently being used. Printed copies of the dictionary are produced by running a script. No further manual processing is required, other than adding front and back matter. Current users The main users of the system are the dictionary writers, who are responsible for collecting headwords, writing the dictionary definitions, sourcing citations or writing usage examples, and providing all of the other information that goes into each dictionary entry. The current writers are Māori language experts, who work remotely from their home locations up and down the country over the Internet. Editors and head office administration staff make up the bulk of the team. None of the current users have extensive IT or lexicographical experience. Workflow Before Mātāpuna was implemented, many of the entries in the dictionary began life as rows in Word document or Excel tables. They were imported into the database using one-off Perl scripts. However, the bulk of the entries were entered into the system on-line using the Add Headword function. 206

220 WORDS IN ASIAN CULTURAL CONTEXTS The system has a built-in corpus search facility to assist with entry writing, as well as a function to search data within the dictionary database. When a writer has finished adding an entry, they can move on to perform a self-check of their work using My Editing function. The first thing a user is presented with on each screen is a preview of what the entry will look like in its final form, followed by form boxes in which information can be edited. 207

221 ASIALEX 2005 The system performs a number of sanity checks on each entry, whether all words appearing in the definition are themselves defined in the dictionary, whether definitions contain non-māori characters, whether critical fields have been left blank, etc. Users must check an ignore errors box to save entries that do not pass one of the automatic checks. The system also maintains cross-references of synonyms and variants of each entry, and prohibits circularity. When satisfied with their own work on an entry, a user will pass the entry, and the entry will be queued for the editor managing the next stage in the editing process. At any time, a user can send an entry to a specific user for their comments, advice, or other assistance. This dialog between users is maintained in the editorial comments field in each dictionary entry. After working on an entry, a recipient can either send it back to the person who was originally working on the entry, or reinsert the entry into the main work queue. All previous versions of the entry are saved in an archive table in the database, so that work can be recovered even if a drastic editorial mistake has been made and not discovered until months later. The history of each entry can be viewed by a user at any time. Printing a draft copy of the dictionary or the final work is accomplished by running a batch script, which produces an RTF file. 208

222 WORDS IN ASIAN CULTURAL CONTEXTS Reporting The system can produce numerous reports for cross-checking information in the dictionary as well as progress monitoring and overall project management. Reports include finding entries with missing information, reporting on daily progress by user, examining the structure of synonym groups, etc. The reports are based on SQL queries and are very easy to add or modify. Collaborators Wanted The Mātāpuna Dictionary Writing System is a free, simple, multiuser system being used to compile a major lexicographical work. It is particularly suited to lexicography projects on tight budgets using staff without extensive IT or lexicography experience. Being browser based, it can be used by geographically remote staff. It handles Unicode well, and can be translated into and used for any language with a UTF-8 encoding. Thinktank Consulting and Te Taura Whiri i te Reo Māori invite any parties interested in using, or extending the system to find out more about the project and join our developer community. For more information, see 209

223 ASIALEX 2005 Strategies used by international participants in discussion on an electronic bulletin board Kyoko Motobayashi University of Toronto Tomoko Takeda Keisen Women s College Satoko Tokumaru Japan Women s College of kmotobayashi@oise.utoronto.ca taketomo90@hotmail.com Physical Education sa-toku@air.linkclub.or.jp Abstract This study reports the participation strategies used by international students, most of whom are from Asian countries, in the discussion in Japanese among Japanese university students and international students in a unique project using an electronic bulletin board called Sakubun org. It focuses on 37 utterances by the international participants and analyzes the data qualitatively. As a result of the analysis, it was observed that the international participants used the following strategies: (a) bringing in personal experiences, including I as nationally/culturally neutral, I with national/cultural identity, and I as a foreigner in general; (b) referring to Asian identity, as a comprehensive category; (c) discussing Japan and/or Japanese; and (d) other strategies such as remaining an objective discussant without bringing in their identities or experiences. These findings show that the international participants in this project used a variety of strategies in order to join in on the discussion with their Japanese peers, oftentimes by using their international status as resources, creating solidarity with other discussants, or taking an outer/objective position. These characteristics also helped to contribute broader views to the discussion on the bulletin board. Introduction Recently, there has been an increasing use of computers and the Internet in the field of education, accompanied by many theoretical debates both supporting and opposing the use of technology in the classroom. Barker and Kemp (1990) introduce their pioneering practices and highlight the natural affinity between the Internet and a postmodern pedagogy presenting the Network theory. Cummins and Sayers (1995) introduce eight practices connecting students of diverse backgrounds from all over the world, aiming to foster cultural and critical literacy among the students. Bigum and Kenway (1998) describe four types of reactions to the use of computers in education, and Cuban (2001) points out that computer use would be very ineffective and unimaginative unless an appropriate goal is set. Some researchers have also emphasized the importance of pedagogy in educational practice using computers (Cummins, 2004; Scourtou et al., in press; Murray, in press; Warschauer, in press). For second language users, computer-mediated communication seems to have some positive impact. Murray (in press) mentions that authentic and communicative language use can be facilitated through technology and/or the Internet. Murray also summarizes studies reporting that shy students are encouraged to use the target language more frequently than in face-to-face conversation. In addition, research shows that in computer-mediated online discussions, silent students increased their 210

224 WORDS IN ASIAN CULTURAL CONTEXTS participation compared to in face-to-face discussions (Warschauer, in press). In other words, it is possible that technology can facilitate second language users participation in authentic communication in the target language. The following study describes the participation strategy used by international students in a written discussion on an electronic bulletin board. The electronic bulletin board, which enables asynchronous discussion mediated by computers, is a unique medium different from computer-assisted classroom discussion (synchronous discussion) or exchanges (asynchronous correspondence) summarized in Warschauer (in press). Therefore, it will be beneficial to examine the interaction occurring on it. In this paper, the Sakubun.org project will first be reviewed. Then an analysis of the participation strategies will be given. The project: Sakubun.org The Sakubun.org project is a unique practice exploring the possibility of creating an educational public sphere (Habermas, 1994) on an electronic bulletin board, connecting university classes both in Japan and abroad. The students participate in this activity as a course requirement but the writing itself is not graded. This project employs a 3 steps and 5 principles text exchange activity : the 3 steps comprise first essays, readers responses, and writers comments to the responses and the 5 principles include (1) non-judgmental attitude, (2) frank and open discussion, (3) receptive and active listening attitude, (4) anonymity, and (5) voluntary participation. This text-exchange activity with 3 steps and 5 principles was first practiced on paper where students initially wrote the essays on paper and later exchanged them with each other to read. Based on this practice, the participants had the opportunity to express their own feelings openly, experienced mutual acceptance and understanding and intimate feeling (Tokumaru, 2000a, 2000b). After 10 years of practice on paper, this project was transferred to the Internet, where the same process as the paper-based one took place (Tokumaru, 2001). Participants and the data The session analyzed in this study took place from August to November in 2002, on the topic of What the World Cup soccer left for us. About 170 university students participated in this session including 99 Japanese university students, 57 international st udents studying at Japanese universities, and 10 learners of Japanese as a foreign la nguage studying in Australia. In this paper, students in the latter two categories are defined as international participants. The countries/places of origin for these international participants based on their own report are as follows: China, 38; Taiwan, 13; Korea, 6; Australia, 4; Hong Kong, 1; Indonesia, 1; France, 1; Vietnam, 1; Malaysia, 1; Myanmar, 1; Unknown, 1. Of this data, 14 threads which consist of 72 compositions (37 by international participants and 35 by Japanese participants) were analyzed here, selected in the procedure described below. Analysis This research employs qualitative analysis. As a first step in analyzing the multi-voice threads, we looked at the contents of the first essays. A researcher coded the 211

225 ASIALEX 2005 contents of each essay with short words or phrases and then discussed with the other two researchers to modify and confirm the code. Each essay was then coded by the three researchers. When the coding differed, we discussed the matter until a common code was decided upon. After the coding, we selected threads (the sequence of essay-reactions-comments) including the three most frequently occurring contents. These threads numbered 34, and the reactions and comments were also coded with the same codes to select threads in which essay, reactions, and comments all contained these three contents. As a result, the thread selected was 31. Among these, the threads in which international students participated were 14 and the writing samples of international participants (regardless of essay, reactions, or comments) were 37. These writings were analyzed in this study. For this data, a researcher coded the participating strategies of the international participants with a word or a short phrase. Then the code was discussed with the other two researchers. The codes will be discussed in the next part. Findings The following codes emerged from the analysis: A. Bringing in personal experiences. The first strategy the international participants used was bringing in their own experiences. This category is further divided into three subcategories: I as nationally/culturally neutral, I with national/cultural identity, and I as a foreigner in Japan, more generally. A-1. The I as nationally/culturally neutral strategy brings in the writer s own experience which is unrelated to her/his cultural background. An example of this strategy is as follows: I have been susceptible to sickness since I was a child so my doctor suggested that I play some sports. Therefore, in order to be stronger, I belonged to a soccer team at my school and played soccer. Our varsity team entered the National Title match and won second place. Later, I played as a member of the national team. I love soccer best among all the sports I have played. (From: Taiwan, Current residence: Australia) (Translated from Japanese into English by the researcher, same as below) A-2. The I with national/cultural identity strategy brings in the writer s experience and/or identity as a person of that nationality/culture. Examples of this strategy are as follows: I am Chinese and I cheered for the Chinese team. However, the Chinese team lost three straight games. Unfortunately, the Chinese were not among the best 16 teams, similar to the Japanese team four years ago. But the Japanese team was among the best 16 teams and showed the culture and spirit of the people. I am hopeful that the Chinese team will become one of the best 16 four years from now, showing the spirit of the Chinese people. (From: China, Current residence: Japan) Reading the response to my essay, I remembered my experience. When the Olympics were held in Sydney in 2000, I had not been interested in the Games before they started. However, after I saw the opening ceremony and the first game, I became interested. At that time all the Australians were excited and it was fun, the same as the case of the World Cup in Japan. (From: Australia, Current residence: Australia) A-3. The I as a foreigner in Japan strategy positions her/himself as a foreigner in general, regardless of nationality/culture. An example of this category is as follows: 212

226 WORDS IN ASIAN CULTURAL CONTEXTS I could not watch the Games because I was working part-time at night in January. But I understand the heat and harmony [the writer is saying in the original essay]. There were many people e xcited and dancing, especially after the Games of the Japanese team, when I went home after my job. Seeing this scene, I also became excited as a foreigner. (From: China, Current residence: Japan) B. Referring to identity as a comprehensive category. The second strategy used by international participants referred to a broader definition of identity. In particular, because of the high proportion of Asians among the international students in this project, the creation of an Asian identity was reported by some participants. In other words, this is a We as Asians category, including their own country in relation to the discussion on Japan and Korea as the host country of the World Cup. In my country soccer is not very popular, and [my country] does not participate in the World Cup, so even though I m in Japan now, I can t understand very well why Japanese young people are so enthusiastic about the World Cup. Because this year the World Cup was held in Japan and Korea, for the first time, Japanese supporters could cheer for soccer players in Japan. (From: Taiwan, Current residence: Japan) Conclusion I cannot forget what my Asian friend said at that time: Because this World Cup is held in Asia, let s cheer for the Asian team as Asians, anyway. For a long time, Asian soccer had been undervalued, but many Europeans were amazed at Japan s being among the best 8 and Korea s being among the best 4. This result is not only for Japan and Korea, but for all Asian unity. All Asians are proud of Japan and Korea s achievement. Asian soccer is never weak. It will become much stronger in the future. (From: China, Current residence: Japan) C. Discussion about Japan and Japanese. The third strategy used by the international participants was to discuss Japanese objectively. An example of this strategy is as follows: D. Remaining an objective discussant without bringing in their identities. The writings categorized in this strategy refer to those remaining as objective discussants without bringing in any experiences or identities. An example of this strategy is as follows: Sports are for relaxing the soccer games are not wars between countries but competition s of the techniques of their soccer. However, many peop le forget this and regard the nature of the soccer games too seriously.it is necessary to distinguish between wa rs and sports. (F rom : Hong Kong, Current residence: Australia) Second language users tend to be placed in a peripheral position in the discussion community in general. However, international participants in this project actively participated in the discussion with their Japanese peers, using a variety of strategies. In particular, they tended to bring in their experiences and identities related to their own countries/cultures. In other words, they used their cultures as resources to participate in the discussion community (A). In addition, international participants create a solidarity with other discussants by finding a common category to which most can belong (B), stand in an objective position to analyze Japan (C), or remain in the same position as their Japanese peers regardless of their nationality/culture (D). These characteristics contributed to broader views in the discussion on the bulletin board. 213

227 ASIALEX 2005 References Barker, T. T. and F. O. Kemp (1990), Network theory: A postmodern pedagogy for the written communication, in C. Handa (ed), Computers and community: Teaching composition in the twenty-first century (Portsmou th, N H: Heinemann), p p (C h. 1). Bigum, C. and J. Kenw ay (1998), Ne w info rmation technologies and th e ambiguous futu re of schooling - Some possible scenarios, in A. Hargreaves, A. Lieberman, M. Fullan and D. Hopkins (eds), International handbook of educational change (Dordte cht, Netherlands: Kluwer), pp Cuba n, L. (2001), Oversold and underused: Computers in the classroom ( Cambridge: Harva rd Univers ity Press). Cummins, J. (2004), Technology, literacy, and youn g second langu age learners: Designing educational futures. Paper prepared September 2004 for the Universi ty of California project, An Exploration of Technology in Suppor t of English and Reading Skills for Young Second Language Learners. Cummins, J. and D. Saye rs ( 1995), Brave new schools: Challenging cultural illiteracy through global learning networks (New York: St Martin s Press). Habe rmas,j. (1994), Koukyousei no kouzou tenkan [The structural transformation of the public sphere] (S. Hosoya and M. Yamada, Trans.) (Tokyo: Miraisya), Original work published 1990, Strukturwandel der Offentlichkeit: Untersychungenzu einer Kategorie der burgerlichen Gesellchaft, mit einen Vorwort zur Neuauflage (Frankfurt am Main: Suhrkamp). Murray, D. E. (in press), Computer-assisted approaches to English language teaching and learning, in J. Cummins and C. Davison (eds), International handbook of English language education (Dordrecht: Springer). Skourtou, E., Kourtis-Kazoullis, V. and J. Cummins (in press), Designing virtual learning environments for academic language development, in J. Weiss and J. Nolan (eds), Handbook of virtual learning environments (Dordrecht, Netherlands: Kluwer). Tokumaru, S. (2000a), Daigakusei wo taisyou to sita sakubun koukan katsudou ni okeru kojin katei no bunseki [A Composition Exchange Project by the anonymity of university student mutuality] Ningensei Shinrigaku Kenkyuu [The Japanese Journal of Humanistic Psychology], 18(1), pp Tokumaru, S. (2000b), Ryuugakusei to nihonjin gakusei no sakubun koukan katsudou ni okeru kojin shinri katei. [Individual mentality process in Composition Exchange Project between foreign university students and Japanese university students] NihongoKyouiku [Journal of Japanese Teaching], 106, pp Tokumaru, S. (2001), Sakubun koukan katsudou no inta-netto riyou no kokoromi. [Composition exchange project on the Internet] Gengobunka to Nihongo Kyouiku [Japanese Language Education], 22, pp Warschauer, M. (in press), Technology and writing, in C. Davison and J. Cummins (eds), Handbook of English language teaching (Dordrecht, Netherlands: Kluwer). Retrieved March 15, 2005, from Translations of topic-comment structures of Vietnamese into English Pham Phu Quynh Na University of Western Sydney ppquynhna@yahoo.com Abstract Pr evious research on the syntactic typology of Vietnamese has led to two controversial findings. One set of studies has claimed that the language is a subject-prominent language (Ly 1948, Chinh & Le 1973, Ban 1987, Thin 2001). The other set states that the basic structure of Vietnamese manifests a topic-comment relation rather than a subject-predicate relation (Thompson 1965, Dyvik 1984, Hao 1991, Anh 2000). Rosen 214

228 WORDS IN ASIAN CULTURAL CONTEXTS (1998), i n her PhD thesis, has elaborated on the subject of Topic and Empty pronouns in Vietnamese and has proved that topic-comment constructions and empty elements are ba sic cons tructions in Vietnamese. According to her, there are five types of relations that may exist between a Noun Phrase topic and the comment in Vietnamese. With the assumption that Vietnamese is a topic-prominent language, the aim of this study is to investigate whether Vietnamese students have problems with the specific task of locating and translating the dropped subject and empty elements in the topic-comment structures of Vietnamese sentences into English. The study is based on a preliminary report of an error analysis of Vietnamese-English translations by 95 Vietnamese EFL students from the first, second, third, and fourth years of the Department of English Language and Literature at the University of Social Sciences and Humanities, Ho Chi Minh City. The data will be the translation texts of these students from the same source text. From the error patterns found in 95 translation texts, this paper summarises the possible problematic areas when translating topic-comment structures and recommends some practical translation strategies in translating topic-comment structures of Vietnamese. Introduction Although many previous studies have held that Vietnamese is a subje ct-prominent language, recent scholars have attempted to analyse Vietnamese sente nces from a different perspective in which the topic-comment structu re represents the basic structure (Hao (1991), Rosen (1998), Thanh (2003). Hao (1991) states that wh ile the subject is an obligatory constituent of a sentence structu re and always occupies the initial position of a sentence in English, it is not the case in Vietnamese. Vietnamese sentences often start with a topic which can be taken over by any part of speech aside from the grammatical subject. This leads to two phenomena: (1) the subject can be dropped in Vietnamese sentences and (2) the position of a grammatical subject in Vietnamese can be occupied by an adverb, an object or indirect object or sim ply a word semantically related to the comment discussed in the topic. Rosen (1998) has classified five main types of topic constructions often found in Vietnamese sentence types. (1) The topic may be understood as filling a gap in the comment. (2) The topic may be coreferential with a noun phrase or pronoun in the comment. (3) The topic may be semantically related to any specific constituent in the topics, its referent is simply what the comment is about. (4) The topic may not be related to any constituents, empty or overt, in the comment. (5) The topic may be a verb phrase, a clause or an embedded topic-comment construction. Regarding the empty pronoun, Rosen (1998) has identified four types in Vietnamese: (6) Empty pronouns in simple sentences. (7) Empty pronouns in embedded clauses. 215

229 ASIALEX 2005 The Study (8) Empty pronouns in consecutive clauses. (9) Empty pronouns with indefinite reference. Four groups of 95 students in their first, second, third and fourth years of the Department of English Languages and Literature of University of Social Sciences and Humanities of Ho Chi Minh City were chosen to be the informants for this study. The subjects were asked to translate a 250-word text from Vietnamese into English in 180 minutes without being told which structure was being tested. The Vietnamese text is an 18-sentence extract from an article named Meâ Thaûo-thôøi vang boùng from Tuoåi Treû (The Youth), a well-known newspaper in Vietnam. The text was chosen because of its richness in sentence types: many different types of authentic Vietnamese sentences can be fou nd in the text, including minor sentences, i.e. elliptical sentences, sentences without either Topic or Comment, sentences in which the Topic is identical or not identical with the Subject, and sentences in which the Topic or Comment itself is an embedded Topic-Comment structure. With such a variety of Topic-Comment structures, the text promises to be a good environment for the author to discover whether the subjects have problems with locating and translating the subjects from Vietnamese to English and what these error types are. Content Analysis The five types of syntactic errors found to be prevalent in the data in the task of locating and rendering the subject into the target text include: Omission of Subject, Repetition of Subject, Inappropriate Choice of Subject, No Logical Connection Between Subject and Predicate, and Inappropriate Connection Between Subject and Passiv e Verb. The data reveals that Omission of Subject is the most frequent type of error, with the second most frequent type being Repetition of Subject and No Logical Connection between Subject and Predicate. Types of errors Year 1 Year 2 Year 3 Year 4 Total Omission Of Subject (Error Type 1) Repetition Of Subject (Error Type 2) Inappropriate Choice Of Subject (Error Type 3) No Logical Connection Between Subject And 3 Predicate (Error Type 4) Inappropriate Connection Between Subject And Passive Verb (Error Type 5) Table 1. Distribution of errors relating to the translation of subjects As Table 1 shows, omitting the subject is the most frequent type of error among all types of errors relating to the sentence subject. Although most Vietnamese can understand (or have the impression that they understand) the empty pronoun or dropped subject when they read a text, they do not always correctly identify the referent of the missing subject when they need to translate these sentences into English. The omission of the subjects may be traced to two reasons: firstly, the students may not be able to identify the referent of the missing subject or, secondly, they do not know that it is necessary to find the missing subject for it to be rendered into a correct English sentence. Besides these types of errors, students also make inappropriate 216

230 WORDS IN ASIAN CULTURAL CONTEXTS choices of subject, repeat the subject, or write sentences in which the subject does not semantically match the predicate, or the passive verbs. The distribution of errors across 18 sentences is discussed in detail below. Sentence Number Error Type Error Ty pe Error Type Error Type Error Type Table 2. Frequency of errors relating to translating subjects in 18 sentences Sentences 1 and 13 show the highest percentage of subject omission. In these two sentences, the subject is completely dropped and the translators have to refer to the context to identify the subject, which, in this case, is I referring to the narrator/writer of the source text. Errors where the subject is repeated are observed in sentences 3, 15, 17 and 18. These sentences have complex structures with embedded topic-comment constructions, requiring the skill of manipulating the subject and the sentence structure. Therefore students who persist with the literal meaning of the text and the surface structure of the source text are more likely to make this type of error. Five students made the wrong choice of subject in Sentence 2, which shows they did not make a very careful analysis of the text before they began their translating. The dropped subject of sentence 2 is I ; however, these students erroneously chose Tröôùùc heát (Firstly) as the subject of the sentence, which is actually a transitional adverb. 16 cases of errors occur when the subject does not match the predicate in sentence 5. In this sentence, the topic Caûnh thaû ñeøn trôøi (scene of dropping lanterns) does not coincide with the grammatical subject of the main verb: it is the object of the verb laøm (make/carry out) in the comment. The actual subject is ngöôøi ta or ngöôøi xem, which is dropped. To translate this sentence, students are required to identify the referent of the empty pronoun ngöôøi ta or ngöôøi xem. However, many of them chose the wrong subject when they selected the topic of the sentence Caûnh thaû ñeøn trôøi (scene of dropping lanterns), an inanimate referent, as the subject of the verb töôûng nhö in their translation. This led to numerous errors appearing in the translations of this sentence. The last type of error, which can be attributed to the poor handling of the subject, is the mismatch in meaning between the subject and the passive verb, which occurs most often in sentence 18. The following tables ( ) display sample sentences for the five error types mentioned above with the distribution of errors by student year. Year Sample Sentences Distribution 1 Being fascinated by named and nameless characters again. 2 2 Also, being fascinated by the named characters and nameless ones Still fascinated by named and anonymous characters Supposed (supposing) that it was (were) named MT, sounds (it would sound) candid and more unique. 1 Table 8.1: Sample Sentences of the Error Type Omission of Subject 217

231 ASIALEX 2005 Year Sample Sentences Distribution 1 Almost each topic, it s permeated with the soul of Vietnamese 1 2 To, the singer is dramatic, wandered (wandering) but dignified, 5 3 amorous but loyal, seductive and serious, petite but unimportant. The scene of burning furniture, the explosive sound of furnishings, the 10 scene of a gunshot firing at the doll, all of them make viewer s 4 blood run cold. 1 Tam, a player of the moon shaped guitar was so wonderful that he was desperate, frustrated but giving, the gesture (gestures) totally suited the mood. Table 8.2: Sample Sentences of the Error Type Repetition of Subject Year Sample Sentences Distribution 1 First, it is the being fascinated by images full of film s language 1 2 Among successful supporting characters, the old servant is so vivid and moving and that (he) is worthy to be nominated as the best supporting 5 3 character The first is fascinated by all language of motion picture images. 1 Table 8.3: Sample Sentences of the Error Type Inappropriate Choice of Subject Year Sample Sentences Distribution The scene of dropping outside lantern is made (can happen) only by charm. The sight of dropping lant erns is considered that can be only carried out by magic. The sight of dropping lanterns seems that only magic could make it. The scene of dropping lantern is believed that only magic can do it Table 8.4: Sample Sentences of the Erro r Type No Logical Connection between Subject and Predicate Year Sample Sentences Distribution 2 Miss To, who is beautiful but miserable, vagabond but dignified, 5 amorous but loyal, both seductive and serious, petite but not important is played the role best by no one except Thuy Nga Table 8.5: Sample Sentences of the Error Type Inappropriate Connection between Subject and Passive Verb It is observed that these errors are not the most frequent of all kinds of syntactic errors. That is to say, although students do have errors because of the typological difference in topic-comment structure between Vietnamese and English, they seem to have less difficulty with locating and translating subjects than with other fields, such as article and subject-verb agreement. In sentences in which the topic-comment structure is significantly different from the subject-predicate structure, students did have problem in locating the subject and made the above-mentioned five types of errors. However, they seem to have more problems in the task of handling the relationship between the subject and the verb: they are puzzled to recognize the difference between verbal predicates (in the form of a verb) and substantival predicates (in the form of an adjective) in Vietnamese and they often omit the verbs in the target text versions. This issue, however, is not the focus of the 218

232 WORDS IN ASIAN CULTURAL CONTEXTS present paper. Students also have difficulty in handling the relative clause in sentences with two topic-comment structures and have problems in constructing the sentence in an appropriate way. Suggested Strategies for Translating Topic-Comment Structures of Vietnamese Sentences in Vietnamese will be divided into two types: those with double su bject constructions (type 3, 4, 5 in Rosen s categorisation) and those without double subject (type 1, 2 in Rosen s categorisation). The examples will be taken from the Elicitation Task. 1. Sentences with double subject constructions of Vietnamese: the topic is not identical with the subject. Tam, cöû chæ thaät chính xaùc vôùi taâm traïng. Tam gestures so exact with mood. In this sentence the topic of the sentence is Tam and the subject of the embedded subject-predicate structure is cöû chæ (gestures); the relationship between them is possessor - possessee. There are three suggested ways to translate this type of double subject construction: 1) to use a noun phrase using possessive case Tam s gestures or a noun phrase his gestures ; 2) to use the phrase As for for the topic Tam and the subject of the embedded subject-predicate constructions cöû chæ will be retained as the main subject in the English sentence; 3) to choose Tam as the subject of the sentence and express the possessor-possessee relationship through the verb have ; the subject of the subject-predicate construction will become the object of the verb have. 1. Tam s gestures precisely reflect his mood. 2. As for Tam, his gestures precisely reflect his mood. 3. Tam has gestures which exactly reflect his mood. The referent of the topic can also be related in such a way to the referent of some other Noun Phrases in the comment, for instance the direct object of the verb in the comment clause. This type of relation is demonstrated by sentence 5 in the Elicitation Task. (5) Caûnh thaû ñeøn trôøi ( ) töôûng nhö chæ coù ma thuaät môùi laøm noåi. Scene drop light sky ( ) think only magic new do manage We/You/One may think that only magic could do such a scene of dropping lanterns. (The scene of dropping the lantern that we/one may think can only happen by magic) In this case, there are two possibilities for translating: (1) to start the sentence by the subject of the comment clause and to locate the topic Caûnh thaû ñeøn trôøi in the position of an object; (2) to locate the topic separately in a phrase at the sentence beginning As for the scene of dropping the sky lanterns, the viewers may think only magic could do that. However, in the sentences where the topic is the object of the verb in the comment clause but the main verb refers to perception processes such as look, seem, sound, taste, we do not need to locate a new subject for the sentence and separate the topic with as for. As English syntax allows such structures as The scene seems to be done by magic or The food tastes good, we can retain the topic as the subject of the sentence with these verbs as the predicate. To illustrate, in example 5, the main verb is 219

233 ASIALEX 2005 seem, and we can retain the topic Caûnh thaû ñeøn trôøi as the subject in the English translation. These are possible ways of translating this type of topic-comment structure in sentence 5: 1. We/You/One may think that only magic could do such a scene of dropping lanterns. 2. The scene of dropping the lantern that we/one may think can only happen by magic. 3. The scene seems to be done only by magic. 4. As for the scene of dropping the sky lanterns, the viewers may think only magic could do that. The referent of the topic can also be the adverb or adverbial clause to modify the verb in the comment clause (Type 4 suggested by Rosen). The example from the Elicitation Task is sentence 6. (6) Caûnh quay naøo (ngöôøi ta) cuõng thaáy moà hoâi. Scene any (..) also see sweat You/we/one can see the sweat behind every scene In this case, the subject of the comment clause has to be identified and located in the position of the subject in the English sentence. The topic will become the adverb phrase which we can place at the beginning or the end of the sentence, depending on what the focus of attention is. 2. Sentences without double subject construction of Vietnamese: the topic is identical with the subject of the embedded subject-predicate construction: Taâm caây ñaøn nguyeät tuyeät vôøi tôùi möûc tuyeät voïng Tam classifier guitar moon excellent to degree desparation In this sentence, the topic of the sentence Tam is semantically identical with the subject of the comment clause caây ñaøn nguyeät (player of moon-shaped guitar), which is a noun phrase to illustrate the identity of the topic Tam. This is the sentence type 2 suggested by Rosen. There are many possible ways of translating this structure. Firstly, the topic can be located at the beginning with As for, then the subject of the sentence must be a pronoun to avoid the repetition. Secondly, the topic will be Tam and the subject caây ñaøn nguyeät will become the appositive phrase or a relative clause. 1. As for Tam, he is a great moon-shaped guitar player who is so excellent to the extent of desperation 2. Tam, the player of the moon shaped guitar, is so excellent to the extent of desperation. 3. Tam, who is the player of the moon shaped guitar, is so excellent to the extent of desperation. When the topic of the sentence is exactly the same as the subject of the comment clause (as shown by Rosen s type 1), we choose the topic to be the subject of the sentence in English. Alternatively, As for can be used to block the topic and the same pronoun will be repeated at the beginning of the sentence. Toâi thì (toâi) khoâng nguû ñöôïc. I TM (I) not sleep manage I couldn t sleep As for me, (I) couldn t sleep. 220

234 WORDS IN ASIAN CULTURAL CONTEXTS References Anh, N. T. M. (2000). Vò töø Tieáng Vieät theo quan ñieåm ngöõ phaùp chöùc naêng. Baùo caùo taïi Hoäi Nghò Khoa Hoïc, Ñaïi Hoïc Khoa Hoïc Xaõ Hoäi & Nhaân Vaên TPHCM (Proceedings of university-level conference in September 2000, University of Social Sciences and Humantities of Ho Chi Minh City, Vietnam). Ban, D. Q. (1987). Caâu ñôn tieáng Vieät. Haø Noäi. Nhaø Xuaát Baûn Giaùo Duïc (Publisher of Education, Vietnam). Chinh, T.V & Leâ, N.H. (1973). Khaûo luaän veà ngöõ phaùp Vieät Nam. Hueá: Ñaïi Hoïc Hueá (University of Hue). Dyvik, H. J. J. (1984). Subject or Topic in Vietnamese? University of Bergen. Hao, C. X. (1991). Tieáng vieät sô thaûo ngöõ phaùp chöùc naêng. Nhaø Xuaát Baûn Khoa Hoïc Xaõ Hoäi (Publisher of Social Sciences, Vietnam). Rosen, Victoria (1998). Topic and empty pronouns in Vietnamese. Published Ph.D Thesis. Thanh, T.M. (2003). Caùc vai nghóa trong caâu traàn thuaät tieáng Vieät vaø tieáng Anh. (Different semantic roles in Vietnamese and English declarative sentences). Unpublished Ph.D Thesis. University of Social Sciences and Humanities of Ho Chi Minh City. Thompson, L.C (1965). A Vietnamese reference grammar. University of Hawai I Press, Honolulu. Multiple word class entries in advanced learners dictionaries of English Sadayuki Nakane University of Fukui nakane@edu00.f-edu.fukui-u.ac.jp Abstract This paper compares how converted or semi-converted words are treated by monolingual and bilingual (English-Japanese) learner s dictionaries of English and discusses their treatments of some problem cases mainly from a morphosyntactic viewpoint. They have traditionally attempted to structure the lexicon of English on the basis of word class distinctions. Although known to be fuzzy, these distinctions are extremely important for advanced learners understanding of syntactic structures. Provided they stick to using word class labels, lexicographers should be careful not to confuse learners in their understanding of English grammar. Unfortunately there are innumerable words that belong to more than one word class and there is also a good deal of overlap between word classes. This has produced borderline cases of lexical word class membership. It is not always easy to distinguish a syntactic composite nominal or noun phrase from a morphological compound noun (e.g. a concrete wall/a concrete mixer). Two structurally similar noun phrases (e.g. a garden chair/a huge chair) need to be kept away from each other, in terms of determining the lexical properties of the first constituents. The out in He was out when I called him 221

235 ASIALEX 2005 might be regarded as an adjective but one may also claim more reasonably that it is an adverb as it functions as an adverbial complement. Despite many demarcation problems of this sort, lexicographers should establish a more unified set of criteria for word class determination based on the common grounds on which most grammarians agree. It has been one of the traditions of lexicography to label each entry in terms of its word class or classes. This is perhaps because lexicographers believe that word class labelling is necessary for lexical categorization and useful to dictionary users in that it contributes to their expectation of the linguistic behaviour of a lexeme. The number of the identified word classes differs among grammarians, although eight categories have generally been recognized: noun, verb, adjective, adverb, pronoun, preposition, conjunction, and interjection. This classification has a long tradition, traceable to Latin grammar, and has been adopted by many grammars and dictionaries, though sometimes in modified form, with some classes renamed and others added. The word classes are to be divided into two broad categories, open and closed. Many grammarians seem to agree that the open category comprises four classes: noun, (full) verb, adjective, and adverb, while the rest are closed, so named because they are not usually open to new members. The closed word classes may be classified in different ways. However, present-day grammarians agree in recognizing at least eight classes: noun, verb, adjective, adverb, pronoun, preposition, conjunction, and determiner, the last two of which may, however, be named differently. Interjections and numerals are ignored by some grammarians as peripheral, but most think it worthwhile to recognize the class of auxiliaries as distinct from that of full or lexical verbs. The criteria for word class identification may be morphological, syntactic, or semantic, but a combined consideration of form, function, and meaning is necessary. Leech et al. (1984: 43) remark in this connection function is most important, form next most important, and meaning least important. If so, function should be prioritized but meaning may still play an important role in determining the word class of a word. The task of word class identification, however, is not as easy as it seems, for many lexemes overlap across word classes. There are indeed borderline cases of word class membership, and yet lexicographers are obliged to favour one classification over another by employing the criteria that may be regarded as most appropriate by as many grammarians as possible so as to help dictionary users to understand how a word actually does or could behave. Let us examine some cases, and how they are treated in dictionaries. Grey/gray, for example, is usually labelled as adj., noun, and verb by most learner s dictionaries, and there is no serious problem about this. Only CALD labels greying as adj., while the others treat it under the verb grey as one of its inflected forms. CALD gives this illustrative example: He is greying now but still elegant. This is, however, an unfortunate choice, because is greying might be felt as forming a verb phrase. One may insert such a phrase as a little after greying, which would indicate the verbness, not the adjectiveness of the form (cf. His hair is a little grey, where grey is an adj.). The adjective greying is usually used in attributive function, as in her long and greying hair or the greying baby-boomers. LDOCE4 enters the same form greying as noun (e.g. the greying of classical music audiences, where the -ing form may be regarded as a noun because it is preceded by a determiner and followed by a postmodifier in the 222

236 WORDS IN ASIAN CULTURAL CONTEXTS form of a prepositional phrase, a typical noun phrase structure). This is a nominal use of the -ing form of the verb grey, but we are not sure whether we should claim that the greying here has been converted to a noun. FAV, NCol7, and WIS list the form as noun. Perhaps it would be best for us to recognize three word classes of the lexeme GREY: adj., noun, and verb, greying being treated as an inflected form of the verb grey. It is difficult to say in some cases whether a word that comes in the premodifying position in a noun phrase is an adjective or a noun. It is a position that people tend to associate with adjectives because of their frequent occurrence there, but a noun may also appear in the same position. Crystal (1995: 211) analyses the garden party and by pointing out that it shows the morphological properties of a noun and an adjective, argues that garden should belong to a mixed word class. However, dictionaries do not assign garden to the adjective but to the noun class. It appears in such combinations as a garden chair, garden tools, a garden shed, a garden apartment/flat, a garden center, a garden city. We may divide these combinations into two groups: noun phrases or composite nominals (the first three) and compound nouns (the last three). A garden chair refers to any type of chair used in a garden, but a garden party is a formal social party held in a large garden, not any private party in a garden, and hence is now considered a compound noun. Generally speaking, compound nouns have a primary stress on the first constituent: a gárden pàrty, whereas in noun phrases a primary stress usually falls on the head: a gàrden cháir. It should be noted that all Japanese English-Japanese dictionaries indicate compound nouns by using primary and secondary stress diacritics. Garden in garden party must have been a nominal premodifier in origin, but has now become a base, and garden party should be labelled as noun. It would be absurd to conclude that gardens that appear in the position just before a noun are adjectives only because it is a position often occupied by adjectives. Garden-variety (e.g. garden-variety science, musical, criminals, cases of fraud) makes an interesting case for comparison. It is labelled as adjective by all dictionaries I have consulted, although its form might suggest that it may be a compound noun, and indeed, variety in a variety show or a variety store is considered by some dictionaries as a noun. The difference between garden-variety commonplace and variety a form of theatre or TV entertainment is that the latter may be used as a noun in various functions, but the former is used only as a premodifier having a sense equivalent to the adjective ordinary or commonplace. The original sense of garden-variety seems to be a variety (of flower, plant, etc.) that is grown in an ordinary garden, which would suggest its nounness, but as it is used metaphorically only in attributive function, it may be judged to be an adjective. Concrete is also a multiple word class word, being an adjective, a noun, and a verb. It has been used for centuries as an adjective semantically opposite to abstract, but upon gaining the noun sense building material in the mid 19th century, it went on to become an adjective with the related sense made of concrete. In a concrete building/floor/ wall/pavement, concrete may apparently be interpreted either as a noun or an adjective. It may be coordinated with a noun in the premodifying position, as in concrete and clay products, where clay is not normally recognized as an adjective by dictionaries. On the other hand, it may be coordinated with an adjective-equivalent, as in a concrete or tiled floor. Which word class should it be assigned to then? A secure test of determining membership of the adjective category is to check whether the target word occurs in predicative as well as in attributive function. The floor is concrete and These products are concrete are acceptable, and the meaning of concrete here is made 223

237 ASIALEX 2005 of concrete ; hence, concrete in a concrete building is an adjective. Note that concrete in a cóncrete mìxer is a noun base. A concrete mixer is a mixer of sand, water and cement to make concrete, not one made of concrete. How about night as used in a night train? It is not as easy as it appears to determine the word class of the form night. Night may occur freely in attributive position (e.g. night air, night light, night nurse), but unlike concrete it hardly ever occurs in predicative position (*Our train was night, *The nurse was night, etc.). Night in these examples is treated as an adjective by MED, LEXIS, and NCol7. Day may also be used in a similar way, as in day work, a day game, day temperatures, but day in this use is labelled as adj. only by LEXIS. Perhaps it would be more appropriate to include this use of night and day under the noun entry. The other Japanese bilingual dictionaries than LEXIS and NCol7 include this night under the noun entry, and interestingly they all add the explanation that it may be used adjectivally. However, attributively would be a more relevant term, because the explanation actually refers to function, not word class. As with many nouns in attributive function, the premodifier night is semantically very variable, its senses being freely determined according to what noun it premodifies. MED defines the adjective night as happening or existing at night, followed by four sub-senses used at night, travelling at night, working at night, and active during the night. None of these senses seems to exactly apply to the night in I ve spent half my windswept life recently on night roads (Cobuild4 English Wordbank). The question about a night out, which may appear to be structurally similar to a night train is different: which is the modifier, night or out? In other words, what is the head of the noun phrase? Deletion of night from a night out leaves an out, but the nominal out usually means an excuse or reason for avoiding an unpleasant situation (CALD), which is different from the meaning of a night out, where out means not at home. Out in this sense is frequently used as an adverb: Let s eat out tonight, which means Let s eat in a restaurant tonight. The out in a night out and that in Let s eat out are semantically identical. Originally a night out seems to have referred to a night spent out (i.e., not at home) at a restaurant, theatre, etc., in which case the noun head is night and the adverb out functions as a postmodifier, just as in the case of a way out. But whatever the original word formation of a night out may have been, night out is now felt as a compound noun rather than a noun phrase. People in general would agree that the prepositional complement of on in Marla Maples wore it on a night out with Donald Trump in New York (The Times, Jan. 1995) is not a night but a night out. There is also phonological evidence for this: a juncture exists between night and out, but there is no pause there. Thus, night out has been converted to a noun. We should be aware, however, that its plural form is nights out (not night outs), as in It s one of the many nights out organised by Flint (The Times, Jan. 1995). As these examples show, night out may occur in other functions than as a direct object, but three of the Japanese bilingual dictionaries I have used list it as part of the idiomatic or fixed expression have a night out. Incidentally, it is translated correctly ( gaishutsu-shite-yoru-o-tanoshikusugosu, FAV) or misleadingly ( ichiya-o-soto-de-asobi-akasu, NCol7 and SA3). The latter translation suggests people staying out all night and not going back home. Out is usually used as an adverb or an adverbial particle, as in Did you take that out or did I?, but is it an adverb or an adjective if it appears after a linking verb, as in It was already out actually? MED treats out as a function word, and gives the information that it may be used after the verb to be : You were out when I called (s.v. out), but the dictionary is silent about the word class it belongs to. Cobuild4, LDOCE4, 224

238 WORDS IN ASIAN CULTURAL CONTEXTS and OALD6 label out in the same use as an adverb, but COD11 and CALD regard it as an adjective. In the predicative position after a linking verb, an adverb (esp. of space) may occur, functioning as an adverbial complement (Greenbaum and Nelson, 2002; Quirk et al. (1990) call this an obligatory adverbial and Biber et al. (2002) a subject predicative ), and there is no well-grounded reason for deciding that such an out is an adjective. LUM is the only Japanese dictionary that claims adjectival status for this use of out. Its illustrative examples include They are out in the garden (s.v. out, adj. 2). It seems to me, in this particular sentence, that out is an adverb phrase and in the garden is a prepositional phrase, both serving as adverbial complements, and standing in apposition to each other. OALD4 does not recognize out as an adjective, but some outs (e.g. Maxi skirts are out now, The machine is out again ) are labelled as adj. by most dictionaries, usually with the usage note not before noun. The outs in these examples, however, are shortenings of out of fashion and out of order respectively, which normally work as adverbial complements. Out used attributively, as in the in and the out directions or Which direction is the out one?, does not occur predicatively, and is not fully adjectival. Out in the sense open about one s homosexuality (COD11), however, is known to work attributively as well as predicatively: an out lesbian politician / Are you out to your parents? (MED, s.v. out, adj.). LUM mentions this sense under the adverb out, with no example. Out in this connection seems to have come from some such expression as Their eldest son had come out (NDAS), where out is undoubtedly an adverb and He came out of the closet last year and his parents damn near died (NDAS), where out of may now be called a complex preposition, but originally out was an adverb and of a preposition meaning from. This slangy out is undoubtedly adverbial in origin, but as it has now come to be used not only in predicative but also in attributive function (though the attributive use still seems pretty rare), one may choose to label it as adjective. Borderline cases of multiple word class membership sometimes defy clear-cut classification of words, but lexicographers should not adopt a too simplified set of criteria for assigning the entry lexemes to their word classes when editing a dictionary, especially one designed for advanced learners. Such criteria have to be based on the grammar, if not codified, which the lexicographers believe in, but unfortunately learners are not provided with enough information on what grammar it is. Lexicographers cannot be too careful in their efforts not to mislead dictionary users, who may be learning grammar from their inappropriate word class labelling. Several good grammars are accessible as university textbooks (see References below), and even if they use different terms for the same linguistic phenomenon, they have a lot in common with regard to word class demarcation, and this should be used as a basis for making more uniform criteria for word class assignment. Appendix: Word class treatment of selected items in UK and Japanese dictionaries for advanced learners entry e.g. CALD Cobuild4 LDOCE4 OALD6 MED greying (1) He is greying. adj. (adj.) (v.) v. v. greying (2) the greying of n. garden a garden chair n. n. n. n. garden-variety garden-variety science adj. adj. adj. variety a variety show n. c.n. n. n. 225

239 ASIALEX 2005 concrete (1) a concrete wall n. n. adj. adj. adj. concrete (2) a concrete mixer c.n. c.n. c.n. c.n. night (1) a night train n. n. n. adj. night (2) a night out n. n.(idm) n. out (1) You were out. adj. adv. adv. adv. f.w. out (2) Is he out? adv. adv. adj. Table 1. Treatments in UK monolingual dictionaries for learners entry E-Gate FAV G3 LEXIS LUM NCol7 SA3 WIS greying (1) (v.i.) v.i. (v.i.) v.i. (v.i.) (v.i.) (v.i.) (v.i.) greying (2) n. (v.i.) n. (v.i.) n. (v.i.) garden n.(adj.) n.(adj.) n.(adj.) n.(adj.) n.(adj.) n. n.(adj.) n.(adj.) garden-variety adj. adj. adj. adj. adj. adj. adj. variety c.n. c.n. c.n. c.n. c.n. n. n. concrete (1) adj. adj. adj. adj. adj. adj. adj. adj. concrete (2) c.n. c.n. c.n. c.n. c.n. c.n. n./c.n. night (1) n.(adj.) n.(adj.) n. adj. n.(adj.) adj. n.(adj.) n.(adj.) night (2) n.(idm) n.(idm) n.(idm) n.(idm) out (1) adv. adv. adv. adj. adv./adj. adv. adv. out (2) adv. (adv.) Table 2. Treatments in Japanese bilingual dictionaries for learners References Biber, Douglas, Susan Conrad, and Geoffrey Leech (1999), Longman grammar of spoken and written English (Harlow: Longman/ Pearson Education). Biber, Douglas, Susan Conrad, and Geoffrey Leech (2002), Longman student grammar of spoken and written English (Harlow: Longman/Pearson Education). Booij, Geert (2005), The grammar of words (Oxford: Oxford University Press). Crystal, David (1995), The Cambridge encyclopedia of the English language (Cambridge: Cambridge University Press). Freeborn, Dennis (1987), A course book in English grammar (Basingstoke, Hampshire: Macmillan Education). Greenbaum, Sidney (1996), The Oxford English grammar (Oxford: Oxford University Press). Greenbaum, Sidney, and Gerald Nelson (2002), An introduction to English grammar, 2nd edn (Harlow: Longman/Pearson Education). Greenbaum, Sidney, and Randolph Quirk ((1990), A student s grammar of the English language (Harlow: Longman). Huddleston, Rodney, and Geoffrey K. Pullum (2002), The Cambridge grammar of the English language (Cambridge: Cambridge University Press). Huddleston, Rodney, and Geoffrey K. Pullum (2005), A student s introduction to English grammar (Cambridge: Cambridge University Press). Leech, Geoffrey, Margaret Deuchar, and Robert Hoogenraad (1984), English grammar for today (London: Macmillan). Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik (1985), A comprehensive grammar of the English language (London: Longman). Sinclair, John, et al. (1990), Collins Cobuild English Grammar (Glasgow: HarperCollins). Dictionaries (listed by abbreviated titles) CALD: Woodford, Kate, and Guy Jackson, managing eds. (2003), Cambridge advanced learner s dictionary (Cambridge: Cambridge University Press). Cobuild4: Sinclair, John, founding ed.-in-chief (2003), Collins Cobuild advanced learner s English dictionary (Glasgow: HarperCollins). 226

240 WORDS IN ASIAN CULTURAL CONTEXTS COD11: Soanes, Catherine, and Angus Stevenson, eds. (2004), Concise Oxford English dictionary, 11th edn (Oxford: Oxford University Press). E-Gate: Tanaka, Shigenori, et al., eds. (2003), E-Gate English-Japanese dictionary (Tokyo: Benesse). FAV: Asano, Hiroshi, et al., eds. (2002), Advanced favorite English-Japanese dictionary (Tokyo: Tokyo-shoseki). G3: Konishi, Tomoshichi, and Kosei Minamide, eds. (2001), Taishukan s Genius English-Japanese dictionary, 3rd edn (Tokyo: Taishukan). LEXIS: Hanamoto, Kingo, et al., eds. (2003), Obunsha Lexis English-Japanese dictionary (Tokyo: Obunsha). LDOCE4: Summers, Della, dir. (2003), Longman dictionary of contemporary English, 4th edn (Harlow: Longman/Pearson Education). LUM: Takebayashi, Shigeru, et al., eds. (2001), Kenkyusha s Luminous English-Japanese dictionary (Tokyo: Kenkyusha). MED: Rundell, Michael, ed.-in-chief (2002), Macmillan English dictionary for advanced learners (Oxford: Macmillan Education). NCol7: Takebayashi, Shigeru, et al. (2003), Kenkyusha s New College English-Japanese dictionary, 7th edn (Tokyo: Kenkyusha). NDAS: Chapman, Robert L., ed. (1987), New dictionary of American slang (London and Basingstoke: Macmillan Press). OALD6: Wehmeier, Sally, ed. (2000), Oxford advanced learner s dictionary of current English, 6th edn (Oxford: Oxford University Press). SA3: Yamagishi, Katsuei, ed.-in-chief (2003), The Super Anchor English-Japanese dictionary, 3rd edn Tokyo: Gakushu-kenkyusha). WIS: Inoue, Nagayuki, and Ichiro Akano, eds. (2003), The Wisdom English-Japanese dictionary (Tokyo: Sanseido). Language rationalization of technical terms in Asian ICT perspective Rhoderick V. Nuncio De La Salle University-Manila, Philippines rhodnuncio@yahoo.com Abstract There is an increasing number of new words within the practice and domains of information and communication technology today. These words typically lend hands in constructing a technical language diffused in consumerist societies understood one-sidedly by its creators/producers/owners of technology. These words if compiled and studied reveal the following categorizations: (1) technospecialist (2) technoacademic (3) technoeconomic, and (4) technocultural. In order to construct users/consumers own meanings of technical terms as they interface with gadgets and other technological tools they negotiate and invent words to approximate their understanding of the practice and the experience. On the one hand, the increased technical rationalization of modern society has paved the way to reflexive construction of knowledge of which technical and everyday languages, on the other hand, are more disparate and widely delineated than ever. However, it is obvious that functionality and sophistication of these machines like computers, cellphones, handhelds decrease our understanding of its technicality. A mere utterance of WAP suggests nothing but a buzzword in cellphones and palmtop use. It is understood as a feature and bears no 227

241 ASIALEX 2005 lexical meaning as far as the users/consumers are concerned. It is a sign of ICT s complexity. Hence, this paper explores the epistemological gap between technical and everyday languages and relates this assumption with ICT s rapid innovations/evolutions and society s modernization. How do technical words fulfill the vision of a common language, a rationalized language, as Asians embrace technology? Does it lead to that direction given the present technological condition? Or does it command a new direction to be explored due to ICT s accelerated complexity and rationality? Introduction Technological innovations are involved in the increasing rationalization of social practices. The rate of change evident in a society driven by Information and Communication Technology (ICT) is tremendous. For only about two decades, people are exposed and are led towards a new generation of technological way of life with the upsurge of computers, the Internet and mobile phone use. The precious commodity today is information. In fact Albert Borgmann (2000) pronounces that this is the age of technological information, i.e., information as reality. Efficiency, calculability and speed are basic engines of technological rationality. The blueprint for these wonders is framed by ICT. What is apparently happening is the novelty of appearance of knowledge space, which dramatically brings about the phenomenon of information revolution. Orality was the carrier of unmediated information, until the ecriture (the system of writing), the printing press, radio, television, and computer came. Now, information traverses wide space of actual and virtual communities within a borderless world in just a short time, in real time. Using available technology today, this amount of information, however, is mostly communicated in technical terms. Hence, a new language is being socially constructed, as ICT is being diffused and popularized at the same time to meet the demands of social change. It is at this rate of hyperchange that language is also deemly rationalized to swim through the paths of ICT, that is, to be complicit too with the demands of codifying (encoding, decoding) and translating new ways of expressions and communications. This is what the paper intends to share by mapping a chink in the domain and practice of language use of ICT technical terms in relation to the rate of technological change. The domain is typically related with modernization and rationalization laying down the groundwork for the possibility of ICT, whereas, the chink in that domain points to a need for a de-rationalized language of ICT in Asian perspective. The process of language rationalization of technical words If technology should have been the key towards alleviating the condition of man, its democratization in terms of accessing, understanding, communicating, utilizing and distributing it, should correspond to the level where its growth becomes expansive across cultures, nation-states, race, gender, class and time. Its democratization should start with forging a common language a vision that can be communicated today and in the future, and that can be shared and utilized for human development and societal advancement around the globe. Al Gore describes this in part as the rising standards of living and literacy, and an even widening circle of democracy, freedom and individual empowerment (Leer 1999: 7). Technology should itself be a language, a common language to all. This is the aim of the rationalization of technical language. 228

242 WORDS IN ASIAN CULTURAL CONTEXTS Accessibility to that vision necessitates familiarizing with technical use through learning and mastery of technical words in meaningful discernment and actions. How do technical words fulfill the vision of a common language, a rationalized language? Does it lead to that direction given the present technological condition? First and foremost let s discuss the basic element of this language: technical words. Here is a simplified definition of technical words: these are specialized terms bearing material, ethereal and operational concepts about systems, processes, products, structures and services made possible by or through the use of technology. Concepts Material (physical attributes) Etheral (nonphysical attributes) Operational (functionality) Words/terms Computer, monitor, mouse, keyboard Cyberspace, MUD, blog, VR games MS word, , chat, sync This brings people closer to what we call as technical knowledge through the mediation of technical language and its relation to a technological reality. Technology is always connected with gadget, machines and equipments. It is first and foremost a material creation. Technology talks about a purpose; moreover, it gives answer to perennial questions/problems of man. It becomes a social condition when it is communicated, when it becomes part of the lives of many people. Technical words are by-products of its corresponding reference to the material inventions/creations with these following categories: its physical/nonphysical attributes (what it is), its operations (what it offers, how it works), its users (who s the intended user), its output/benefit (what it does), its limitation (what it can t do), its extent (how far it can work). There are four ways in considering how technical words and technical language evolved: 1. Technospecialist domain independent-private, commercial, or state-funded explorations, experimentations and inventions of new technological systems, processes, structures, services and products for the future. Experts, scholars, scientists, engineers and specialists converge to bring into life scientific and technological discoveries and inventions that answer present needs and anticipate future worries. Information on this regard is highly classified, restricted to few personnel and not yet for media consumption. An example of this is the birth of the Internet, which was originally clouded in secrecy by the military and the academe in the United States in the 1960s (Dery 1996). The need at time was to maintain military communication in the event of a nuclear attack (future anxiety). Technical words are very specialized in this regard like downlink, http, speech codex, chip rate, etc. (Lightman 2002) 2. Technoacademic domain public sphere, opening the gates of selective and premium results of research to the public, which is actually reserved to academics. It oftentimes invites debates and confirmation or negation of claims of new discoveries in scholarly journals, in conferences, symposia and lectures. As it survives the rigorous academic test, technical knowledge is passed on through teaching, dissemination, and evaluation. Technical words are used in academic discourse. 3. Technoeconomic domain national to global sphere: patenting the invention, copyright of intellectual properties, mass production, distribution and consumption. Technology enters the domain of popular culture and politics mediated by State intervention or non-intervention and by media hype with loads of publicity gimmicks. Hence, technical words constitute a market-driven/informalized language in 229

243 ASIALEX 2005 consumerist societies (Shortis 2001). 4. Technocultural domain inner sphere, sphere of consumption, appropriation, internalization (Shortis 2001). ICT users and consumers negotiate and invent words to approximate their understanding of the practice and their experience as they interface with gadgets and technological tools. It is this domain that translating technical words as part of/ as ordinary language and other forms of language (literary, humanistic, creative) becomes evident. Thus, technical words meaningfulness rests on the attributes and the functionality of referents being used and understood within technical/technological conventions. It is therefore arbitrarily coined and exclusively used initially by specialists in the fields of science and technology. Furthermore, as technicality reinforces functionality and utility of technology, and less its complexity, it therefore, rationalizes human and social actions, expectations, behaviors, and conditions. A computer becomes technical because it is functional and useful, however, as soon as it chunks out, it ceases its inherent technicality it becomes an artifact with meaningless and unpurposive end. Hence, technical words common in ICT terminologies face an indeterminate erasure in each passing time as ICT revolution accelerates. The trend today is this: as ICT accelerates its technicality (upgrading and ceaseless innovation of functions and utility), the more it transforms to complexity, the more information overload comes in our way, the more technical words become meaningless, and the more people lose the direction to map out what lies ahead. technospecialists Producers and inventors of New technologies technoacademic technoeconomic Users of ICT are narrowed down and become exclusive with higher economic & cultural capital technocultural Figure 1. Technicality pyramid- Hierarchy of language rationalization of technical words Figure 2. Complexity inverted pyramid as a digital divide Complexity and derationalization As the level of ICT s rapid change and innovation increase in just a short time, the amount and kind of knowledge gained or expected from it contracts or diminishes in the marginalized sectors or in new growth areas of developing countries (like Asia). In this scenario, while the rest of the Western World and NIC (newly industrialized countries) welcome the dawn of the new millennium with the advent of 3G and 4G technologies; unfortunately, other parts of the world are still lurking in limbo: no 230

244 WORDS IN ASIAN CULTURAL CONTEXTS telephone lines, no Internet, no information revolution. Innovations must deccelerate so that developing countries could catch up. A greater space of digital divide is created if these developing countries are forced to embrace the present technologies. Technical words become useless as its referents are phased-out because of the fast moving upgrades and innovations in hardware and software technologies. A formula of one step back and one step up means considering the potential need to know and use immediate past technologies in relation to existing ones and how this would have productive effects in anticipating and using new ones. Another scenario is this: technological complexity takes over the lives of people and not the other way around. This is what Neil Postman (1993) calls as the surrender of culture to technology. It becomes a way of life where human beings could not live without technology, no matter how complex it is. Donald Norman (1998) also points out the complexity of computer in the age of ICT. According to him: There are three major reasons for the complexity: the attempt to make a single device do too many things; the need to have a single machine suffice for every person in the world; and the business model of the computer industry. (Norman 1998: 77) Related to the third reason of complexity, the best way to sell technogadgets is to leave room for curiosity and ignorance. People will become curious as media hype reign the publicity of new products. As a result, people become ignorant as they become compulsive buyers without knowing what they are actually buying and why they are buying it. Therefore it incites false curiosity and dumbfounded ignorance of mass consumers. The technoeconomic domain takes over and controls the direction of ICT as it detaches from technospecialist and technoacademic domains. What we have now is an information marketplace, which is: the collection of people, computers, communications, software, and services that will be engaged in the intraorganizational and interpersonal informational transactions of the future. These transactions will involve the processing and communication under the same economic motives that drive today s traditional marketplace for material goods and services. (Dertouzos 1997: 10) However, we have an information or ICT marketplace without a standard, singular, global information infrastructure. Ironically this is what Dertouzos (1997: 16) admittedly proclaimed in his book, What will be; as well as in Lightman s (2002: 27), Brave New Unwired World; and in Alvin Toffler s essay, Shocks, waves and power in the Digital Age (Leer 1999: 23). And since, the major economic and political gatekeepers of ICT are the highly industrialized and developed countries, the distribution and acceleration of ICT revolution and evolution are selective, narrowed down and exclusive to those who might have the capacity to buy out or implement technologies from multinational companies of these countries. Developing countries mostly from Asia and Africa have no choice but to subscribe to one-sided ICT infrastructure and to be coerced in embracing the present technologies. In relation to this, language rationalization has manifested the following conditions: artificial rise of English as the standard medium of ICT and the increasing complexity of ICT. There is a tendency to have a wider epistemological gap between technical language in English and ordinary language use in Asia. Technological innovations and inventions originated mostly from the Western countries and hence it demanded the imposition of English as the language of ICT. However, innovations and inventions are not anymore the monopoly of the West. Asian countries have also earmarked on the technicality pyramid 231

245 ASIALEX 2005 of improving and competing with American and European technologies. 1. Japan, China and South Korea have earned respect from global market as they develop their own technospecialists reshaping the global information infrastructure (Lightman 2002: ). 2. Singapore and Japan boasted academic centers of excellence in the fields of science and technology increasing the public sphere of expert discussion and academic/scholarly exchange in the region. 3. A dirigiste approach or the State interventionist model in policymaking, implementation and improvements of ICT in Singapore, Japan, Korea, Taiwan, China has produced alternative and viable program of actions as against the neo-liberal approach of the West (Moore 1998: ). 4. Technical language and knowledge are derationalized as translation, codification and internalization of technical words become evident in China using Mandarin, in the Philippines using Filipino and other Asian countries in their respective lingua franca. There is a movement of derationalizing the language of ICT to slim down the digital divide and to popularize/democratize ICT in local communities. 5. More popular use of cellphones and the Internet in Asia have led to technocultural domains of inventing and appropriating new words as Asians interact with these technologies. Conclusion The increasing complexity of ICT and the breakdown of an English-based ICT language have led to derationalization. Asian countries are continuing to become potent partners in attaining the vision of ICT and not just as consumers of surplus technologies from the West or from developed countries. It would need more time to realize this as Asia s role becomes significant in slowing down the acceleration of technological change. It is reshaping the information highway or dismantling the ICT marketplace as it redistributes the effects, benefits and wonders of ICT. Soon enough, Asia would become a major playing field in closing the gap of digital divide and in increasing its stakes in technicality beyond complexity. The derationalization of ICT language brings forth greater heights and higher leaps of technological change which center on meaningful actions towards human development and societal advancement. This is the heart of an ICT perspective/vision in Asian context. References Borgmann, Albert (2000 ), Holding on to reality: the nature of information at the turn of the milennium. (University of Chicago Press). Dertouzos, Michael (1997), What will be. How the world of information will change our lives. (London: Judy Piaktus Limited). Dery, Mark (1996), Escape velocity: cyberculture at the end of the century. (Hodder & Stoughton). Gore, Al (1999), Putting people first in the information age, in Anne Leer (ed), Masters of the wired world: cyberspace speaks out. (London: Pearson Education Limited), pp (Ch.1). Lightman, Alex with William Rojas (2002), Brave new unwired world: the digital Big Bang and the infinite Internet (New York: John Wiley & Sons, Inc.). Moore, Nick (1998), Confucius or capitalism? Policies for an information society, in Brian Loader (ed), Cyberspace divide: equality, agency and policy in the information society (London and New York: 232

246 WORDS IN ASIAN CULTURAL CONTEXTS Routledge), pp (Ch.9). Norman, Donald A. (1999), The invisible computer (Cambridge, Massachusetts: The MIT Press). Postman, Neil (1993), Technology: the surrender of culture to technology (Vintage). Shortis, Tim (2001), The language of ICT (London: Routledge). Toffler, Alvin (1999), Shocks, waves and power in Digital Age, in Anne Leer (ed), Masters of the wired world: cyberspace speaks out. (London: Pearson Education Limited), pp (Ch.1). Electronic dictionaries in the classroom Shinya Ozawa Hiroshima Shudo University James Ronald Hiroshima Shudo University Abstract In some Asian countries, and in countries where many Asian students are studying, electronic dictionary ownership is reaching levels where it can no longer be ignored. The use of these dictionaries in the language classroom is affecting learner behaviour in the classroom in at least two important ways: in how learners deal with the language and in their perception of their language teacher. This changed learning environment is also affecting how language teachers feel about the presence of electronic dictionaries in the classroom and how they respond to the changed circumstances. This paper reports an in-depth survey of English teachers attitudes and behaviour regarding electronic dictionaries. The survey was conducted among both Japanese and native-speaker English teachers at a mid-ranking Japanese university. Analysis of survey results has revealed a wide range of attitudes and responses to these dictionaries that will be of interest to language teachers, and to those involved in the making and marketing of dictionaries. In particular, there are revealing differences according to whether the teachers are electronic dictionary users themselves or whether they share the mother tongue of their students. There are also significant differences attributable to teachers ages. Based on findings from the survey, suggestions will be made for how teachers can respond to the opportunities and challenges offered by the presence of electronic dictionaries in the classroom. Specifically, guidelines will be offered to minimize the potential of electronic dictionaries as barriers to communication and learning, and to maximize their potential as aids to language learning. Introduction In recent years, especially in a number of Asian countries, we have witnessed unprecedented student ownership of, and widespread reliance on, handheld electronic dictionaries (henceforth ED) in the language classroom. In many Japanese university language classrooms, for example, ownership has reached over 80% (Kobayashi 2004). Despite the important position that the ED has claimed in the classroom, and the way that it may affect so many aspects of language teaching and learning, this classroom revolution has largely been student-led, with the language teacher often little more than a passive bystander. Teachers do, however, devote much thought to the effect of these 233

247 ASIALEX 2005 dictionaries on different aspects of the language classroom, whether individually or in staffroom conversations with other teachers. The aim of the research described in this paper has been to gather these private beliefs and opinions, analyse them, and consider their implications for language teaching. There are two main reasons for pursuing this goal: Teachers vary widely in their attitudes and behaviour towards ED use in the classroom, and these differences may depend in part on factors such as the teacher s age, native tongue, or teaching environment. As teachers, we have a lot to learn from each other, not least understanding and appreciating each other s differing perspectives. While some teachers may feel that to interfere with students ED use in the classroom is to impede learner independence, as language teaching professionals we do need to be able to offer informed guidance to students who very often may not know how best to learn a foreign language. To be able to do this, we need to know about the different issues raised by the presence and use of EDs in the language classroom. This paper begins by reporting a survey of the attitudes and practices of university-level language teachers with regard to handheld electronic dictionaries. We will describe how the survey was conducted and present an analysis of the results. This will be followed by a discussion of the findings of the survey in which we will consider the consequences of the various issues raised by these as they affect language learning, classroom roles and dynamics, and teacher development. The Survey As Hatherall (1984), Lew (2002) and others have noted, questionnaire-based research into dictionary use is not always the best means of investigating dictionary use. At the same time, questionnaires do provide an important means of gathering information, especially attitudes, from large numbers of people. With reference to Dörnyei s (2003) general guidelines for questionnaire design, and Lew s (2002) specific recommendations in the context of dictionary use, a questionnaire for English teachers was prepared in both Japanese and English. It was composed of three sections: questions which would provide a profile of the respondent, including their ownership and use of EDs; questions about the respondent s attitudes towards EDs in the classroom; and questions about the respondent s behaviour with regard to student ED use in the classroom. Both Japanese and English versions of the questionnaire were piloted then revised with the help of university-level English teachers not involved in the survey itself. The survey was conducted of all the English teachers, both full-time and part-time, at Hiroshima Shudo University, a middle-ranking Japanese university. Questionnaires were distributed to each of the 56 Japanese teachers of English and 23 non-japanese teachers of English, with all but one of the latter being native speakers of English. Response rates were 50% for Japanese teachers and 61% for non-japanese teachers. While these response rates were disappointingly low, there is still much that we can learn from this survey. We will begin by briefly summarising the profiles of teachers who responded before going on to focus on the behaviour and attitudes of the teachers towards student ED use. After this, we will investigate what relationships there may be between individual teacher profiles and behaviour or attitudes regarding EDs in the classroom. 234

248 WORDS IN ASIAN CULTURAL CONTEXTS Finally, we will look at Japanese and non-japanese perceptions of the merits and demerits of EDs in the classroom. Respondent profiles are summarised in Table 1, with figures for Japanese (J) and non-japanese (N-J) respondents presented separately. At once we can see that there are major differences between Japanese and non-japanese teachers in terms of their typical ages, sex, subjects taught, and dictionary ownership. Most notably, over half the Japanese teachers possess and use an ED, usually of a model similar to that used by their students, while few of the non-japanese teachers have access to such dictionaries. J N- J J N-J J N-J J N-J Age Sex Subject Own ED 20~ 1 0 M Language Yes ~ 6 6 F 13 2 Literature 8 1 No ~ 5 5 Culture ~ 8 3 Linguistics ~ 8 1 Other 4 0 Totals Table 1. Profiles of ED survey teachers The next issue is that of teacher behaviour regarding ED use in the classroom. Here we will seek to identify correlations between behaviour and other factors, focusing mainly on the responses of Japanese teachers. Teacher ED ownership and use in class is an important factor, with 15 of the 28 teachers owning an ED and 6 of these using the ED in class. As we can see from Table 2, there is a clear relationship between whether teachers use their EDs in class and what they use their EDs for. ED-owning teachers who don t bring their EDs to class typically only use it for simple or basic uses such as checking meanings or spellings while those who bring their EDs to class all use the dictionary for more elaborate or sophisticated purposes such as checking collocations or example sentences. Teachers ED use purposes Elaborate Simple N/A Totals Teacher ED in class Yes No N/A Totals Table 2. Japanese teachers ED use There is a similar relationship between teachers use of EDs in class and teachers regulations of students using EDs in class. As we can see from Table 3, teachers who use the EDs in class allow their students to do so as well, while those who do not are much more likely to restrict or forbid student ED use. 235

249 ASIALEX 2005 Teachers regulation of student classroom ED use Allow It depends Forbid N/A Totals Teacher ED in class Yes No N/A Totals Table 3. Japanese teachers regulation of student classroom ED use We might expect that teachers who use and allow the use of EDs in class would also be likely to provide ED use guidance or training for students. This is true in some cases but as Table 4 shows, as many do this as do not. Interestingly, a more significant correlation is between the age of the teacher and dictionary training, with 5 of the 9 Japanese teachers who provide some kind of training being in their sixties. This tendency is also found among older non-japanese teachers, with the only two who do provide training being in their fifties or sixties. Dictionary use guidance Yes No Totals Teacher ED in class Yes No N/A Totals Table 4. Japanese teachers provision of dictionary use guidance We will now go on to look at teachers attitudes towards EDs in the classroom. Specifically, we will focus on the perceived merits and demerits of student ED use. Answers from Japanese teachers were given in Japanese and summaries are presented here in English. The most often cited merits of EDs for both Japanese and non-japanese teachers are listed in Table 5. The numbers in brackets indicate the number of respondents giving a particular answer. Here, the answers given by Japanese and non-japanese teachers are largely similar, with both most commonly citing convenience to carry and speed of access. One difference between the two groups of teachers, expressed in relation to ease of use and ED contents, is the Japanese teachers familiarity with the EDs the students use. The opposite is evident for many non-japanese teachers from their use of hedges, not included in these summaries, such as I guess, appear to be or from what I ve seen. A further difference, although only cited by two non-japanese teachers, is the issue of learner autonomy. 236

250 WORDS IN ASIAN CULTURAL CONTEXTS Japanese teachers Easy to carry (10) (they bring it to class) Quick for looking up (9) (saves time, students can look up a lot) Easy to use (5), like a mobile phone (4) (used more often, like a game, satisfying) Contact with various dictionaries (6) (easy to consult various dictionaries, good for preparation) Non-Japanese teachers Quick for looking up (9) (easier than paper, more words in less time) Easy to carry (5) (most students won t bring heavy paper dictionaries) Learner autonomy (2) (students have control over a very useful tool) Table 5. Perceived merits of EDs in the classroom As we go on to consider the perceived demerits of EDs in the classroom, we can see, in Table 6, that the perspectives of the two groups of teachers give them very different perceptions of the demerits of classroom ED use. The two main concerns of the Japanese teachers relate to the experience of getting lexical information from the ED and to the effect of ED use on vocabulary acquisition. Again, we can see that these are the views of teachers many of whom are themselves ED users. The widespread consideration given by Japanese teachers to vocabulary acquisition-related issues is also related to these teachers own experience as both teachers and learners of the foreign language. This will be considered in more detail below. For the non-japanese teachers, the majority of answers regarding demerits are concerned with the practice of communicating in English: whether the interruption of classroom conversation by ED use, or the inability to use or develop other communication strategies when students resort to ED use without reflection. A further issue includes implicit recognition, and criticism, of the fact that most ED use in Japan means bilingual dictionary use. The criticisms, then, are both towards the unwelcome intrusion of ED use into classroom communication and towards the deleterious effects of the repeated intrusion of Japanese into English learning environments. It is interesting to note that, in contrast with the Japanese teachers, there is little direct reference by the non-japanese teachers to the effect of ED use on L2 vocabulary acquisition, although the objection to Japanese in the classroom is, in part at least, due to the perception of the L1 as an obstacle to acquisition of the target language. 237

251 ASIALEX 2005 Japanese teachers Difficult to get the full picture on the small screen(8) (only see top of entry, not example sentences, grammar, words either side) Easy but no retention (6) (no vocabulary growth, no effort to memorize) Use without thinking (4) (paper dictionary use time is important: learning depends on time and effort) Too convenient (4) (fuzzy spelling inhibits learning, not quick for example sentences, idioms) Non-Japanese teachers Disrupts flow of class/conversation (4) (breaks flow of conversations, focus on English) Hinders students creativity, natural self-expression (4) (don t try to use vocabulary they know, EDs discourage Ss from finding a natural way to negotiate generation of ideas) Dangers of overuse, dependency, too much Japanese (4) (easy to use, so danger of automatic/reflexive use, don t guess meanings from context) Discussion Table 6. Perceived demerits of EDs in the classroom We will begin by considering the survey data from the perspective of the changed learning environment brought about by the presence of the ED in the language classroom. We will then reflect on the consequences that these changes may have for the role of the language teacher. As Kent (2001) observes, the presence of the dictionary in the language classroom represents a challenge to the authority and language expertise of the teacher. This is arguably even truer of the ED, considering its greater presence and widespread use. The issues of authority and expertise affect both Japanese and non-japanese teachers, but they differ in how it affects their status and how they respond to these changed circumstances. The Japanese teachers widespread ownership and use of EDs give them a greater understanding both of the ED and of their students ED use. One way by which these teachers may regain the status of classroom expert is by being an ED specialist more familiar with the technology in the students hands than the students are, and more knowledgeable about how to best use the ED as a tool for learning the foreign language. In this way, although the teacher may lose the role of all-knowing linguistic authority, this is exchanged for that of a model language learner for the students to emulate. Another way that Japanese teachers may retain authority is by maintaining control of ED use in the classroom, coupled with language learning expertise about how, when, and where to gain the greatest benefits from ED use. For the non-japanese English teacher, solutions are not so easy to find. Twice alienated by the ED It s the students generation s technology and it s in their language responses have been various and many teachers are still not sure about how to respond to this alien/alienator in the classroom. One worthwhile response may be reasoned, and explained, guidance for ED use according to the classroom language activity. Another could be to insist on monolingual ED use; this would make dictionary use less easy and automatic and so prevent overuse. It may, arguably, also make classroom ED use more beneficial and memorable. Finally, ED training is worth considering for both Japanese and non-japanese 238

252 WORDS IN ASIAN CULTURAL CONTEXTS teachers, especially in environments where ED ownership is approaching 100%. Learner autonomy is an important goal, but it only becomes more than a fashionable catchphrase when students learn how to make best use of the resources available to them. As teachers themselves learn more about the values and weaknesses of classroom ED use, they will be more able to instruct their students in using their electronic dictionaries wisely. References Dörnyei, Zoltàn (2003), Questionnaires in second language research: construction, administration, and processing (Mahwah, NJ: Lawrence Erlbaum Associates). Hatherall, Glyn (1984), Studying dictionary use: some findings and proposals, in Reinhard Hartmann (ed), LEXeter 83 proceedings: Proceedings from the international conference on lexicography at Exeter, 9-12 September 1983, (Tübingen: Niemeyer), pp Kent, David (2001), Korean university freshmen s use and perceptions regarding dictionaries, The Korea TESOL Journal 4(1): Kobayashi, Chiho (2004), Pocket electronic or printed electronic dictionaries?, paper presented at the national conference JALT 2004, Nara, Japan, November Lew, Robert (2002), Questionnaires in dictionary use research: a reexamination, in Anna Braasch and Claus Povlsen (eds), Proceedings of the tenth EURALEX international congress, EURALEX (Copenhagen: Centre for Sprogteknologi, CST), pp Words, meaning, speech acts and vocabulary in Far East cultural contexts Sang-Jun Park Hongik University, South Korea jeremypsj@hanmail.net Abstract A wider understanding of the word requires taking into consideration all contexts in which a word could be employed. A simple correspondence, suggested frequently in many foreign language manuals and beginner-oriented two language dictionaries, could not help distorting not only the realities reflected in the word, eventually in the discourse, but also the messages aimed at in the linguistic activities, which compels us to widen and to enrich the definition frame of the vocabulary of a concerned language. What this paper aims at is to give an example of this approach to the words with special reference to the Korean word mal, word, speech, tongue, language and to other words in the (para-)synonymic relations. Their connotation network reflecting Korean social hierarchy, Confucian tradition and epistemological evaluations, distinguishes the Korean language from Japanese, which shares many syntactic and semantic properties, and from Chinese, which has 239

253 ASIALEX 2005 enormously influenced Korean in its vocabulary. The shared cultural elements, for example, honorific distinctions between (para-)synonyms of mal, contribute to a profound understanding of the semantic network that defines the cultural boundaries testified in Korean, Japanese and Chinese. However Korean s proper aspects, for example Korean s particular conceptualization of social relations and speech acts, could lead to a specific understanding of Korean language within the Far East cultural background. Finally this paper will provide a trial comparative research and corpus of equivalents of the Korean word mal in three languages mentioned above, followed by a brief list of French and English words having the same meaning. Introduction The influence of social hierarchy is one of the most remarkable linguistic phenomena, especially in Korean language. It determines not only the inflectional verbal forms but also the vocabulary choice in quotidian conversation. This paper is a short version of the work including a trial comparative corpus of equivalents of the Korean word Mal, word, speech, tongue, language in two Far East languages Japanese and Chinese, and in two European languages French and English. The similarities between Korean and Japanese sometimes make us overlook the subtle differences reflecting the particularities of the two cultures. On the other hand, the enrichment of Korean vocabulary via Chinese words leads to the lexical specialization which has enlarged the vocabulary signifying speech act. In examining this lexical enlargement of the Korean word mal, we could put forward the assumption that semantic particularities of (para-)synonyms of mal should be formed according to Korean cultural particularities. Among many culture related Korean particularities, the social hierarchy is considered to be one of the most impacting pragmatic factors. Given the semantic universality of mal, it will be very fruitful to scrutinize its interrelations with linguistic honorification devices such as honorific suffixes, case markers, verbal inflections and lexical choice. It is generally accepted that the honorific expressions in Korean are classified into three categories for subject referent, object referent and hearer 1. With the view of giving a coherent explanation, precedent works on Korean honorific expressions have been oriented toward verbal affixation by specific morphemes, for example -si- or on several honorific markers and correlating postpositions. However the semantic distribution of nouns, at least in Korean and Japanese, demands that the honorific systems should be taken into account for an extensive lexical description, which will be useful in academic linguistic researches and also in foreign language teachings. Mal, its (para-)synonyms and their semantic network The dictionary definition of the Korean word mal can be given briefly as (the letters representing) the sounds expressing human thoughts or feelings. However its 1. For another categorization of honorifics, some linguists propose the addressee honorific (Geertz, 1960), the bystander honorific (Dixon, 1972) and the referent honorific (Levinson, 1983). 240

254 WORDS IN ASIAN CULTURAL CONTEXTS pragmatic meanings are rather extendable, including different types of speech acts and its correlated meanings as shown by the following comparative table between Korean expressions employing the word and English equivalents 1. hangukmal Korean language seoulmal Seoul speech (dialect) dongjeongui mal words of sympathy gasidochin mal a caustic remark hal mali eopda have no say mal junge in the course of a conversation maleul dolrida divert the talk change the subject almajeun mal An appropriate word/expression tteodoneun mal a rumor yangjjokeu mal the cause of either party maleul ttareuda follow one s directions/advice Table 1. The semantic distribution of mal On the other side of this semantic enlargement, however, we notice a lexical extension of mal. Proportionally to the importance of the speaker, Korean has a set of words for the same meaning speech : mal, malsseum, goron, goui, samieong. We present them in accordance to the degree of each word s honorificity. This set of vocabulary shares the same semantic core, speech act or content of speech, but each word doesn t appear in the same distributional environment. The honorific frame of Korean is, integrating the three honorific categories mentioned above, simplified as follows: [NP 1 -hon.suff. -hon.nom.marker NP 2 -acc. marker V-si-(p)ni-term.infl.] If we take temporarily the established explanation for Korean honorific system among Korean linguists 2, 1) the subject referent honorific is taken charge of mainly by a specific suffix attached to NP1 and the corresponding honorific normative marker, -kkeseo-. 2) The second category, object referent honorific, is realized by the lexical choice for NP2 and corresponding verbs. 3) The last one, hearer honorific, is assumed by a verbal inflectional element, -(p)ni-. 1. The transcription follows in this article Revised Romanization (2000) for Korean and Nippon-shiki for Japanese. The large semantic field of mal can be attested to in Japanese also. The Japanese equivalent of Korean mal, hana- delivers a rather extensive range of meanings such as hanasu (speak, say), hanasuni tariru hito (someone for one to discuss a matter with), hanasiga umai (story-telling, the art of conversation), hanasiokaeru (change the subject), totini tutawaru hanasi (folk tale, legend), hanasini yottewa (according to the situation), etc. For this similarity, we can assume that the universality of speech acts in human societies would permit this semantic enlargement. 2. For example, see Sook Lee (2004) and Sangsoo Park (1994). 241

255 ASIALEX 2005 These oversimplified rules seem to neglect some important consequences of Korean honorific frame. See the following examples. a) seonsaeng-i mal-eul hae-ss-ta. NP 1 -nom.mark. NP 2 -acc.mark. V-past-term. b) seonsaeng-nim-i mal-eul hae-ss-ta NP 1 -hon1.suff.-nom.mark. P 2 -acc.mark. V-past-term. c) seonsaeng-nim-kkeseo mal-eul hae-ss-ta NP 1 -hon.suff-hon1.nom.mark. NP 2 -acc.mark. V-past-term. d) seonsaeng-nim-kkeseo malsseum-eul hae-ss-ta NP 1 -hon.suff-hon1.nom.mark. NP 2 -acc.mark. V-past-term. e) seonsaeng-nim-kkeseo malsseum-eul ha-si-eoss-ta NP 1 -hon.suff-hon1.nom.mark. NP 2 -acc.mark. V-hon1.-past-term. f) seonsaeng-nim-kkeseo malsseum-eul ha-si-eoss-eupni-ta NP 1 -hon.suff-hon1.nom.mark. NP 2 -acc.mark. V-hon1.-past-hon3.-term. One consequence neglected in the honorific frame rules is the reinforcement effect in the degree of honorificity. Except for a) having no honorific marker, the others have at least one honorific marker: b) has just one honorific marker, -nimsuffixed to seonsaeng teacher, and the rest have more than two honorific markers: -si-, -(p)ni- attached, before ta terminal inflection, to hae- do. In the case of f), there is the superposition of five honorific markers, if we include semantic feature [+honorific] of malsseum in honorific markers. In the past, Korean children were rebuked for their mistakes in employing correct honorific combinations. When teacher takes NP 1 place, the sentence should have had a correct honorific noun as well as an honorific verbal form, since teacher has as part of its connotation, respectable person. In the case of a), children and even ordinary adults would have been treated as truly rude ones, but these days all of the examples seem to be easily said and considered as relatively correct forms. If they are, however, asked to choose the correct one, their answer must be still e) or f), according to utterance situations. Especially b) through e) have tendencies of being intuitively analyzed as reflecting the degree of honorificity. Otherwise, it should be noticed that the object referent honorific, reflected in the lexical choice of NP 2, NP 1 should be degraded in the degree of honorificity, accompanied with a corresponding verb as follows. g) nae-ga mal-eul hae-ss-ta. I-nom.mark. NP 2 -acc.mark. V-past-term. h) * nae-ga malsseum-eul hae-ss-ta. I-nom.mark. NP 2 -acc.mark. V-past-term. i) jae-ga malsseum-eul deuri-eoss-ta. I-nom.mark. NP 2 -acc.mark. V-past-term. 242

256 WORDS IN ASIAN CULTURAL CONTEXTS The example g) is neutral, so it gives no honorific effect. If only NP 2 is realized by malsseum and there is no other honorific index as in h), the utterance loses rapidly its acceptability and even grammaticality. The only way to preserve its correctness is to have an honorific verb and polite 1st person pronoun jae ( jeo). It should be noticed here that the honorificity delivered by NP 2 is oriented toward the honored person, who is in i) the addressee of malsseum. From this observation, a second consequence of Korean honorific frame should be induced: each part of the sentence is not totally destined to a specific category of the Korean honorific system. Instead all parts are correlated to have an utterance participant in the focus of honorificity 1. Modification of Korean honorific frame: Concordance of [±honorific] between sentence constituents The honorific superposition makes it necessary to modify Korean honorific frame presented above, producing non-negligible lexico-syntactico-pragmatic effets. In this section, we present our trial Korean honorific scheme as follows. [NP 1[+hon] -hon.suff.-hon1.nom.mark NP 2[+hon] -acc.mark. V [+hon] -si-(p)ni-term.] This is regulated by following rules: 1) NP 1 s honorific should satisfy semantic feature concordance constraint, otherwise the utterance would produce some humoristic or pejorative effects; 2) NP 2 should meet the same constraint, but as there is no possible honorific index in NP 2 headed phrase, the honorific concordance should be established with [+hon] in the other sentence constituents; 3) VP constituent should orient NP 2 s honorific toward a target speech participant. In lieu of a conclusive remark, we draw an anticipation that the modified Korean honorific frame could contribute to developing a Korean foreign language teaching method that would allow foreign Korean learners to acquire properly some Korean vocabulary, as well as an explanatory outline for one of the most significant linguistic phenomena in Far East languages. References Fillmore, Charles J. and B. T. S. Atkins (2000), Describing polysemy: the case of crawl, in Yael Ravin and Claudia Leacock (eds), Polysemy: theoretical and computational approaches (Oxford: Oxford University Press), pp (Ch. 5). (10 point, hanging indent by 10 mm) Comrie, Bernard (1976), Linguistic politeness axes: Speaker-addressee, speaker-referent, speaker-bystander, in Pragmatics Microfiche 1.7: A3-B1. (Oxford: Oxford Microform). Dixon, Robert M. W. (1972), The Dyirbal language of North Queensland. (Cambridge, England: Cambridge University). Geertz, Clifford (1960), The religion of Java. (Chicago: University of Chicago). Lee, Sook (2004), On nominative case marking in iss- exist constructions, in Korean Journal of Linguistics, 29-3 (Seoul: Linguistic Society of Korea), pp Lemaréchal, Alain (1989), Les parties du discours (Paris : Presses Universitaire de France). Lemarechal, Alain (1997), Zéro(s) (Paris : Presses Universitaire de France). Levinson, Stephen C. (1983), Pragmatics. (Cambridge, England: Cambridge University). 1. We define utterance participants as all entities, directly or indirectly, related to a speech act. Therefore someone who is simply mentioned in an utterance can play a role of a speech participant. 243

257 ASIALEX 2005 Pae, Kyung-Seon (2004), Requestive hints and politeness, in Studies in Modern Grammar (Seoul: The Society of Modern Grammar), pp Park, Sang-Jun (2004), Contribution à l étude de la transitivité en français, Thèse de doctorat (Paris : Université de Paris IV-Sorbonne). Park, Sangsoo (1994), Checking Korean honorific morphology in Studies in Modern Grammar (Seoul: The Society of Modern Grammar), pp The introduction of the lexicon of grammar into China: typologies of neologisms and historical background of their creation Tommaso Pellin Ca Foscari University of Venice, Italy tommaso.pellin@unive.it Abstract During the 19th century, the great flow of scientific texts coming from the Western countries into China made it necessary for the Chinese scholars to embark on the tremendous task of creating brand new sets of words in order to give a Chinese name to the new concepts introduced from Western learning and that were becoming part of their own learning. The studies carried out in the last fifteen years demonstrated that this process of creation of neologisms underwent an evolution during the period between the second half of the 19th century and the first two decades of 20th century. In this paper, I report the results of the lexical and socio-historical analysis of Chinese grammatical lexicon, as created in the first two Chinese texts dealing with grammar: Yingwen juyu 英文舉隅 (Examples of English language), by Wang Fengzao 汪鳳藻, published in 1879, maybe the first Chinese textbook of English language for Chinese students; Mashi wentong 馬氏文通 (General rules of language by Mr. Ma), by Ma Jianzhong 馬建忠, published in 1898, unanimously acknowledged as the first Chinese grammar of Chinese language. In this presentation I carry out a combined interpretation of lexicological data together with the relative historical background. The aim of my enquiry is to see whether and how the differences of historical context influenced the creation of Chinese grammatical lexicon. In general, the purpose of such an investigation is to formulate hypotheses on the communicative motivations that determined different lexical choices and different patterns of creation of lexical terms in a given period and regarding a given lexical domain. Introduction 244

258 WORDS IN ASIAN CULTURAL CONTEXTS During XX century, there have been many investigations on the lexicon of the Chinese language, which had the great merit of listing a number of words related to foreign cultures (one of the most relevant is Gao and Liu, 1958). In the last fifteen years, some researchers (for instance Masini, 1993; Liu, 1995; Lackner et al., 2000) even accomplished the task of drafting the historical trends of the coinage of Chinese foreign neologisms, i.e. the neologisms created after linguistic interference. The goal of this paper is to highlight the importance of deepening such researches through the interpretation of lexicological data and historical data together. The aim of such multidisciplinary enquiry is to formulate some hypotheses on the communicative motivations that determined the choice of some patterns of coinage of neologisms in a certain period and regarding a certain lexical domain. First, I will illustrate the outcome of the lexicological analysis carried out on the first two Chinese essays on grammar; then I will recall some historical and cultural details of the lives of the two authors. Finally, I will discuss the influence that historical contexts had on the process of lexical coinage for both authors. The texts examined were: 1. Yingwen juyu 英文擧隅 (Examples of English Language; hereinafter: Examples), published in 1879 and written by Wang Fengzao 汪鳳藻. According to Zhang and Ding, 2005, it is the first Chinese grammar of English language. 2. Mashi wentong 馬氏文通 (General Rules of Language by Mr. Ma; hereinafter: Rules), by Ma Jianzhong 馬建忠, published in It is unanimously considered the first Chinese grammar of Chinese language. Lexicological analysis At first, I selected thirty-four words and phrases belonging to the semantic field of grammar from the two texts; then, I subjected them to lexicological and lexicographical investigation to find out the different patterns of coinage according to which they were created. The lexical items were first subsumed under the most common categories of neologisms established by the scholars (namely, phonetic loans, graphic loans, semantic loans and loan-translation). After this subdivision, it turned out that just a small number of them fitted traditional lexical types; above all, a great number of compounds could not be satisfactorily grouped into them. Thus, two special typologies were identified: explicatory neologistic compound : these are compounds whose attributive part is rendered through the explication of the relationship between the general term (the head of the compound) and the specifying term (the compound itself) or through the explication of the function of the compound. They thus appear to be motivated compounds. For instance, in the compound jìjìngzì 繫靜字 adjective, jì 繫, the attributive part, means concerning and jìngzì 靜字, the head, is noun, so the term literally is concerning with the noun, what complements the noun ; likewise jìdòngzì 繫動字 adverb is what complements the verb (dòngzì 動字 ). denominative neologistic compound : these are compounds whose attributive part is quite an arbitrary denomination, so that the compounds turn out to be unmotivated. For example, in nèidòng(zì) 内動 ( 字 ) intransitive verb and 245

259 ASIALEX 2005 wàidòng(zì) 外動 ( 字 ) transitive verb, the attributive parts intransitive and transitive are rendered imbuing nèi 内 inwards and wài 外 outwards with the meanings of intransitiveness and transitiveness. Nominative case (zhŭcì 主次 ) and objective case (bīncì 賓次 ) are rendered through the compounding of cì 次 case and zhŭ 主 host and bīn 賓 guest respectively. As it can be noticed in the following table, these two particular forms of neologisms are the most frequently occurring in the corpus selected, while the others occupy only a little share. Moreover, it can be seen that the Examples has no denominative compounds, while in the Rules the explicatory compounds are very rarely used. Historical facts Typologies of neologisms Examples Rules Denominative neologistic compounds 0 11 Explicatory neologistic compounds 12 3 Graphic loans 1 0 Native phrases 1 0 Native words 3 6 Neologistic phrases 2 0 Phonetic loans 0 1 Semantic loans 5 6 Loan-translations 10 7 Table 1. Results of the lexicological inquiry In order to understand the reason why the two authors chose different words, it can be useful to investigate their historical and cultural backgrounds. During the second half of the 19th century, the bureaucrats, who constituted the leading class of the Chinese empire, promoted the introduction of a great number of Western works, mostly about scientific subjects, as a means to grasp a solution to resolve the crisis of the State, provoked by the military and economic pressures brought about by some Western countries. The goal of the Chinese leadership was to learn Western science, to modernize the army, to overcome the economic crisis and to dispel the invaders. But until the last decade of the 19th century, the leadership of the empire did not have a clear perception of the gravity of the institutional, social and economic downturn of China. They still trusted the traditional organization of the State and of the society. Therefore, Western texts were just meant to provide the basics of Western science, considered as a mere tool for strengthening traditional institutions. There was no intention at all of introducing the values and the institutions of the West all along with the science, let alone to adopt them. To provide the army, the public administration and the native entrepreneurs with Western texts, in 1862 Chinese government set up the Capital Institute for Foreign Languages (Jingshi Tongwenguan 京师同文馆 ). The Institute had two main tasks: the 246

260 WORDS IN ASIAN CULTURAL CONTEXTS first was to teach foreign languages and some elements of Western scientific disciplines to future statesmen, public managers, diplomats, translators and interpreters. The second was to translate Western texts on scientific matters, which were at first employed as schoolbooks for the lessons at the Institute (Wang, 1987: 39-44). Wang Fengzao ( ), the author of the Examples, was first a student and then one of the principal translators of the Institute. Born in a family of the provincial bureaucracy, he began his career as a translator at the Institute; then, after acquiring the status of bureaucrat by passing the imperial exams, he climbed up the bureaucratic ladder until the appointment, in 1892, as Ambassador in Japan (Su, 1978: ). Ma Jianzhong ( ), the author of the Rules, had a very uncommon cultural background. He was born in a Catholic family: thus, since a child, he was taught Western sciences, Western languages (among which English, French, ancient Greek and Latin) and was imbued with Western values (Liu and Xue, 2002: 4-8). After studying at the Shanghai school of French Jesuits, in 1870 he was enrolled into the entourage of Li Hongzhang, one of the main promoters of the modernization of China, and worked as a translator and interpreter. In 1877 he was sent to France with the first group of Chinese students who went overseas to study and graduated in international relations at the École Libre des Sciences Politiques; in addition, he practiced his foreign languages very much. After his return, he continued to work for Li Hongzhang until In that year, in spite of all the efforts to modernize the army and the country, the Chinese underwent a major defeat by Japan. Ma Jianzhong then was involved in the decline of Li Hongzhang and of that generation of leaders and reformers and retired. This allowed him to fulfil the long-dreamt project of writing a Western-type grammar of Chinese language (Liu and Xue, 2002: ). Influence of communicative goals on lexical choices After reporting the principal lexical differences about the patterns of coinage of neologisms of the authors and pinpointing some relevant elements of their lives, it is possible to discuss the influence of the historical context on the process of lexical coinage. Wang Fengzao s only purpose for writing the Examples was to transmit, as clearly and as simply as possible, the basic rules of English language, in order to let the Chinese students of the Institute master them, even if quite roughly. This hypothesis is confirmed by three facts. Wang Fengzao wrote a very simple grammar book, not a very structured essay on linguistics. Secondly, he himself states that his model was some Grammar by the American grammarian Simon Kerl. In fact, Kerl wrote only simple manuals for primary or middle schools; moreover, even if Wang Fengzao states that his work is a translation, he actually just made a summary of the key concepts of Kerl s grammar. Finally, Wang Fengzao entitled his work examples, in the sense that he did not have any intention of theoretically speculating on the English language. Considering the scope of Wang Fengzao, it can be maintained that, according to his training as a translator, in the process of coinage of words he felt only the necessity of rendering as easily as possible the terms of the text he summarized. It was for this reason that, save for some other kinds of neologisms, the type of neologism he most often employs is the explicatory compound, which is a motivated compound. Considering that Chinese linguistics had never had such concepts like parts of speech, cases, syntactic functions, etc., Wang Fengzao coined the words of the whole 247

261 ASIALEX 2005 system of English grammatical lexicon as specifying lexemes of zì 字, the one root concept of traditional Chinese linguistics, indicating at one time morpheme, syllable and grapheme. By doing that, the new concepts of Western grammar could be apprehended more easily because put in the framework of Chinese own linguistic concepts. Ma Jianzhong had not been trained as a translator, devoted to the comprehensibility of the original text. He was a learned scholar with a deep knowledge of many Western languages. Nevertheless, he had to serve all his life as a simple translator and interpreter. Therefore, when in 1895 he retired from political life, he had the time for accomplishing a most honourable academic task: writing the first Chinese grammar of Chinese language in accordance with Western linguistic categories. This consideration is supported by the fact that the models of Ma Jianzhong were not easy primers of grammar but quite surely Latin and Greek grammars that he studied first at the Jesuits school and then in Paris; according to Peyraube, 2000, he even took as model the Grammaire générale et raisonnée, which he certainly studied in France. In fact, the title of Ma Jianzhong s work is general rules of the language (Peyraube, 2000: 341). According Liu and Xue 2003, Ma Jianzhong had also the goal of providing Chinese language with a scientific grammatical description so to make clear to Western sinologists that also Chinese language had a grammar (Liu and Xue, 2002: 167). It is probably for the sake of scientific purpose that the comprehensibility of his work was not Ma Jianzhong s first concern, and in fact it has been acknowledged that the Rules was quite hard to understand at his time and it is still now. Ma Jianzhong did not care for creating immediately understandable lexemes but felt it necessary to coin scientifically correct terms. It is for this reason that he based the system of grammatical specifying compounds upon three concepts, zì 字 part of speech, cí 詞 syntactic function and cì 次 cases. Thus, he applied for the first time Western grammatical categories to Chinese language and demonstrated that Western grammatical categories could explain Chinese syntactic structures as well. But as for lexical coinage, he discarded motivated words. He coined the compounds simply by imbuing some concept with a completely new meaning, often not easily referable to the traditional one. In fact, it is possible to understand thoroughly the meaning of Ma Jianzhong s target lexemes only if already acquainted with the source lexemes (as Ma Jianzhong was). For instance, the meaning of zhuàngzì 狀字, literally zì of the mood (zhuàng 狀 ) is not so immediately clear if one does not already know that it is adverb and what adverb is, while jìdòngzì 繫動字, literally complement of the verb, is motivated. Conclusion In this presentation, I have reported the results of the analysis of the neologisms of two Chinese essays on grammar and some information about their authors. The goal was to provide all the necessary data to make out not only how these neologisms originated but also why, in which context, and for which purpose. The lexical data and the historical facts highlighted in this enquiry are valid for an analysis of the development of the neologisms of the lexicon of grammar in the last two decades of the 19th century. On the basis of the results, it can be maintained that the creation of the words in these two Chinese essays on grammar appears influenced by 248

262 WORDS IN ASIAN CULTURAL CONTEXTS historical factors, with the above exposed results. It is highly likely that the lexicons of other semantic domains had different patterns of generation and development. As I tried to show in this paper, a careful investigation on the composition of the lexicon, and in particular on its neologisms, and a historical research on the lives and the background of the authors, can help explaining the reasons for certain patterns of coinage of a lexicon in some period of time. References Gao Mingkai 高名凱 and Liu Zhengtan 劉正埮 (1958), Xiandai Hanyu wailaici yanjiu 現代漢語外來詞研究 (A study upon the loanwords in modern Chinese). (Beijing: Wenzi gaige chubanshe). Gusmani, Roberto (2003), Saggi sull interferenza linguistica (Essays on linguistics interference). (Firenze: Le Lettere). Liu, Lydia H. (1995), Translingual practice. Literature, national culture and translated modernity. China (Stanford: Stanford University Press). Liu Zhengwei 劉正偉 and Xue Yuqin 薛玉琴 (2002), Ma Xiangbo Ma Jianzhong Ma Yuzhang 馬相伯馬建忠馬玉章. (Shijiazhuang: Hebei jiaoyu chubanshe). Masini, Federico (1993), The formation of modern Chinese lexicon and its evolution toward a national language: the period from 1840 to (Berkeley: Journal of Chinese Linguistics - Monograph Series number 6). Peyraube, Alain (2000), Some reflections on the sources of the Mashi wentong, Lackner, Michael et al. (eds) (2000), New terms for new ideas. Western knowledge and lexical change in late imperial China. (Leiden: Koninklijke Brill), pp Su Jing 蘇精 (1978), Qingji Tongwenguan ji qi shisheng 清季同文官及其師生. (The Foreign Languages Institution and its teachers during the Qing dynasty). (Taipei). Wang Daming, 王大明 (1987), Jingshi Tongwenguan ji qi lishi diwei 京師同文館及其歷史地位 (The Foreign Languages Institution in Peking and its historical role), Zhongguo keji shiliao 中國科技史料 (Documents on the history of science and technology in China), 8(4): Zhang Tongbin 張同冰 and Ding Junhua 丁俊華, Zhongguo waiyu jiaoyu fazhan shi huigu 中國外語教育發展史回顧 (Historical review of the development of the teaching of foreign languages in China), available online, (accessed 31 March 2005) Abstract Corpus planning and language shift in Malaysian law Richard Powell Nihon University ricpowell@yahoo.com Lexical creation and standardisation have been at the centre of policies to shift the medium of law in Malaysia from English to Malay, but are they sufficient to effect this shift? While some proponents of language shift believe that Malay corpus planning is rendering the need for courtroom English obsolete, one argument commonly made by those opposed to abolishing the central role of English is that Malay terminology is not 249

263 ASIALEX 2005 up to the complexities of administering a system of justice heavily derivative of English common law and continually influenced by Commonwealth precedents. Using the example of Istilah Undang-undang, a growing corpus of Malay equivalents for English legal terms sponsored by Malaysia s Institute of Language and Literature, this paper will highlight key constraints upon the creation and dissemination of a corpus of legal Malay, outlining microlinguistic problems involving orthography and semantics before moving on to macrolinguistic issues such as the difficulty of establishing a body of legal discourse in Malay to imbue new terms with cultural acceptability and professional authority. In conclusion it will be argued that both the advocates of language shift and its sceptics have tended to focus too narrowly on terminology to the detriment of wider linguistic and non-linguistic restraints on language shift. The microstructure for a dialect dictionary of agricultural vocabulary in Tamil S. Raja Annamalai University Annamalainagar, Tamil Nadu, India rajaling@rediffmail.com Abstract Due to the influence of science and technology many indigenous occupational vocabularies are slowly getting replaced by English. In this context, much culturally related vocabulary like occupational items are slowly replaced and much textural hybridity is taking place. At this juncture, this paper will mainly focus on different issues related to textural hybridity in Tamil agricultural vocabulary. Apart from this, this paper shows the significance of dialect surveys and the compilation of a dialect dictionary. The significance of dialect surveys is to bring out the variant forms of different regions. Recent research in the field of metalexicography has focused on the structure of the dictionaries identified by Hausmann and Wiegand (1989) and Weigand (1989) are made use of. The proposed dictionary will examine the different structures of a dictionary the macrostructure and the microstructure. Here the macrostructure is the selection of lemma sign is the spoken variety only. The microstructure of the proposed dictionary has two important structures; the information structure is not a lexicographic definition but the lexical description given by the farmers in Tamil. Another important structure in the microstructure is providing the variation found in the different dialect regions. This is the main aim of the proposed dictionary. For this, the microstructure may be different when compared with the microstructure of a general monolingual dictionary. Here all the variant forms are clearly marked by different labels. These variants are treated in the microstructure as a dialect variant form (DVF). 250

264 WORDS IN ASIAN CULTURAL CONTEXTS Lexicography is an applied field, the theoretical background for which is provided by linguistics. Almost all the branches of linguistics provide information for the making of a dictionary. Phonetics and phonology are useful for providing information about the pronunciation, standardization of the script and alterations in the form of headwords. Grammar (morphology and syntax) is essential for providing grammatical information of different types, like parts of speech of words, gender distinction and causal relation of nouns, conjugation of verbs etc., Dialectology, Sociolinguistics, Psycholinguistics etc., are useful in giving usage and other types of labels to the headwords. But the branch of linguistics which is most important for the compilation of dictionaries is semantics, especially lexical semantics, as meaning is the central and vital component in the structure of a dictionary entry. This paper mainly focuses on the method of preparation of an article or dictionary entry for agricultural vocabulary in Tamil. Dictionary Definition According to the Oxford English Dictionary (OED), the word dictionary is derived from the Latin form dictionarius, used in 1225 by English poet and grammarian Joannes de Garlandia (John of Garland) as the title of a collection of Latin vocables and the word dictionarium, used about a century later by Peter Bechorius (or Becharius). The first recorded appearance of the word dictionary as such is dated to 1526 by the OED; and later the word was used by Thomas Elyot in The French word dictionnaire seems to have been used for the first time by Robert Estienne in 1539 (Bejoint, 2004). In Webster s third new international dictionary (W 3 ), the definition of dictionary is slightly more detailed, but very similar: A reference book containing words usually alphabetically arranged along with information about their forms, pronunciation, functions, etymologies, meanings and syntactic and idiomatic use. Matore (1968) finds all definitions of dictionary in dictionaries less than satisfactory. The only one that satisfies him is the one that is proposed by the OED. It runs: A book dealing with the individual words of a language (or certain specified classes of them), so as to set forth their orthography, pronunciation, signification and use, their synonyms, derivation and history, or at least some of these facts: for convenience of reference, the words are arranged in some stated order, now, in most languages, alphabetical; and in larger dictionaries the information given is illustrated by questions from literature. Zgusta (1971) uses a slightly more precise formulation: A dictionary is a systematically arranged list of socialized linguistic forms compiled from the speech-habits of a given speech-community and commented on by the author in such a way that a qualified reader understands the meaning of each separate form, and is informed of the relevant facts concerning the function of that form in its community. Dictionary Typology Zgusta (1971) opined that when a lexicographer sets out to compile a dictionary, he has to take two basic decisions (1) what part of the total vocabulary of a language the proposed dictionary will cover and (2) of what type the proposed dictionary will belong. Both the aspects of typological classification of dictionaries are quite useful in understanding the classification and locating the proposed dictionary. Dictionaries can 251

265 ASIALEX 2005 be classified into various types on the basis of different criteria. Zgusta, classifies dictionaries into two major divisions namely, linguistic and non-linguistic dictionaries respectively. The linguistic dictionaries are concerned with the words or lexical units of languages, called word books. The non-linguistic dictionaries are not concerned with the words, but with realia or denotata (thing) and they are called encyclopedias, or thing books. Classification of linguistic dictionaries has been attempted by a number of scholars such as Shcherba (1940), Sebeok (1962), Malkiel (1967), Cornyn (1967), Gelb (1968), Zgusta (1971), Al Kasimi (1977), Singh (1982) and Svensen (1993). Contrasting features of dictionary typology are divided into two basic classes viz., internal features and external features. The internal features are concerned with the aspects of the nature of language such as paradigmatic vs. syntagmatic contrast, aspects of form/ meaning, time and area of vocabulary covered. External features are concerned with the target group or user of the dictionary for whom the dictionary is compiled, size, purpose, arrangement of entries or words, number of languages etc. Internal features are theoretical in nature, whereas the external features are applicational or are concerned with the use or practical utility of the dictionary. Methodology Agricultural vocabulary have been collected from fifty points comprising of the five major dialect regions of Tamil Nadu namely (1) Eastern dialect, (2) Western dialect, (3) Northern dialect, (4) Southern dialect and (5) Central dialect. The distance between each point was twenty five to thirty kilometre distance. Informants were selected from densely populated areas. The informants are of different age groups and caste groups, both male and female. Old persons in the age group of fifty to seventy, who have no educational background, were also selected. Lexical items relating to various agricultural areas like soil, irrigation, water lifting instruments, weather and season, ploughing, leveling the field, seeds, crops, harvesting, manure, cattle, etc., were included in the detailed questionnaire which was prepared for this purpose. The questionnaire contained 527 direct questions. In addition to this, a number of additional questions were also asked to collect sufficient data. The data were recorded in a tape-recorder. Data were analysed and presented in the form of a dictionary format. Description of Lexical Meaning Zgusta (1971) says that there are four basic instruments which are used for the description of lexical meanings. They are: (1) The lexicographic definition, (2) synonyms, (3) exemplification and (4) glosses and labels. A number of other theoreticians include synonyms as a kind of definition (Svensen, 1993; Kipfer, 1984; Landau, 1989; Hartmann and James, 1998). What Zgusta calls exemplication, others such as Landau (1989) refer to as illustrative quotations, Ilson (1986) as illustrative phrase, Svensen (1993) as examples of usage and Kipfer (1984) calls illustration or exemplification used as theorganic part of the definition within parenthesis. Label is another instrument used to describe the lexical meaning especially in the compilation of a dialect dictionary (proposed dictionary), and in connotation. Zgusta, regards labels as a species of glosses. But, others such as Svensen (1993) discuss it separately, while Landau (1989) and Kipfer (1984) describe it as a part of usage. Lexicographic Definition 252

266 WORDS IN ASIAN CULTURAL CONTEXTS There are different types of lexicographic definitions (1) Paraphrase, (2) Referential/ analytical/ intentional definition, (3) extensional definition, (4) formulaic definition and (5) definition by synonyms. The second type, what Zgusta (1971) calls lexicographic definition, Ilson (1986) calls as referential or analytical definition, while Svensen (1993) calls intentional definition, which are modified versions of lexical definitions genus proximum and differentia specifica. The concept represented by the headword is called the definiendum and the definition (verbal description) is called the definiens. Intention denotes the content of the concept which is defined as the combination of distinctive features or what Zgusta calls criterial features which the concept comprises. It expresses a generic conceptual relationship whereby concepts are arranged in classes according to similarities and differences noted between them. This has resulted in a hierarchical system with super-ordinate (hyperonym), sub-ordinate (hyponym) and co-ordinate (co-hyponym) concepts. The process of definition involves stating the superordinate concept next to definiendum, i.e., (genus proximum) together with at least one distinctive feature typical of the definiendum (differentia specificum) (Svensen, 1993). This type of definition is useful to explain a class of related words constituting a set of hyponyms with or without superordinate, meronyms and also for preparing a dialect dictionary. E.g., kilai (n) marattin oru paakam/ pakuti Ital (n) puuvin oru pakuti/ paakam Component parts and structure of a dictionary The present article focuses on the dictionary type in general and very particular to the structure of an entry. The word dictionary has two meanings in a textural manner i.e. (1) the whole book and (2) the word list, which constitutes the main part of the book. Both of them have two structures, the textural book structure and the textural word list structure respectively. The word has several important units and structures. The basic unit of the dictionary is the treatment unit. The treatment units have a form and information relating to that form are brought together. The relation of form and information is that of topic and comment (Hausmann and Wiegand, 1989). A form and information relating to a form are brought together under the addressing procedure. Each information item is addressed to a form called address. In any dictionary the most important item is the definition, but there may be hundreds of other information types, i.e., items. The most important address is lemma (headword or entry word), because the lemma belongs to the alphabetical access structure of the dictionary. Normally all the ordered set of lemmata of the dictionary forms the macrostructure. The lemma and the whole set of information items, which are addressed to the lemma, form the dictionary article. Normally speaking, the structure of information within the article is called the microstructure. The Structure of an Entry The proposed dictionary entry is given in the following format. The headword or article is given in Tamil script with bold letters, followed by the phonemic script (transliteration). The grammatical indication is also given and in addition to this, the English equivalent is also provided. After that, the lexicographic description is given in Tamil, which is added for better understanding of the target language group (Tamil language group). Many types of dictionaries will have the same structure but the 253

267 ASIALEX 2005 proposed dialect has a different structure. That is, in addition to the lexicographic description, the variations found in the different regions are also presented in the same entry. Those variations are labelled. If there is no variation in any district to the headword, then it is believed that the speaker of particular district use the same form (i.e., the central region form). In such cases, no form is given in that entry. Since the central region (Thanjavur) has the standard spoken variety, the headword is given in that variety only. All the regions are labelled according to the nesting method. Conclusion Due to the influence of science and technology, many new technical terms in English come into existence in almost all fields of Tamil. Because of this reason, much native vocabulary of indigenous occupations are slowly getting replaced by English language. This may lead to permanent loss of the rich occupational vocabulary in Tamil. This study will be quite helpful in preserving the occupational vocabulary in Tamil and recording it in a dictionary format for future purpose. This study may provide a model for the future survey and to record the various occupational vocabularies in Tamil. References Bejoint, Henri (2000), Modern Lexicography: An Introduction (Oxford: Oxford University Press), pp.6-32 (Ch.1). Cruse Allan, D (1986), Lexical Semantics (Oxford: Oxford University Press). Hartmann, Reinhard R.K. (1983), Lexicography: Principles and Practice (London: Academic Press). Hartmann, Reinhard R.K. and G. James (1998), Dictionary of lexicography (London: Routledge). Hausmann, Franz J. and H.E. Wiegand (1989), Component parts and structures of General Monolingual Dictionaries: A Survey, in Franz Josef Hausmann and Oskar Reichmann (eds), An International Encyclopedia of Lexicography (Berlin: Walter de Gruyter), pp Ilson, R.F. (1996), Lexicography, in Asher (ed), The Encyclopedia of Linguistics (London: Routledge), pp James, Gregory (2000), Colporul: History of Tamil dictionaries (Madras: Cre-A Publications). Kipfer, B.A. (1984), Work book on lexicography, in R.R.K. Hartmann (ed), Exeter Linguistic Studies (Exeter: University of Exeter), Vol.8. Landau, Sidney I. (1989), Dictionaries: The art and craft of lexicography (New York: Scribner). Malkiel, Yakov (1967), A typological classification of dictionaries on the basis of the distinctive features in Householder and Saporta (eds), Problems in Lexicography, pp Matore, G. (1968), Histoire DES dictionaries francais (Paris: Larousse), pp (Ch.1). New Well, L.E. (1995), Hand book on lexicography (Manila: Linguistic Society of the Philippines). Sebeok, T.A. (1962), Materials for a typology of dictionaries Lingua, 1: Singh, Ram A. (1982), An introduction to lexicography (Mysore: CIIL). Svensen, Bo. (1993), Practical lexicography (Oxford: Oxford University Press). Zgusta, Ladislav (1971), Manual of lexicography (Mouton: The Hague). Roman letter words in Mandarin Chinese 254

268 WORDS IN ASIAN CULTURAL CONTEXTS Helena Riha Ohio State University Abstract China s efforts since the 1980s to become fully integrated into the world community have led to changes in Mandarin Chinese. Mandarin is showing the effects of China s opening and internationalization through language contact with English. Although lexical borrowing from English through phonetic loans and calques is an established phenomenon, increasing use of lettered words and phrases that combine a roman letter component with native morphemes (e.g., X guang X ray, VIP piaojia VIP ticket price ) is a new and growing trend in PRC Standard Mandarin. These lettered words constitute an unprecedented interpretation of conventions in lexical borrowing and word formation in Chinese which traditionally required adaptation of foreign elements through nativization via Chinese morphemes. Some lettered words and phrases in Mandarin are native creations, while others constitute some of the first types of pure loanwords in Mandarin. Both provide evidence of an increasingly international Chinese worldview and challenge established notions of what is native in the national language. While transliteration, phonetic loans, and loan translations have been studied extensively in Chinese, lettered words receive little attention in research on Mandarin morphology and have yet to be described systematically. This study identifies basic types of lettered words and phrases currently in use and describes their morphological structure. The emergence of lettered words is described as an extension of existing morphological and lexical borrowing processes. The study concludes with thoughts on the significance of the entry of lettered words into the Mandarin lexicon on changing notions of Chineseness. A comparison of paper and electronic monolingual dictionaries: location, comprehension and retention of secondary senses James Ronald Hiroshima Shudo University ronald@shudo-u.ac.jp Akira Tajino Kyoto University akira@tajino.mbox.media.kyoto-u.ac.jp Abstract This paper reports a study comparing location, comprehension and retention of secondary word senses by advanced Japanese learners of English using the same monolingual learner dictionary as either a paper or electronic dictionary (ED). Research into ED use has focused on primary senses of words, often resulting in higher location rates for EDs than for paper dictionaries. While ED accessing mechanisms may favour the rapid location of these senses, the same may not be true for secondary senses of L2 words. Findings from the study reported here suggest that advantages 255

269 ASIALEX 2005 attributed to EDs in some studies may not apply for users of monolingual learner dictionaries in their task of finding secondary senses of words. Introduction Full-content handheld electronic dictionaries (henceforth EDs) are increasingly widely used by foreign language learners. Lighter and smaller than paper dictionaries, they are widely seen as being faster to use and there is some research to support this (Koyama and Takeuchi 2004). However, although research using electronic dictionaries has a history as long as these dictionaries themselves (e.g., Krantz 1991; Knight 1994), there is as yet little or no evidence for any advantage of using EDs from the perspective of vocabulary acquisition. One important study that does address this issue is that reported by Osaki, Ochiai, Iso and Aizawa (2003). We will begin this paper with a brief review of their study. This will clarify the issues involved in such research, and lead us to identify questions that remain unresolved and require further attention. We will then report a study that sets out to address these issues. Osaki et al s (2003) study investigated the value of using bilingual paper and electronic dictionaries for finding 15 unknown words encountered in a reading passage, in assisting comprehension of the passage, and in immediate and delayed retention of the target words. The study was conducted with groups of Japanese university students with higher and lower English proficiency levels. The dictionary used was the same edition of a widely available English-Japanese dictionary in both paper and ED formats. ED-using participants at both higher and lower proficiency levels were able to locate more correct target word meanings than paper dictionary users at their relevant levels. The same was true for comprehension questions, with the ED users performing better at both levels than their paper dictionary counterparts. As for the immediate retention test and that conducted one week later, there was no significant difference between electronic and paper dictionary users. Regarding these findings, we need to consider two important issues: that the study is concerned with bilingual dictionary use, and that almost all the target items were for the first or only sense of the words. With bilingual dictionaries we may generally assume that location of the correct entry will guarantee its comprehension. With a monolingual learner dictionary (MLD), we cannot take the comprehension of L2 entries for granted, especially in the case of less advanced level learners. The issue of the focus on the first or only sense of target words is of particular importance in studies of ED use. While all the senses of a word will usually be visible to a paper dictionary user who has found the right entry, only the first senses will appear on the screen of most EDs, who will either have to scroll down the entry to view subsequent senses or to select a particular sense or part of speech from a menu screen in order to view it. Both of these factors may offset much of the advantage found for EDs with regard to the successful location of relevant senses and, consequently, comprehension and retention of these senses. The Study This brings us to the focus of the study of this paper: comparing the location, comprehension, and retention of secondary senses of L2 words in paper and electronic formats of a monolingual learner dictionary. There were three main research questions: 1. What difference, if any, is there between electronic and paper dictionary users in the successful location of specific secondary senses of targeted 256

270 WORDS IN ASIAN CULTURAL CONTEXTS words? 2. What difference, if any, is there between electronic and paper dictionary users in the comprehension of definitions for the specific senses of targeted words? 3. What difference, if any, is there between electronic and paper dictionary users in retention of sense meanings as measured in a delayed retention test? The dictionary used for this study was the popular Longman Advanced American Dictionary (Summers 2000), with the paper version kindly supplied for this study by Pearson Education Limited, and the electronic version contained in the handheld electronic dictionary Casio EX-word AZ-H9100, kindly lent by Casio Limited. The text of the two versions of the dictionary was identical, although there were some differences in the fonts and styles, especially for signpost words, used in the entries. A total of 94 participants with upper-intermediate or advanced level English proficiency took part in the first part of the study, with 86 also taking part in the retention test one week later. They were from two classes of first year students, one majoring in arts and science subjects and one majoring in science. Their average estimated vocabulary size was about 5,800 English words (see Okamoto 2005), sufficient to cope with the comprehension of definitions in the MLD used in this study. Both classes were divided into two, with half the participants from each class given paper dictionaries to use (the PD group) and half given electronic dictionaries (the ED group). Administration of Nation s Vocabulary Levels Test (Nation 1990: ) confirmed that there was no statistically significant difference between the groups in this respect. In order to focus on secondary senses of words, the materials chosen for this study were newspaper headlines with polysemous words or idioms whose senses rendered the headlines ambiguous. Here are two examples, with ambiguous items underlined: Farmers Wives Win Prizes for Fat Calves. Lung Cancer in Women Mushrooms. All but one of the targeted items were secondary senses of words, or were idioms that appeared towards the end of their respective entries. The procedure of the experiment is summarized in Table 1. Time ED group PD group Week 1 10 mins MLD use guidance and task explanation 10 mins ED guidance, familiarization Dictionary familiarization 40 mins Dictionary task: sense location and comprehension Week 2 20 mins Retention test Table 1. Experiment Procedure Before the dictionary task, the participants received 20 minutes guidance and practice in using monolingual learner dictionaries, both paper and electronic, and were given instruction sets for this and the task in Japanese to keep and refer to as needed during the task. For each task item, the participants read the headline, together with the translation of the wrong, or comic, meaning of the headline, and looked up the underlined 257

271 ASIALEX 2005 ambiguous word to find the intended sense of this word. They then wrote down two or three keywords from the definition of the sense they had identified as the correct one. Following this, they would provide a Japanese translation equivalent or explanation of the sense. Here is one example: New Vaccine May Contain Rabies ( 新しいワクチンは狂犬病を含んでいるかもしれない ) [ = There may be rabies included in new vaccine] [key words] 単語の意味 [meaning] The full list of target items is as follows: mushrooms, cut, give the slip, fine, contain, try, case, left, waffles, board, appeal, drop off, march, by, calves, held, battery, charge In the delayed retention test held one week later, the participants were presented with the same headlines with translations as for the task and asked to recall the other, real, sense of the underlined word. This method was chosen to produce a sensitive test in which the original lexical environment would assist the participants in recalling the targeted sense. For the location of senses, a correct answer was one where the appropriate sense for the word could be identified from the keywords supplied by the participant. For the tests of sense comprehension and retention of senses, a list of correct and partially correct answers was drawn up for each item. The participants answers were then rated according to this list, with answers rated as correct, partially correct, wrong, or no answer. For the purposes of statistical analysis, correct and partially correct scores were collapsed into one score. A summary of results from the experiment is presented in Table 2. As the data shows, scores for location of the relevant sense and comprehension of the sense are very similar for the two groups, and t-tests confirmed that there was no significant difference between the two groups in these respects. For location of the relevant senses, participants of both groups were typically successful for just over two thirds of the items. Comprehension rates for the definitions of these senses were just a little lower, with an average of just under two thirds of the items rated as correct or partially correct. As for the retention of senses, a t-test confirmed that there is a significant difference between the two groups (t = , d.f.=84, p< 0.05), with the ED group scoring higher than the PD group. SD Sense Location Sense Comprehension Sense Retention Mean SD Mean SD Mean PD (N = 47) (N = 44) 9.20 ED (N = 47) (N = 42) Table 2. Results from Experiment A breakdown of how comprehension of senses was rated for the two groups is shown in Figure 3. We can see that the ED group gave fewer wrong answers than the PD group but on average gave no answer for a larger number of items. Proportions of 258

272 WORDS IN ASIAN CULTURAL CONTEXTS correct and partially correct answers were very similar for the two groups, with the ED group participants giving, on average, a slightly higher proportion of correct answers. 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% PD ED No answer Wrong Partially Correct Discussion Figure 3. Rating of Sense Comprehension Before proceeding to address the questions about differences attributable to the medium of the dictionary used, we should ask whether the participants understood what the task required of them, and whether they were allotted sufficient time to complete the task. The participants in both groups were given 40 minutes to complete 18 items: for each item, their task was to locate and comprehend the correct secondary sense of a specific word, and to record that they had located and comprehended the sense. The average success rate for both groups of close to 70% in locating the correct sense suggests that for most participants the requirements of the task were clear, and that sufficient time was allocated for the completion of the task. This is confirmed by a response rate (including incorrect responses) approaching 80% for comprehension of the relevant senses. We will now return to the three questions raised earlier regarding the comparative efficacy of electronic and paper dictionaries in the location, comprehension, and retention of secondary senses. As we consider the results of our study that address these questions, we will also refer to those obtained in previous studies. As regards location of the relevant sense for each item, average results for the ED and PD groups were almost identical, with both identifying the relevant sense for over 12 of the 18 items. This differs markedly from previous research in which ED group participants either located (and comprehended) a significantly higher number of entries (Osaki et al. 2003) or located entries significantly faster (Koyama and Takeuchi 2004) than PD group participants. The lack of difference in our study may in part be a consequence of the time allocated for the task, but two factors that may be more 259

273 ASIALEX 2005 influential are the use of monolingual dictionaries rather than bilingual dictionaries, and the focus on secondary senses rather than single or primary senses. In our study, although ED users may have located the entry for the targeted item faster than paper dictionary users, this advantage appears to have been offset by the relative ease of scanning entries for the correct sense afforded by the paper dictionaries. Comprehension rates for the relevant senses were also very similar for the two groups, with participants in both groups giving meanings for the senses that were rated as either partially correct or correct for just over 11 of the 18 items, or around 62%. The closeness of these figures to the location rates shows how much the tasks are interdependent; identifying the correct sense depends on comprehension of the sense from the definition or example sentences. As for the similarity between groups scores for comprehension, another reason for this is the lack of difference between the task for the two dictionary types: reading the entry for the sense and understanding the definition and example sentence. One difference is that the ED users may need to scroll down to see the whole entry for the sense, while another is that the ED entry is less cluttered and easier to read. One further factor may be the style in which signpost words are presented in the two dictionary types. In the ED, they are presented like this for try: 4. food/drink. In the paper dictionary, the reader sees this: 4. food/drink. This does, however, appear to have had little effect on participant comprehension rates, perhaps because the highlighting effects of the two styles are similar. Alternatively, data suggests that this may be because signposts assist sense location in some cases where the signposts are effectively abbreviated definitions, and to work as distracters in others such as that shown above where the signpost indicates in which field the sense is used. Finally, we come to the retention rates of the 18 targeted senses. The mean retention rate for PD group participants was just over 50% at 9.20, with the ED group participants scoring a significantly higher Two questions need to be addressed: why, for both groups, retention rates are so high, and how we can account for the ED group s significantly higher scores. The retention rates are higher than those in comparable studies (e.g., around 30% for delayed retention test scores in Osaki et al. s study) for two main reasons. One is that there would have been prior knowledge of some of the items which would be reinforced through the dictionary task. The other is that the test was designed to be a sensitive measure of word knowledge retention. The dictionary task was a kind of puzzle, with the correct sense being the answer to the puzzle. In the retention test, the participants were presented with the same puzzle and asked to recall the answer; this is undoubtedly easier to do than recall a particular required sense of a word presented in isolation. Although it could be said that the ambiguous newspaper headlines are not typical lexical environments for the targeted items, they are arguably more so than words presented in isolation. Two main factors may help us understand the significant, and unexpected, higher retention rates for ED participants: the different actions involved in the use of the two types of dictionaries, and the focus of attention during dictionary use for the two groups. For the ED users, the item search started with the typing of the word, while paper dictionary users started by turning pages. ED users have to make a series of decisions, such as identifying part of speech then consciously rejecting or selecting the senses that appear on the screen. Paper dictionary users may go through the same motions or they may just scan the entry until they pick out sense that seems to match the context. If the users are experienced at this, scanning time might be focused, efficient, and effective for L2 vocabulary learning. 260

274 WORDS IN ASIAN CULTURAL CONTEXTS If not, their time may be better employed in typing the word on an ED keyboard and following the various menus and signposts to the sense. In conclusion, the motivation for this study was to investigate areas of paper and electronic dictionary use that appear to have been neglected: monolingual learner dictionaries and secondary senses of targeted words. It is also hoped that the study may be useful in offering a counterweight to research that made use of materials and instruments which may favour EDs over paper dictionaries. There were limitations to this study, such as the fact that some participants did not take the retention test, and these may have had some effect on the results. Clearly, although further research is needed, it is too early to say that either the electronic or the paper dictionary is the better learning dictionary. These factors may not affect many learners decisions about which dictionary to buy and use, since portability and attractiveness seem to be primary considerations, but it is hoped that this and similar research will continue to merit the attention of both language teachers and dictionary makers. Finally, our profound thanks to Della Summers of Longman Dictionaries, Mr Hirayama of Casio Computer Co. Ltd, Masamichi Mochizuki, Kazumi Aizawa, M. Okamoto, Naoko Koizumi, and Midori Kanmei for their generous help in making this study possible. References Knight, Susan (1994), Dictionary use while reading: the effects on comprehension and vocabulary acquisition for students of different verbal abilities, The Modern Language Journal 78 (3): Koyama, Toshiko and Takeuchi, Osamu (2004), Comparing electronic and printed dictionaries: how the difference affected EFL learning, JACET Bulletin 38: Krantz, Gösta (1991), Learning vocabulary in a foreign language: a study of reading strategies (Gothenburg: Acta Universitatis Gothoburgensis). Nation, I.S.P. (1990), Teaching and learning vocabulary (Boston MA: Heinle and Heinle). Okamoto, M. (2005), University students lexical acquisition and attrition in English as a foreign language, unpublished MA thesis (Kyoto: Kyoto University). Osaki, Satsuki, N. Ochiai, T. Iso and K. Aizawa (2003), Electronic dictionary vs. paper dictionary: accessing the appropriate meaning, reading comprehension and retention, in Minoru Murata, Shigeru Yamada, and Yukio Tono (eds), Dictionaries and language learning: how can dictionaries help human and machine learning? Papers submitted to the third ASIALEX biennial international conference. (Urayasu: Asian Association for Lexicography), pp Summers, Della (ed) (2000), Longman advanced American dictionary (Harlow: Pearson Education). A lexical semantics approach to designing a more user-friendly EFL dictionary Kensei Sugayama Kobe City University of Foreign Studies ken@inst.kobe-cufs.ac.jp Abstract In this paper I will discuss meanings in EFL dictionaries. From a Lexical-Semantic point of view, I shall try to throw some light on the description of meanings in EFL 261

275 ASIALEX 2005 dictionaries in order to make it more user-friendly. More specifically, I will show that a Lexical-Semantic analysis of the English verb sweep gives a new taxonomy of senses of the verb and accordingly it improves classification of senses used in the current EFL dictionaries. Introduction In this paper I will discuss meanings in EFL dictionaries. From a Lexical-Semantic point of view, I shall try to throw some light on the description of meaning in EFL dictionaries in order to make it more user-friendly. I will be concerned only with the English activity verb sweep. Reinterpreting the data collected concerning the verb sweep, I will outline arguments which support a different, and in my view, more adequate and elegant analysis of the semantic structure of sweep in Lexical-Semantic terms and will suggest ways of integrating the lexical-semantic account with the traditional accounts of meaning description in EFL dictionaries. Dictionary Design Let me start off by thinking about the dictionary design. Let us take the learners dictionaries that are easily accessible and consider the kind of information that they provide about words, and see whether that information constitutes a revealing and comprehensive lexical description. We examine first of all what might constitute such a description. Hudson (1988, 1995a) gives a checklist of types of lexical facts under eight headings. The list in (1) is an abridged version of Hudson s more comprehensive list, but it will serve our purpose, which is to show what kind of lexical information the ideal dictionaries should contain: (1) a. Phonology: the pronunciation of a word and variants b. Morphology: a word s morpheme composition, and any irregular inflections c. Syntax: the word class and particular structures a word may enter d. Semantics: the meanings of a word and the semantic structures it may enter e. Context: restrictions on social context and style in which a word may be used f. Spelling the normal and any variant spellings of a word g. Etymology: the origin and history of a word h. Usage: frequency of use, when acquired, any taboos In view of the list, it is obvious that meaning is among the most important information that constitutes lexical description in the dictionary entries. Accordingly we can even say we use a dictionary just to look for the meaning of a word we just do not know as far as the beginners are concerned. The purpose of this paper is to explore the way in which we will best incorporate this kind of semantic information into dictionary description. In what follows I shall use the verb sweep to illustrate my points. Senses of sweep As is well known, the main thing about the EFL dictionaries, the learners dictionaries in particular, is to show the learners of English how words are used or should be used in contemporary English. Therefore, it is quite reasonable to observe at 262

276 WORDS IN ASIAN CULTURAL CONTEXTS the start how the word sweep is used in contemporary English, i.e. the real facts about the word itself. Before taking a closer look at the facts, let us first consider what the reference grammar says about senses of the verb sweep. According to WordNet, the verb sweep has eight senses. WordNet The verb sweep has 8 senses (first 7 from tagged texts) 1. brush, sweep (sweep across or over; Her long skirt brushed the floor ; A gasp swept cross the audience ) 2. sweep, sail (move with sweeping, effortless, gliding motions; The diva swept into the room Shreds of paper sailed through the air ; The searchlights swept across the sky ) 3. sweep, broom (sweep with a broom or as if with a broom; Sweep the crumbs off the table ; Sweep under the bed ) 4. embroil, tangle, sweep, sweep up, drag, drag in (force into some kind of situation, condition, or course of action; They were swept up by the events ; don t drag me into this business ) 5. cross, traverse, span, sweep (to cover a wide area; Rivers traverse the valley floor, The parking lot spans 3 acres ) 6. sweep (clean by sweeping; Please sweep the floor ) 7. sweep (win an overwhelming victory in or on: Her new show dog swept all championships ) 8. swing, sweep, swing out (make a big sweeping gesture or movement) Here I shall limit myself to discuss only two senses of the verb sweep, enclosed by a square above, i.e., the sense of sweep with a broom or as if with a broom as in 3 and that of clean by sweeping as in 6. It seems to me that these two senses are core meanings of the verb. The metaphorical or extensional uses of these senses, being their extensions, are excluded from the discussion here. The data This section gives the real facts about two senses of sweep mentioned earlier. Some of the examples are taken from various corpora, and others are tested against my informants. Based on a detailed observation of the data, one can make the following statements: The verb can be used intransitively given an appropriate context. When it is used as the transitive, its object is construed as a surface of some kind. Compare these with less acceptable or ungrammatical ones. The object when it appears can occur with an additional predicate representing a change of state of what the object refers to. When it is used in the sense of move something by sweeping, the goal phrase as well as the object is required to appear. The crucial point here is that missing out the goal phrase will lead to making the resulting sentence ungrammatical. 263

277 ASIALEX 2005 What lexical semantics tells us about sweep Following Pustejovsky (1995), Rappaport Hovav and Levin (1998), Levin and Rappaport Hovav (1991, 1995, 1999), Pinker (1989), Jackendoff (1990), among others, verb meanings may be encoded using event templates. An event template consists of a constant (< >), the verb-specific core meaning that supplies spell-out, and an event structure, common to all verbs of a class, that specifies the relation between arguments of the verb. For example, sweep of activity sense is a manner verb which takes its name from a manner constant <SWEEP> enclosed by angle brackets. Because of the nature of sweeping, this constant is associated with two arguments or participants: a sweeper (x) and a surface (y) (or the place swept). Notice here that the stuff that might be on the surface is not among the minimum set of participants, because one can sweep a surface without anything being on it. Therefore it is an optional participant, which is represented by underlining the element. (2) An activity sense (a) [x ACT <SWEEP> y] (b) Terry swept [the floor]. Sweep of activity sense has an event structure template consisting of a single event. Both participants (x, y) are associated with the same subevent. Only one of the two (i.e. x) needs to be expressed and the other (y) is a semantically invoked argument, rather than a structural participant, therefore, not necessarily expressed. A change of state sense of sweep, on the other hand, has the following semantic template as in (3). (3) A change of state sense (a) [[x ACT <SWEEP> y] CAUSE [BECOME [y/z <STATE>]]] (b) Terry swept the floor clean. (y:y) (c) Terry swept the leaves into a pile. (y:z) The causing subevent the result subevent [change of state] Each subevent is identified by a predicate (sweep, <STATE>) and associated with an argument NP. Omitting the object would leave the subevent without an associated NP, leading to ungrammaticality: *Terry swept clean. From an event-structural point of view, this sense can be analysed in such a way that the activity (i.e. the first event) caused the second event (i.e. a change of state in y or z). This sense is actually grouped further into two sub-types: one case where an activity caused the change of the state of the place (or location) as in (4) and the other where the activity caused the change of the stuff on the surface. (4) A removal sense (a) [[x ACT <SWEEP> y] CAUSE [BECOME [z <PLACE>]]] (b) Terry swept the leaves off the pavement. Semantic structure of this sense is the same as that of change of state sense except that semantic participant (y) is not the same as the structural participant (z). 264

278 WORDS IN ASIAN CULTURAL CONTEXTS In passing, structural participants, however, cannot be omitted, because there is a general rule saying that each predicate has to be associated with at least one NP, which is a structural participant (or argument). Each structural participant (or argument) has to be associated with a subevent. If PP is omitted, there would be no subevent associated. Participants y and z cannot be omitted because either of them is a structural participant and being a structural participant, it has to be expressed at the surface structure. To sum up, a lexical semantic analysis of sweep can show that there are at least three different types of semantic structures for the verb sweep: a single event and complex events of different two types. What dictionaries tell us about sweep Taking into account the lexical-semantic information identified above, let us then have a quick look at how this kind of information is reflected in EFL dictionaries currently available. What follows are a couple of reproductions of what current EFL dictionaries describe about the senses of sweep we are concerned with. LDOCE 4 1 CLEAN SOMETHING [transitive] to clean the dust, dirt etc. from the floor or ground, using a brush with a long handle ᅳ synonym brush Bert swept the path in front of the house. sweep something off/out/up etc Will you sweep the leaves off the patio? 2 PUSH SOMETHING SOMEWHERE [transitive always + adverb/preposition] to move things from a surface with a brushing movement I swept the papers quickly into the drawer. LDOCE 3 Sweep 1 1 CLEAN STH [transitive] to clean the dust, dirt etc from the floor or ground using a special brush: Bert swept the path in front of the house. Sweep the floor clean for me please. 2 PUSH STH SOMEWHERE [transitive always + adv/prep] a) to clean a surface by pushing something to a particular place or in a particular direction with a special brush: Could you sweep the snow off the patio for me? b) to move something to a particular place or in a particular direction with a brushing or swinging movement: The wind swept the dead leaves away. I swept the papers quickly into the drawer. OALD 6 265

279 ASIALEX 2005 Sweep verb with brush 1 to clean a room, surface, etc. using a BROOM (= a type of brush on a long handle): [VN] to sweep the floor / street / stairs Chimneys should be swept regularly. [VN-ADJ] The showroom had been emptied and swept clean. [also V] 2 [VN+adv./prep.] to remove sth from a surface using a brush, your hand, etc: She swept the crumbs into the wastebasket. He swept the leaves up into a pile. She swept the clothes onto the floor and invited him to sit down. move quickly / with force 3 [VN+adv./prep.] to move or push sb/sth suddenly and with a lot of force: The little boat was swept out to sea. Their tent was swept away in the storm. She let herself be swept along by the crowd. 4 (of weather, fire, etc.) to move suddenly and/or with force over an area or in a particular direction: [V+adv./prep.] Rain swept in through the broken windows. A fire swept through the store on Tuesday night. [VN] Strong winds regularly sweep the islands. Concluding remarks Having perused the entries of the EFL dictionaries above, let me point out some problems of these descriptions and the accompanying examples. First of all, there is a mixture of examples of different senses under the same entry. For instance, OALD 6 gives the following sample sentences under sense 1 as in (5). (5) to clean a room, surface, etc. using a broom (= a type of brush on a long handle): [VN] to sweep the floor / street / stairs Chimneys should be swept regularly. [VN-ADJ] The showroom had been emptied and swept clean. [also V] Obviously the former example underlined is a sense of activity, while the latter one shaded is that of change of state. There is another mixture of the same kind with OALD 6 under sense 2 as in (6). (6) [VN+adv./prep.] to remove sth from a surface using a brush, your hand, etc: She swept the crumbs into the wastebasket. He swept the leaves up into a pile. She swept the clothes onto the floor and invited him to sit down. Another point to make is that there is no indication of the intransitive use of sweep in the sense of activity in LDOCE 3, OALD 6 and CIDE although CIDE has examples of the intransitive use given in the entry. 1 The only exception on this point is COBUILD 3, which quite rightly and in a user-friendly way, gives a verb pattern of intransitive use. 2 Lastly, the third point is the occurrence of the non-subcategorisable object with this verb as instantiated in (7). (7) *Phil swept the crumbs. In fact the EFL dictionaries currently available imply sentences like (7) are not possible but they do not explain why it is not possible. It seems to me that what they really have to do is to explain why sentences like (7) are not possible in English, which should be made in terms of lexical semantics. 266

280 WORDS IN ASIAN CULTURAL CONTEXTS At this point I cannot solve the problems raised above. Much depends on the overall approach to event-structural lexical semantics and the relation of semantic representation and its use in the dictionaries. Notes 1, 2 The relevant entries in CIDE and COBUILD3 are not given here for lack of space. References Cambridge International Dictionary of English (1995), (Cambridge: Cambridge University Press). [CIDE] Collins Cobuild English Language Dictionary (2001, 2004), (Glasgow: Harper Collins). [COBUILD3, 4] Collins Cobuild grammar patterns 1: Verbs (1996), (London: Harper Collins). Hudson, Richard A. (1988), The linguistic foundations for lexical research and dictionary design, International Journal of Lexicography 1(4): Hudson, Richard A. (1995a), Identifying the linguistic foundations for lexical research and dictionary design, In D.E. Walker et al. (eds), Automating the lexicon (Oxford: Oxford University Press), pp (Ch. 2). Hudson, Richard A. (1995b), Word meaning (London: Routledge). Jakcendoff, Ray (1990), Semantic structures (Cambridge, Mass.: MIT Press). Jackson, H. and E. Z. Amvela (2000), Words, meaning, and vocabulary (London: Cassell). Levin, B. (1993), English verb classes and alternations (Chicago: University of Chicago Press). Levin, B. and M. Rappaport Hovav (1991), Wiping the slate clean, Cognition 41: Levin, B. and M. Rappaport Hovav (1995), Unaccusativity (Cambridge, Mass.: MIT Press). Levin, B. and M. Rappaport Hovav (1999), Two structures for compositionally derived events, SALT 9: Longman Dictionary of Contemporary English (1995, 2004), (Harlow: Longman). [LDOCE3, 4] McCawley, James D. (1986), What linguists might contribute to dictionary making if they could get their act together, In P.C. Bjarkman and V. Raskin (eds), The real-world linguist (Norwood: Ablex Publishing), pp (Ch. 1). Oxford Advanced Learner s Dictionary (2000), (Oxford: Oxford University Press). [OALD6] Pinker, Steven (1989), Learnability and cognition (Cambridge, Mass.: MIT Press). Pustejovsky, J. (1995), The generative lexicon (Cambridge, Mass.: MIT Press). Rappaport Hovav, M. and B. Levin (1998), Building verb meanings, In M. Butt and W. Geuder (eds), The projection of arguments (Chicago: CSLI), pp (Ch. 4). WordNet 1.6 (1997), (Cambridge, Mass.: MIT Press). Intermediary languages in bilingual lexicography for lesser-used languages: the pros and cons Mats-Peter Sundström European Parliament, Swedish Translation Unit msundstrom@europarl.eu.int Abstract With a total of twenty official languages as of May the European Union translation services face a situation possibly of relevance in an Asian lexicographic 267

281 ASIALEX 2005 context also. In this situation, a number of questions come to the fore. Are there enough dictionaries available to aid translations from one lesser-used language into another (e.g. from Latvian into Greek) or do such translations have to be re-routed through what is called an intermediary language (e.g. translating first from Latvian into English and then from English into Greek)? Does recourse to an intermediary language invariably increase the risk of subject matter and shades of meaning being lost in translation? Or are intermediary languages in fact an aid to translation, since bilingual dictionaries involving at least one major language as source or target language are generally more comprehensive both in lemmata selection and article size than bilingual dictionaries centered exclusively on lesser-used languages? Finally: is the effort involved in compiling bilingual dictionaries between lesser-used languages offset by the advantages they offer, viz. increased precision in equivalents and lemmata descriptions? The present paper attempts to elucidate this issue, on the basis of research conducted in seven lesser-used language translation units of the European Parliament (the Estonian, Finnish, Lithuanian, Maltese, Polish, Slovene and Swedish units). Attention is focused on whether in those units it is found difficult to use intermediary language dictionaries and whether available bilingual dictionaries between lesser-used languages are considered satisfactory, hoping the results, mutatis mutandis, will lend themselves to a modicum of application to words in Asian cultural contexts also. A casual observer of the lexicographic scene may well be pardoned for concluding that lexicography, either bilingual or monolingual, is somehow bound to involve the English language. Or in the words of Bejoint: The field of international metalexicography is clearly more interested in English-language dictionaries than in dictionaries of any other languages [Bejoint 2003:4]. Again, if we take the expression English-language dictionaries to include also bilingual dictionaries where English is either source or target language, we arrive at a surprisingly accurate picture of today s lexicographic world. Most notably, bilingual dictionaries where both the source and target language are lesser-used languages are the odd man out among bilingual dictionaries. The underlying reasons are easily understood. Bilingual dictionaries are mostly commercial products and thus subject to constraints of feasibility and marketability. Even so, bilingual dictionaries between lesser-used languages do exist. As an example may be mentioned the three-volume Swedish Finnish General Dictionary [some lemmata] and the two-volume Finnish Swedish General Dictionary [close to lemmata] both published in Finland in the 1980 s and the 1990 s respectively and directed at language communities of nine million speakers [Swedish] and five million speakers [Finnish]. Also, the largest dictionary publishing house in Finland, the WSOY, has a standard policy of publishing dictionaries involving Finnish and above all other European languages. In this context, culture-related considerations also play a role in company policy, apart from purely commercial ones [Hosia 2005]. The plethora of bilingual dictionaries in Finland mainly stems from a desire to break the linguistic isolation from mainstream Europe, due to Finland s having a non-indo-european language, i. e. Finno-Ugric Finnish, for its major national language. This goes a long way towards proving that it generally takes some extralinguistic sociocultural factor to promote the development of what we may call non-conventional 268

282 WORDS IN ASIAN CULTURAL CONTEXTS bilingual lexicography. An example of this tendency is the Finnish Ndonga, Ndonga Finnish dictionary. This dictionary, involving an African Bantu language, appeared in 1980, as the fruit of then more than one hundred years of activities by the Finnish Missionary Society in what later became the Republic of Namibia. Generally, though, bilingual dictionaries between lesser-used languages are uncommon. Here, often an intermediary language bilingual dictionary (ILBD for short) enters the picture. If a Maltese text has to be rendered into Estonian a situation perfectly conceivable in European Union language practice, the lack of direct bilingual dictionaries, or DBDs for short (as in this case Maltese Estonian), calls for the use of an ILBD. The intermediary language, or IL for short, will most probably be English. Thus, Maltese lexical items will be looked up in a Maltese English dictionary and then, possibly, the resulting English equivalents will be checked with the help of an English Estonian dictionary. Though far from ideal, ILBDs enjoy a kind of tacit approval as an aid to translations between lesser-used languages if not for any other reason, then out of pure necessity. Bearing in mind the role ILBDs play, it may come as somewhat surprising that ILBDs have not to any greater extent attracted lexicographers curiosity. A survey of the International Journal of Lexicography, shows the above-mentioned subject was not treated in any volume between 1988 and Neither was it treated in Szende s major work on contrastive lexicography. In view of this apparent scarcity of studies, the present writer will attempt to shed light on some of the problems posed by ILBDs, hoping to provide impetus to further research. The framing of the issue will follow a dual approach. First the advantages and disadvantages of ILBDs will be brought up for discussion. Then, on the basis of interviews with European Parliament language professionals, further light will be shed on the use of ILBDs. Initially, ILBDs seem fraught with unavoidable disadvantages. They may be summed up as follows: 1. duplication of language competence demands, 2. semantic peripherilization of equivalents, 3. IL induced red herring effects, 4. obfuscating of hierarchies and 5. interfering with search strategies. Duplication of language competence demands is obvious: when resorting to an ILBD the dictionary user will be required to master yet another language in addition to the source and target languages. The resulting additional burden may of course vary considerably. Since the most frequently used IL worldwide is surely English, in geographic regions where knowledge of English is widespread and solid the need to verify the meaning of English-language equivalents in the target language will be less urgent than elsewhere. Even so, the presence of yet another language as a tertius interveniens between source and target language is bound to make itself felt. Semantic peripherilization of equivalents is probably the most serious problem related to ILBDs. As Taczalska pointed out, often there is no one-to-one equivalence in lexical items between two languages, they are only an approximation and do not correspond in meaning [Taczalska 1997:163]. The result, in the words of Svensén, is that the semantic fields between source language entry and target language equivalent 269

283 ASIALEX 2005 do not overlap with ensuing disturbances of communication, in a situation where an important task of a bilingual dictionary consists precisely of eliminating such disturbances [Svensén 1987:135]. Interpreting source language lexeme meanings through an IL involves the risk of veering away at a tangent from the source language semantic field, no matter what quality of lexicographic work has gone into compiling the ILBD. A result of such peripherilization is an occasional loss of idiomacy, unless the equivalents provided by an ILBD are checked out carefully in bilingual dictionaries involving the IL and the target language. One of the results may be redundancies in the equivalents proposed owing to needless disambiguating, i.e., disambiguating that serves no purpose in the target language although perfectly justified in the IL. Consider the expression cabinet minister. In English, the qualifying noun cabinet is relevant, as a minister would otherwise primarily be associated with a clergyman. However, a word-for-word translation into, for instance, Swedish as *kabinettsminister carries a redundancy, as in present-day Swedish, a minister is always a member of a government. Inserting such a qualifier here would constitute a breach of what Berkov called interlingual idiomacy [Berkov 1999:12]. The red herring effect comes into play when a word is polysemous in the IL, though neither in the source nor the target language. Cf. the Finnish word palomies which could translate into English as fireman. Such a translation, however, leaves the question open as to what precisely is meant: a firefighter or a stoker. In several other languages, this ambiguity does not arise. Red herring effects may also affect collocations. A graphic illustration is provided by Finnish and Japanese, where, by pure coincidence, the verb that is used when the object is a dream, corresponds in meaning to the English verb see. Thus, in Finnish nähdä unta and in Japanese yume o miru (*see a dream), whereas the idiomatic English verb here is have. In cases like these, ILBDs certainly risk misleading the user. Ambiguating of hierarchies relates to Koller s statement about translation as a matter of recapitulating language hierarchies into other languages [Koller 1987:191]. Basically, this involves conveying not only the denotational core of an utterance but also the connotational, stylistic, register-related and such aspects attaching to it. Here, an ILBD may act as an unwelcome intruder, notably in cases where the source and target languages have an idiomatic expression lacking in the IL. A striking example is offered by these two idioms in Finnish and Japanese respectively, which both express a feeling of utter shame: in Finnish haluaisin vajota maan rakoon ( I would like to sink into the ground ) and in Japanese ana ga attara, hairitai ( If there were a hole in the ground, I would like to slip into it ). The closest English expressions, like I could ve died for shame carry none of the metaphorical elements Finnish and Japanese quite fortuitously have in common. Interfering with search strategies, finally, refers to Bogaards notion of dictionary users search strategies being dependent on the users native languages [Bogaards 1990:98]. Conceivably, if both source and target language but not the IL belong to the same language family or group, it is quite possible that an ILBD may disturb dictionary searches. After this survey of ILBD drawbacks the examination will proceed to consider their advantages. They may be summed up as follows: to the publishers: feasibility 270

284 WORDS IN ASIAN CULTURAL CONTEXTS to the editors: monodirectionality to the dictionary users: availability also in larger formats than DBD:s. Feasibility is due to the fact ILBDs cater to a larger audience than DBDs involving lesser-used languages. Monodirectionality refers to the fact that ILBDs, by their very definition, are used for purposes of text reception, never text production, something that facilitates the editors task in compiling them. Also, monodirectionality evidently helps the users avoid what Leung called the temptation of immediate insertability of translation equivalent[s] [Leung 2003:153]. From the user s point of view, availability in larger formats results from feasibility. As pointed out by V. Hosia from Finland s largest dictionary publishing company, despite the extensive company policy of putting as many bilingual dictionaries as possible on the market, involving Finnish both as source and target language, the lemmata count in DBDs between Finnish and other lesser-used languages falls in a modest range of between and , and very probably the situation is similar in other lesser-used language areas as well [Hosia:2005]. For a view of the situation from the angle of day-to-day translation work with its accompanying demands for dictionary use, the present author turned to professional translators (and in one case an interpreter) working for the European Parliament. At present the Parliament has twenty working languages. They are, in English alphabetical order, Czech, Danish, Dutch, English, Estonian, Finnish, French, German, Greek, Hungarian, Italian, Latvian, Lithuanian, Maltese, Polish, Portuguese, Slovakian, Slovene, Spanish and Swedish. In a European Union context, the description of major language is reserved for what G. Erasmus called Europe s three federating languages, i.e. French, German and English [Erasmus 2004:46] and thus the remaining seventeen including Spanish! by consequence fall under the heading of lesser-used languages. On a sampling basis, the following questions were put to translators in the Estonian, Finnish, Lithuanian, Maltese, Polish, Slovene and Swedish translation units of the European Parliament: 1. If you have capacity for handling translations between lesser-used languages, do you find DBD coverage satisfactory regarding the number of source languages covered, lemmata selection, dictionary format and accuracy? 2. Do you find ILBDs an adequate tool? What possible disadvantages have you encountered in using them? 3. Would you consider an expansion of DBD coverage into your own language desirable, necessary and urgent? Initially, it must be underlined that translations between lesser-used languages played a markedly limited role, most notably so in the units translating into languages from member states having acceded on May Even so, satisfaction with DBD coverage was fairly low. In this context, however, it should be pointed out that the availability of DBDs, at least in smaller formats, was surprisingly high. Here, the Finnish language ranked very high, for cultural reasons mentioned above (actually there are DBDs between Finnish and all EU language, excepting Maltese and Slovene). The Slovene language also distinguishes itself by broad DBD coverage, a fact which in the opinion of Slovene translation unit head, Valter Mavric, is probably attributable to Slovenia s historical experiences as part of the multilingual Austro-Hungarian empire [Mavric 2005]. Not surprisingly, for the tiny Maltese language there were DBDs from 271

285 ASIALEX 2005 only one lesser-used language, namely Italian. Complaints concerning ILBDs generally focused on the need for additional language (IL) knowledge and notably, the difficulty of using such dictionaries when searching for equivalents relating to technical, legal and other special lannguages. Sometimes, ILBDs were seen as retarding translation work, owing to the need for double equivalent searches: in the source language and the IL. Some of the most perspicacious comments were given by a Lithuanian translator, Giedrė Andreikėnaitė, translating i.a. from Finnish into Lithuanian, and a Finnish translator-interpreter, Christina Skogster, working i.a. between Dutch and Swedish. The former pointed out that the whole mentality in the dictionary, as manifested for instance in the choice of entryword illustrative phrases, would be different and more appropriate in a DBD than in an ILBD, as it would then not be dictated by the social and cultural system underlying the IL. The latter maintained that the use of an ILBD, apart from being time consuming, may lead the user on dead-end searches, mainly owing to IL polysemy not being correctly interpreted, and generally gives the translator a vague feeling of unease for not working in a wholly professional manner [Skogster 2005]. Finally, an extension of DBD coverage was perceived as highly desirable across the board. A graphic illustration was given by Slovene translator Darja Pečovnik, who unbeknownst made herself the spokeswoman of many a colleague in emphasizing she would definitely and for sure consult dictionaries more frequently if a Dutch Slovene DBD were available [Pecovnik 2005]. Also, as pointed out by Swedish unit translator Jonas Winnerlöv, DBDs are not only an aid to translation but also provide a direct link between nations. Here, in sum is perhaps the most important contribution an investment in DBDs can give, also in view of a possible future ever closer union between Asian nations. References Andreikėnaitė, Giedrė (2005), oral communication during interview with author, Mar Bejoint, Henri (2003), Modern Lexicography: Past, Present and Future, plenary lecture presented at the Third Asialex Conference, Tokyo, Aug Berkov, Vladimir (1999), De tospråklige ordbøkers begrensninger [ The defects of bilingual dictionaries ], in Nordiska studier i lexikografi 4, ed. by Peter Slotte, Pia Westerberg, Eva Orava. Bogaards, Paul (1990), Où cherche-t-on dans le dictionnaire?, International Journal of Lexicography 3(2): Erasmus, George (2004), L Europe des langues, Le Jeudi Dec , p. 46. Hosia, Vuokko (2005), oral communication during interview with author, Feb Koller, Werner (1987), Einführung in die Übersetzungswissenschaft (Heidelberg-Wiesbaden: Quelle & Meyer). Leung, Matthew (2003), A bilingual dictionary for translation students, paper presented at the Third Asialex Conference, Tokyo, Aug Mavric, Valter (2005), oral communication during interview with author, Mar Pečovnik, Darja (2005, oral communication during interview with author, Mar Skogster, Christina (2005), oral communication during interview with author, Mar Svensen, Bo (1987), Handbok i lexikografi. Principer och metoder för ordboksarbetet (Stockholm: Tekniska nomenklaturcentralen). Szende, Tamás (1999), Approches contrastives en lexicographie bilingue, in Journées d etude sur la lexicographie bilingue (Paris: Institut national des langues et civilisations orientales). Taczalska, Anna (1997), Professional problems encountered by translators of medical texts, in Zofia Rybińska and Danuta Kierzkowska (eds), On the practise of legal and specialised translation (Warsaw: The Polish society of economic, legal and court translators, TEPIS), pp Winnerlöv, Jonas (2005), oral communication during interview with author, Feb

286 WORDS IN ASIAN CULTURAL CONTEXTS The benefits of CD-ROM dictionaries in teaching Monika Szirmai Hiroshima International University, Japan Abstract This paper is based on experience and information gathered in a reading class for first year non-english major university students at a Japanese university over the last three years. After a short educational background, some of the differences between the different types of dictionaries are discussed. Although the paper version and the handheld one may be more useful in many cases, the CD-ROM version proved to have more advantages in general. In addition to speed of access and ease of cross-reference to related entries and pictures, most dictionaries also offer opportunities for listening and pronunciation through their sound files and recording functions. Many also provide additional example sentences to illustrate typical uses. CD-ROM dictionaries help not only decoding of the foreign language but also encoding. Introduction Dictionaries have always been considered as essential tools for helping communication in a foreign language. In spite of their widely acknowledged importance, according to student questionnaires at the Japanese university I work at, only a negligible few of the students have ever actually been taught how to use a dictionary, which most often meant using the hard copy of bilingual dictionaries. This leaves them ill-equipped to cope with their current learning situation and unprepared for future use of English in academic and specialist work situations. As schools should provide students with skills that they can use in the future, it is essential not only to teach and practice dictionary skills but also to do this using the format that students will be most likely to need and benefit the most from. Considering that modern companies are highly computerized, and most of their everyday communication is done through electronic messages, CD-ROM dictionaries seem to fulfil both expectations. After a brief overview of the educational background and that of the different types of dictionaries, this paper will discuss the main benefits of using CD-ROM dictionaries in regular language classes, such as integration of skills, visual aids, cross-referencing, ease and speed of searching, integration into learning activities, portability, and extra exercises. Educational background At university level in Japan, English is usually taught skills-based, with little or no coordination between the different courses: reading, writing, and oral communication. This latter may also be divided into listening and speaking. Sometimes reading and writing classes are run in a teacher-centered way because these classes often accommodate more than 50 students. If students work individually, they have to rely on their dictionaries and work silently until they finish the task and check it. As Japanese 273

287 ASIALEX 2005 universities are highly computerised, reading and especially writing classes often take place in computer labs. In fact, this paper is also based on information gathered in a reading class for first year non-english major students. There were more than 60 students in the class and the Cambridge Learner s Dictionary with CD-ROM (CLD) was used as a course textbook. Dictionaries are generally considered as reference books that should be consulted occasionally rather than books that should be used for everyday studies. Nowadays, students do not even have to look up words in dictionaries as most textbooks have a word list either right at the end of each unit or at the back of the textbook. However, being a non-native speaker of English, my personal experience as a language learner using the Oxford Advanced Learner s Dictionary of Current English (OALD), popularly called as The Hornby at the time, encouraged me to set a dictionary as a course textbook. As mentioned above, very few students had ever been taught how to use the dictionary. During the first semester, students learnt what information is available in the dictionary and how they could find it. At this stage, students are usually not so familiar with computers, so using the paper dictionary takes some unnecessary burden off the students. However, the worksheets are in an electronic form, so they have to type their answers, and this helps with learning spelling. Once the students are familiar with the hard copy, it takes only one or two classes in the second semester to change to using the CD-ROM version of the same dictionary. In the second semester, texts are chosen both for intensive and extensive reading, so that the students can learn different ways of reading and of using the dictionary. Differences between types of dictionaries Basically, dictionaries today are available in four media: 1. as a traditional paper dictionary; 2. as a handheld electronic dictionary; 3. as a CD-ROM dictionary; and 4. as a web-based dictionary. The information contained in each type does not vary so much, as the information is usually just transferred from one platform to another. Printed versions of dictionaries are just as well based on corpus research, the results are organised on the computer, and the actual hard copy is created using computers before it is printed. However, some changes in the content are necessary for each type of publication. Just to give an example, handheld dictionaries may not have any illustrations or just very few, whereas both hard copies and CD-ROMs do. Dictionaries may have more black and white pictures but CD-ROMs are usually very colourful. Before we go on, it must be clarified that as our main concern is the use of monolingual learner dictionaries, this is what is meant by the term dictionary in most cases. If not, it will be specified. All dictionaries contain or come with some kind of a user s guide that explains how information is displayed and encoded in the given dictionary. Entries provide information about pronunciation, meaning, grammar, collocations, etc., some even about frequency (e.g. COBUILD, Longman Dictionary of Contemporary English, (LDOCE)). Most publishers use a defining vocabulary of two or three thousand words for the explanations in the dictionary. Probably it is not necessary to go through the whole list to prove that it is not so much what is different but how the content is organized and how the information is presented and can be accessed. It must be mentioned though right now that CD-ROMs can and do hold much more information than printed books, which also includes sounds, pictures and moving pictures. In this 274

288 WORDS IN ASIAN CULTURAL CONTEXTS respect, the CD-ROM offers much more to the user because it contains additional learning material in addition to the dictionary itself. Benefits of CD-ROM dictionaries in teaching The strengths of using CD-ROM dictionaries come from the nature of the medium itself. As mentioned above, CD-ROMs can hold text, pictures, moving pictures, and sound, all of which can be displayed and used simultaneously. Very often control over functions is given to the user. For example in the case of pictures, the user can decide whether to see labels or use a quasi test mode when pictures must be matched with the labels by the user. With paper dictionaries, there is no option how much information is shown at once because all must be printed and is revealed by opening the page. There also seems to be a tendency to combine dictionaries in the CD-ROM versions. For example the Longman Dictionary of Contemporary English and the Longman Language Activator are two different publications in their paper version but the CD-ROM of the LDOCE contains both of them. Let us see different aspects of how CD-ROMs enhance teaching. Integration of skills As CD-ROMs combine text with sound, oral skills can be integrated to some extent to the silent reading and writing classes. Students can listen to and practice the pronunciation of words or phrases they have just encountered the first time. This is very important because many Japanese students have problems with matching spelling and sound. It is common knowledge that the Japanese language does not make a distinction between the sounds r and l. As a result, if a Japanese speaker uses r -like sounds, the word lamb will sound like ram. On the other hand, if a speaker uses l like sounds, we might hear a confusing sentence like this: I eat lice every day. This problem is aggravated by the frequent use of English words in everyday Japanese: English words are written down using the Japanese katakana syllabary and pronounced accordingly. That is why this kind of pronunciation is called katakana pronunciation. Another obstacle for Japanese students in acquiring the right pronunciation without an aural model is that phonetic transcriptions, such as IPA, are not taught at school. So consulting a paper dictionary may not result in being able to say or understand a word they have looked up if it is only seen but not heard. CD-ROM dictionaries contain sound files as models, usually spoken by both an American and a British native speaker. These dictionaries often have a recording function as well where students can listen to the model and record their own speech for comparison. Visual aids Visual aids fall into two categories: one helps navigation and cross referencing in the dictionary, the other explains the meaning of words by showing the objects, animals, actions, etc. A clear layout can also increase the speed with which one can find the right information. Figure 1 shows the same entry in the paper version (left) and the CD-ROM version of CLD. The paper version uses light blue for the headword and no other colours, while the CD-ROM version uses dark blue for the headword and red for the guidewords. Even a quick glance at the two versions can convince us that the one on the left looks crammed and information can be found much easier in the one on the right. 275

289 ASIALEX 2005 Figure 1. A comparison of visual clarity in a hard copy (left) and a CD-ROM (right) One might think these seemingly slight differences do not make any difference but in fact they do, especially for students who are usually slower at finding information in an English text. In the paper version, the symbol informs the user that related information is available at a different place. In the CD-ROM, according to the conventions of links, blue underlined words serve the same purpose. In this case, at the end of the first sense of egg, there are two such links: one for Related words and one for the illustration of food. Cross-referencing If we take another look at Figure 1, it becomes clear that in total there are only two signals for cross referencing in the paper version but there are five in the CD-ROM, which means that each sense has at least one link to other related words. As a matter of fact, everything is related to something, even the letters of the alphabet. This is another strength of the CD-ROM version. If students have to find the related information by turning the pages, that may take quite some time. So students tend to look up pictures but they rarely look up other entries. With the CD-ROM, because this can be done simply by clicking on the link, students are curious enough to carry out such an effortless action. Even if they do not read or understand the whole list of words that comes up, they might either just skim though it or choose one or two that they look up. Most students like this kind of browsing in the dictionary, which means that studying English can become more enjoyable, and may improve the students motivation or at least increase their level of tolerance with English-only texts. Other colour-coded links available are: Word Building, Collocation, Learner Error, and Verb Ending. If students regularly look up for example Word Building, it will be easier for them to discover known elements in unfamiliar words as well, and by doing that guessing from context becomes easier. The word work, the following list came up: Nouns: work, workaholic, worker, workings; Adjectives: workable, unworkable, overwork, working; Verbs: work and rework. The Learner Error section 276

290 WORDS IN ASIAN CULTURAL CONTEXTS pointed out the difference between job and work, which are often used incorrectly by students. In addition to that, career, profession and occupation were also mentioned. Searching for words Using a paper dictionary can help users memorize the alphabet, so students who are used to using the dictionary can find information relatively fast. However, students who would need the help of the dictionary the most often get lost in the dictionary and give up their word hunt. With the CD-ROM, these students can search for words just as fast as the other students. By saving time on using the dictionary, students can accomplish more work in one class. When students try to produce a text in a foreign language, they usually use only a bilingual dictionary to look up the word in the mother tongue and find the translation equivalents in the target language. After that, very few would go as far as checking the equivalents they have found in a monolingual dictionary, mostly because of the time and effort involved. However, students tend to use the thematic picture pages because they can quickly identify what they are looking for. With the help of the CD-ROM dictionary, students can also find synonyms for words, related word forms, and many example sentences to help production. Japanese students have difficulties in spelling words they have only heard but not seen. Some CD-ROM dictionaries, for example the LDOCE, make it possible to search for words based on their pronunciation. Of course, this can only be done if students are familiar with the phonetic symbols. Integrating the dictionary into the reading activity When reading text on the computer screen, it is even easier to integrate CD-ROM dictionaries into the task, as some have a function which makes it possible to look up words by moving the pointer over the word (Figure 2). This is extremely useful because students can look up all the words they want to without spending a lot of time typing the words into the dictionary program. Students tend to look up more words than they would without this function. The small window can be moved anywhere in the screen. Here as well, students have instant access to the sound files by clicking on the speaker marks, one of which is for the British pronunciation and the other for the American one. Figure 2. Quick look up window in CALD 277

291 ASIALEX 2005 Portability If some one ever had to take two or more big dictionaries to school from home, that person appreciates the portability of CD-ROM dictionaries. Even if students are required to bring their own dictionaries to classes, very often they turn up without them. A CD-ROM hardly weighs anything and takes up very little space. Many schools have dictionaries already installed in their computer labs, which also solves the problem of students possible forgetfulness. Extra activities CD-ROM dictionaries contain exercises related to different areas. Some are designed to help learning grammar, others for vocabulary. Some also contain corpus examples and corpus related activities. These activities can be done individually as the student can also see the right answers after finishing the task. Conclusion Based on my experience as a teacher and a language learner, it does not seem to be an exaggeration to say that a good dictionary and dictionary skills are the pillars of language learning. Although dictionaries are available in many different formats, CD-ROM dictionaries are superior to other formats because of the nature of the medium. Their relational database structure, which allows linking words in many different ways, the ease and speed with which information can be accessed, and the capability of integrating sound cannot be matched by any other media. In addition, if they are used in regular language classes, they can smoothly be integrated into learning activities regardless of whether students are decoding the foreign language or encoding into it. Preparing students for autonomous learning in the future is one of the key tasks of schools. In language learning, that means helping students acquire good dictionary skills by using CD-ROM dictionaries. References Cambridge advanced learner s dictionary (with CD-ROM) (2003), (Cambridge: Cambridge University Press) Cambridge learner s dictionary (with CD-ROM) (2003), (Cambridge: Cambridge University Press) Collins COBUILD English dictionary for advanced learners, 3rd edn (2001), (Glasgow: HarperCollins Publishers) Longman dictionary of contemporary English NEW EDITION 4th edn (with CD-ROM) (2003), (Harlow: Pearson Education) Longman language activator (1993), (Harlow: Longman) Oxford advanced learner s dictionary of current English, 3rd edn. (1974), (Oxford: Oxford University Press) Oxford advanced learner s dictionary of current English, 6th edn (2000), (Oxford: Oxford University Press) 278

292 WORDS IN ASIAN CULTURAL CONTEXTS Abstract Narrative and the multimodal contexts of lexicography Ismail S Talib National University of Singapore ellibst@nus.edu.sg This paper is a reflection on a dictionary of narrative that I am now compiling. It intends to use this project as a starting point rather than as a basis for comparison with more general lexicographical projects that engage in the explanation of the lexicon of particular languages. On the fringes of my present project are the explanations of terms in the analysis of narrative that cannot be adequately explained using conventional methods, or arguably, cannot be explained at all without misleading the reader. These explanations will not be comprehensively pursued for my present book-length project, as they cannot be done completely, using the present method that I have at hand. However, my paper will prod at possibilities that may not only be relevant for a future specialised dictionary of narrative, but for future more general dictionaries of particular languages that are more closely based on discourse, or are based on an awareness of the multimodal contexts of the lexicon. As a starting point for the discussion although it is hoped that it will not be limited to the given examples the paper will discuss the notions of what on the surface may appear to be the simple concepts of flashback, voice-over narration, and what is known as the point-of-view shot in cinematic narrative. It will demonstrate how conventional print-based definitions, even if they are aimed at encyclopedic exhaustiveness, can never be adequate, and how a dictionary of the future may try to overcome this. Trick or chic: code-switching in colloquial Taiwanese Mandarin revisited Loretta Chung-Wing Tam Macquarie University, Sydney, Australia rettatam@graduate.hku.hk Abstract Taiwanese Mandarin establishes itself as a distinctive spoken variety featuring a strong influence of former colonial languages and local dialects. Its contemporary colloquial counterpart, in particular, consists of a Hokkien-influenced corpus. This paper examines the role of code-switching in colloquial Taiwanese Mandarin, and the issues concerning the nature of such adoption of foreign and/or dialect words in the colloquial, low variety of lingua franca. Though it is believed that a centuries-old tradition of borrowing does exist, the paper points out that the recent adoption of imported lexicon in colloquial Taiwanese Mandarin might well have been a form of code-switching, with the adhocacy in word choices as well as the much varied usage across individual ideolects, whereby personal sociolinguistic backgrounds play a determining and 279

293 ASIALEX 2005 shaping role. The paper also notes certain scholarly suspicion that such use might well contribute to a creolization of Mandarin Chinese in Taiwan, which in turn marginalizes the local variety and its speakers from the Mainland and General Standard Putonghua. While it is still far away from such a process for it mainly remains in the creative media as well as colloquial speech, and rarely formally gets into the formal domains of use as a high variety, the paper proposes a laissez-faire approach to language development, whereas deliberative measures to purify the language should be kept to a minimum in order to preserve an equilibrium of culturo-linguistic heterogeneity. Introduction The multilingualism in Taiwan has always been a focus of attention, amongst which being either from the society as a whole, or the academia. Plurality in language choice and use has long been predominant, and such tendency is nothing new to the general public. In this paper, I shall take note of the code-switching in colloquial Taiwanese Mandarin as a linguistic phenomenon by first, revisiting the differentiation of Taiwan Mandarin and Taiwanese Mandarin, whilst making an attempt to collaborate these two concepts into one, leaving only the term Taiwanese Mandarin. I shall argue that colloquial Taiwanese Mandarin may well be a spoken low form of the lingua franca in the light of a contemporary, egalitarian-based sociolinguistic understanding. Second, I shall relate such diglossia frameworking to the aforementioned claim on Taiwanese Mandarin and its colloquial counterpart, and state that long-established practice of multilingualism of Taiwan has encouraged, or at least not curbed, the import of lexicon from other dialects and/or languages into the existing colloquial corpora. Should borrowing be a quasi-ritual of adapting particular foreign lexis at a high, formal domain of the language, code-switching would have remained a dynamic, ad hoc, virtually individual choice of code(s) within a definite part of utterance(s) that should vary across the local, societal continuum. Sample data of a few identified types of code-switching in colloquial Taiwanese Mandarin are to be given and explicated toward the final part of this paper 2. Brief discussions on certain recent skepticisms concerning the increasingly popularized employment of Hokkien-Mandarin(-Japanese)(-English) code-switching in the local colloquial Mandarin speech, and on the issue of language purification, shall play the role of overall conclusion. Taiwanese Mandarin and its colloquial counterpart Talking about the variety(-ies) of Mandarin Chinese spoken on the island of Taiwan, the terminology employed has remained rather constant. Cheng reinstated the term Taiwan Mandarin ( TM ) for the variety which is learned and primarily as a second language, by the people of Taiwan (Cheng 1997: 29), claiming that most people s mother tongue was to be Taiwanese ( a variety of Minnan 1 ). Whilst, in Fon et al. (2004), that a further distinction between Taiwan Mandarin and Taiwanese Mandarin, echoing Cheng s demarcation. According to their definitions, the former refers to the official language of Taiwan and [m]any speakers of such a variety belong to the younger urban generation of Southern Min, while Taiwanese Mandarin refers to a variety of Mandarin on Taiwan that is heavily 280

294 WORDS IN ASIAN CULTURAL CONTEXTS influenced by Taiwanese phonology and syntax. And their speakers, contrary to the ones of TM, were raised in a Taiwanese environment and consider Taiwanese as the first language. They learned Mandarin through formal education and uses Mandarin mostly in formal contexts. (Fon et al 2004: 250) Such distinction, in a socio-historical sense, might have been logical. Well before the post-world War II baby boom, Mandarin and Taiwanese Hokkien were and made as binary opposites in the society of Taiwan (Tam 2004). Mandarin has been the lingua franca of the Mainland migrants, usually wartime veterans, intellectuals, merchants and their family; and Hokkien being the mother tongue of most local people who were of Southern Min ancestries. The conflicts between different lingua-cultural groups seemed inevitable, and such were further complicated with the polarized anti-communist policies in the 1960 s. However, should socio-cultural formation and identities of individuals be the key factor determining how varieties of a language should be typified, a few doubts on the might well be cast upon the aforementioned differentiation. The first goes to how to linguistically theorize the syntactic and lexical differences between Taiwan Mandarin and Taiwanese Mandarin, other than laying a perhaps overloaded emphasis on the phonological differences. Along the same continuum, from which such Taiwanese Mandarin is likely to draw upon the standardized Mandarin (and Taiwan Mandarin is clearly meant) for the basis of the language, it is obvious such attempt to cut clear boundaries in terms of only phonology and syntax would be far from being a sound enough argument. The second would have been the mother language issue. With a mass migration of Mainlanders and returning of Taiwan-originated people back to the Island since the late 1940 s, of various regional origins and speaking different home dialects, and the popular intermarriage of people from different language/dialect/ethnic group, it can well be anticipated that a single, unified concept of Taiwan Mandarin might elicit suspicion of being stereotyping; and so for the locally-bred Taiwanese Mandarin. And, as bilingualism and multilingualism have long existed in Taiwan, and boomed after the Japanese Occupation ( ), one might easily take note of the fact that many local Chinese children in the Japanese colonial years were speaking both their home dialect(s) and Japanese at home, which is very similar to how Mandarin and home dialect(s)/language(s) act as the mother tongues of most Taiwanese children in dual presence since the beginning of the Republican years (1949-present). A further danger of such terminology would be a marginalization of those whose mother tongues were neither Mandarin nor Hokkien, ruling them out from either the Taiwan or Taiwanese groups; example of such include the Hakkas, the aborigines, and the increasing population of expatriate spouses from Southeast Asia and their future generations. Hence, I shall propose that Taiwan Mandarin and Taiwanese Mandarin might well be incorporated into one single umbrella of referencing as Taiwanese Mandarin, and so as to dilute the often undesirable implications of the local label of Taiwanese 3. Under this definition, the variety of spoken Mandarin referred to by Fon et al. as heavily influenced by Taiwanese phonology and syntax can likely be included into the Colloquial form of the Standardized variety that is usually used in formal domain(s). Code-switching at work in colloquial Taiwanese Mandarin Code-switching is nowadays a general term which gains popularity to others 281

295 ASIALEX 2005 referring to various blending specific in particular forms and contexts, one of which being code-mixing 4. However there is yet another major concern after the clarification on the definition and terminology matters concerning code-switching in Colloquial Taiwanese Mandarin, i.e., the diglossia concern. The work of diglossia, a term and concept and model originated from Ferguson (1959) and later extended for wider coverage and use by Fishman in his 1967 model, is a key answer to the question of why code-switching should be a prominent issue in this colloquial variety rather than in the standardized one. Colloquial Taiwanese Mandarin, a variety often found in the everyday speech of common people, in this way that it might serve the low, informal domains of language use in the Taiwan society (the Low language), while the Standardized Taiwanese Mandarin would play the part as the chief medium of communication in the high, formal domains such as business, government and news media. Such distinction can however be blurred especially with one destabilizing factor: spontaneous language use and choice in the originally formal domains. An origin of it might have been the long-established practices of borrowing and code-switching, and the public tolerance under the Democrat government s current cultural policies, upholding the slogans of localizatiom and globalization. The following example of Mandarin-Hokkien code-switching might well be heard on the street in speech by laymen, but also in critical political commentaries by a guest speaker any local, public television programs (the boldfaced words being Hokkien): (1) gua 2 ka 7 li 2 kong 2, zhe1 ge shi1qing yi1 dian3 bu jian3dan1. (lit. I/ and/ you/ talk, this/ PART NUM/event-situation/ a/bit/not/simple-single) Let me tell you, this event is definitely unusual. And whilst the following Mandarin-Hokkien-Japanese code-switching might be found in local radio programs, the Taiwanese young people would not find it unfamiliar (the boldfaced being Hokkien, bold-italicized as transliteration from Hokkien to Mandarin, and the italicized as Japanese): (2) wo3 a 1 ma 7 sui1ran2 you3 dian3 gui1 mao2, dan4 shi1hai3 man2 kawaii o1! (lit. My/PART FAM-granny/though/but/have/bit/tortoise-hair/, but/is/still/quite/lovely/part STN END) Though my grandma is a bit picky on the things, I still think she s quite a lovely person! However such spontaneity is usually found in the more everyday-related media productions, such as talk shows, advertisements, sitcoms, and phone-in programs; as well as less than formal interpersonal communication like chats or friendly letters. Examples of Mandarin-English and Mandarin-Hokkien code-switching can also be spotted without difficulty, which might also appear in their respective written forms in the printed media or the Internet: the boldfaced being Hokkien, the italicized being English or transliteration from English to Mandarin):- (3) high fan1 tian1 (lit. high/over/sky Really very high ) (4) wo3 men2 de e-mei1-er2 shi4 (lit. I/PLUR/PART POSS/e-mai-l/is Our address is ) (5) bu4 yao4 zai4 king 2 le! 282

296 WORDS IN ASIAN CULTURAL CONTEXTS (lit. not/want/again/stubborn/part STN END- Oh please, don t insist anymore! ) Concluding Remarks Code-switching, a blend of other language codes with the matrix language, is getting more prevalent in Colloquial Taiwanese Mandarin utterances. Unlike borrowing, which is often found in the relatively high language and can only be attained through prolonged processes like transliteration, phonosyntactic adaptations, translation and gradual acceptance, code-switching is relatively fluid, ongoing, and is possible to be employed at will. Given that the choice of language codes and ways of switching is highly personal, and varies amongst ones of different socio-cultural backgrounds, code-switching not only functions as word/phrase or sentence-level blending of language codes but also functions much more than borrowing in terms of individual identity evolution. Similar to Cheng s commentaries on why borrowing in Taiwanese Hokkien has often been voiced against (Cheng 1997: 78-81), skepticisms that code-switching in colloquial Taiwanese Mandarin might lead to an eventual creolization of Mandarin Chinese on the Island is well observed in the recent decade. Not only that Mandarin Phonetics classes flourished in local universities extension schools, but also Chinese writing workshops in both formal education systems and the private tuition schools. This, basing on an implicated fear that the Standardized Taiwanese Mandarin might be further drifted away from Putonghua, its Mainland counterpart, which is used by a vast population and learnt by many speakers of other languages. I should suggest, and so as to conclude, that code-switching is still highly restricted in informal, spoken utterances, and the aforementioned fear does seem premature. Sustainable development of languages is deemed for balance, and unbound; the real-life multilingual Taiwan has proved itself to be such a case. Notes 1 Minnan is also referred to as Hokkien, Southern Min (for people and language from South Fujian). 2 The phonemic schemes adapted here are Hanyu Pinyin (Mandarin) and TLPA Reformed (Hokkien). 3 Implied in colloquial Taiwanese Mandarin expressions hen3tai4 ( very Taiwanese ), Tai2ke4 ( Taiwanese fella ), and the latter has been in use since a few decades back. 4 Relevant works in this area include Blom & Gumperz (1972), Baker (2000), Li (1996) and Luke (1999). Acknowledgements My sincere gratitude goes to M. H. Chen and J. C. Y. Wang for their kind updates on the Taiwanese dialects and languages, and their insightful comments on earlier drafts of this paper. References Chen, Ping (1999), Modern Chinese: history and sociolinguistics (Cambridge: Cambridge University Press). Cheng, Robert L. (1997), Taiwanese and Mandarin structures and their developmental trends in Taiwan II: Contacts between Taiwanese and Mandarin and restructuring of their synonyms (Taipei: Yuan-Liou 283

297 ASIALEX 2005 Publishing Ltd.). Fon, Janice et al. (2004), Production and perception of the two Dipping Tones (Tone 2 and Tone 3) in Taiwan Mandarin, Journal of Chinese Linguistics, 32(2), Hudson, R. A. (1996), Sociolinguistics (Cambridge: Cambridge University Press). Tam, Loretta Chung-Wing (2004), Myth or Miracle? On the multilingual language education policy in Taiwan, paper presented at the International Conference of Multilingualism in the Classroom: Eropean and Asian Perspectives, The Chinese University of Hong Kong, 7-8 April. The Malaysian English Newspaper Corpus: construction and function Abstract Siew Imm Tan University of Hong Kong tansi@hkusua.hku.hk The Malaysian English Newspaper Corpus (or MEN Corpus) is a 5-million-word synchronic corpus representing contemporary Malaysian newspaper English (MNE). The corpus comprises articles selected from two of the most authoritative and influential English-language newspapers in Malaysia the New Straits Times and The Star. The construction of the corpus began in May 2001 and was completed in December The present paper describes the construction of the MEN Corpus and provides a brief overview of the possible applications of this electronic corpus. The first section addresses: (1) issues that were taken into consideration in designing the corpus criteria for selection of newspapers, data acquisition, criteria for selection of texts and sampling of newspapers; and (2) the process of compiling the corpus. The second part of the paper illustrates how specific categories of lexical borrowing and lexical innovation can be identified and extracted from the MEN Corpus. By examining word lists and concordance lines generated using Wordsmith Tools Version 3, it is possible to study the contexts of specific lexical features in MNE. This allows speculation about the linguistic processes that led to this lexical change or creation, in the context of the constantly-changing sociolinguistic and language-contact conditions that have long prevailed in the region that is today Peninsular Malaysia. Introduction The Malaysian English Newspaper Corpus (the MEN Corpus) is a 5-million-word collection of newspaper articles sourced from the most influential English language newspapers in Malaysia. It was conceived as a synchronic database from which evidence of lexical borrowing and innovation in Malaysian newspaper English (MNE) could be drawn and analysed. This paper traces the process of constructing the corpus and provides a brief description of how diverse types of lexical features can be extracted for study. Construction of the MEN Corpus In selecting the newspapers from which to sample data for the MEN Corpus, three 284

298 WORDS IN ASIAN CULTURAL CONTEXTS main criteria were taken into consideration representativeness, influence and circulation/readership. The Star (STAR) and the New Straits Times (NST) were chosen because they have good-quality printed daily editions and they were the most widely-read English language newspapers in Malaysia. In March 2001, I approached the publishers of both newspapers to discuss the possibility of purchasing their newspaper articles in electronic format. NST agreed to sell me the ASCII 1 version of selected articles from each issue of newspaper in the form of zipped files, but STAR declined to sell me their articles advising that I could download them from their online editions. The compilation of data from NST was thus relatively simple mainly unzipping files and saving them in ASCII format while for STAR it involved downloading relevant articles, copying them and pasting them into a word document. From the two newspapers, articles from the news and opinion categories (see Bell [1991] for classification of text categories in newspapers) that were written by Malaysians were selected to be included in the corpus. In order to ensure that a balanced spread of issues across different days of the week was achieved (see Ljung [2000] for the importance of considering the cyclicity of newspapers), it was decided that STAR and NST would be sampled on alternate days, the former beginning 1 August 2001 and the latter 2 August The compilation of the MEN Corpus ended on 30 January 2002, by which time, the 5-million-word target was achieved. The data was then cleaned 2, annotated with some minimal identifying information 3 and saved in ASCII format. The MEN Corpus, unlike some of the major corpora available today, is not grammatically-annotated as the resources necessary for such annotation were not available to me. Function of the MEN Corpus As it stands today, the MEN Corpus constitutes a wealth of linguistic data that can be used to shed some light on the types of lexical borrowing and innovation that have emerged from the evolving language contact situation in the area that is today Peninsular Malaysia. To establish the presence of these specific features as well as their patterns of distinctiveness, I have used the newspaper component of the British National Corpus (henceforth BNC News), a 9.3-million-word collection of British English newspaper texts from 1975 onwards, as a reference corpus. Below, I shall illustrate how the MEN Corpus can be used, either alone or in conjunction with the BNC News, to identify lexical features that contribute to the distinctiveness of MNE. Based on Winford s (2003) review of the phenomena of language maintenance 4 1 ASCII stands for American Standard Code for Information Interchange. It is a code for representing characters as numbers, with each letter assigned a number from 0 to 127. Data in ASCII can be read by most computers and software. 2 Data sourced from STAR had numerous formatting details and encodings, such as font type and size, paragraph formatting, accented characters, ellipses, bullets, fraction characters, em and en dashes and smart quotes, which were not ASCII-compatible. As the data was ultimately to be saved in ASCII format, these features had to be replaced with ASCII-compatible features. Thus, smart quotes were replaced with straight quotes, ellipses were replaced with three consecutive periods, accented characters were replaced with DOS characters, etc. The clean-up process was minimal for NST articles which I received in ASCII format. 3 For each article, a set of five tags was used to indicate the name of the newspaper <T>, date of publication <D>, headline <H>, subheading <S>, note <N> and byline <B>. 4 I am aware that, from Winford s perspective, the case of Malaysian English does not qualify as an example of language maintenance resulting in lexical borrowing. Lexical borrowing, in the sense intended by Winford, is a process of importing features from a foreign language, and the process is always initiated 285

299 ASIALEX 2005 and language shift, I have isolated several categories of lexical borrowing and innovation that have a strong presence in MNE. The process of lexical borrowing results in instances of loanwords, loanblends, loan translations and native creations, all bearing the evidence of contact between the English language and the languages of the larger ethnic groups in Malaysia Malay, Hokkien, Cantonese and Tamil. Besides lexical borrowing, the MEN Corpus also yields evidence of lexical innovation that appears to have resulted from substratum influence and speakers attempts at regularising the structure of the English language. The purpose of this paper is not to explore the whole range of lexical borrowing and innovation in MNE but to show how the MEN Corpus might be used to extract these types of features for study. Accordingly, only a handful of these features will be discussed here. The corpus analysis software that I have used to analyse both the MEN Corpus and the BNC News is Wordsmith Tools Version 3. This software package has three main programmes Wordlist, Keywords and Concord and they can be use to examine how words in a collection of texts are used and therefore also how certain usages or senses of a particular word are more typical of MNE than other varieties of English. Loanwords are the most prevalent and easily identified types of borrowings in MNE. A quick scan of the alphabetical word list of the MEN Corpus will reveal hundreds of instances of words of Malay, Chinese, Tamil or Arabic origin. For example, from the first column of Table 1, we can see instances of possible Malay loanwords such as datuk and datin as well as the different forms of these words. Datuk, for example, has gained three English morphological units the possessive s, the plural -s, and the suffix -ship. The second column contains entries that are likely romanised Chinese characters cheo, cheok, cheong as well as the Cantonese loanword, cheongsam, and its plural form cheongsams. N Word Freq. N Word Freq DATIN CHEO DATIN S CHEOH DATIN-DATIN CHEOK DATING CHEOK S DATION CHEOK-YONG DATO CHEONG DATO-DATO CHEONG S DATOH CHEONGSAM DATOK CHEONGSAMS DATSAKOM CHEOR DATSAKORN CHEOW DATSON CHEP DATSUN CHEPA DATTEN CHEPAKA DATTUYAWAT CHEPOR DATU CHEQUE DATU-SANDAKAN-TELUPID CHEQUE S DATU-TAWAU CHEQUE-DEPOSIT DATUK 8, CHEQUEBOOK DATUK S CHEQUERBOARDS DATUK-DATUK CHEQUERED DATUKABDUL CHEQUES DATUKS CHEQUESCAN 9 by the native speakers of the recipient language. Native speaker agency is a criterion about which I raise further questions in my PhD dissertation. 286

300 WORDS IN ASIAN CULTURAL CONTEXTS DATUKSHIP CHER DATUKSHIPS CHERAKAH 3 Table 1. Fragments of the alphabetical word list from the MEN Corpus illustrating potential Malay and Chinese loanwords In order to examine particular loanwords in context, one has to generate the concordance lines of these words. This is done by specifying a search word in Concord, which will then locate all instances of the word and present each instance together with a pre-specified amount of context. For example, the concordance display illustrated in Table 2 is generated when the search word baju is specified. Baju, in Malay means garment, dress, shirt, etc 1 though the word is used in combination with many other Malay words to refer to different types and styles of attire (e.g., baju alang, baju bersayap, baju dalam, baju hujan, baju kurung, baju kebaya, etc). 2 The word baju is defined in the Oxford English Dictionary (OED) as a short loose jacket worn in Malaya. The presence of this word in the OED seems to be a fair indicator that baju has been absorbed into the lexicon of BrE. This word, however, does not occur in the BNC News, which is not entirely surprising. What is notable is that the word baju, used either in the general sense of garment, dress, shirt, etc or in the sense denoted in OED, does not occur in the MEN Corpus either! If we refer to Table 2, we see that baju occurs 32 times in the MEN Corpus, only once singly (see entry number 32). That is to say, in all but one case, the term is used in conjunction with other Malay words to form compound loans like baju kebaya (5), baju Kedah (2), baju kurung (15), baju Melayu (8) and baju panjang (1). In the one instance when baju occurs singly (as baju s), it is in reference to any of three specific types of traditional Malay dress baju kurung, baju Kedah or baju Kebaya which were mentioned a little earlier in that text. 1 use of intan, was usually worn with the baju kebaya labuh or kebaya pendek and were 2 Rahman s 63rd birthday. Dressed in black baju kebaya with a beige and blue floral motif, 3 as well as the more contemporary curvy baju kebaya. Each of these fashion pieces is said 4 glass cases, waiting for their day on a baju kebaya. At an interview in Aseana, Sim 5 goodness, a whole host of gorgeous silk baju kebaya. The serene gentleness hiding, of 6 choices of baju Melayu cekak musang, the baju Kedah and even some hip togs to add some 7 the classical baju kurung and elegant baju Kedah as well as the more contemporary 8 KLCC. It was unfortunate that the tailored baju kurung and baju melayu for my children 9 coloured stones, they will rival the brightest baju kurung and give the Christmas tree a run 10 pieces in the collection include the classical baju kurung and elegant baju Kedah as well as 11 guests also turned up in batik-print shirts, baju kurung and baju Melayu. Some were 12 baju kurung. We can see that some of their baju kurung are so thin that one can see through 13 geometric prints, this alternate range of baju kurung comes in multi-hued attractive 14 spade a spade, often eschews the traditional baju kurung for the modern pant-suit and openly 15 hands wearing bangles, the other with the baju kurung sleeves and the other one just plain. 16 who only made traditional attire such as baju kurung, kebaya and cheongsam but I ve 17 One for school comprising mainly of baju kurung, longer skirts or even kebaya labuh 18 Sri Lumandeng, who was clad in a green baju kurung, looked calm throughout the 19 this year. It s easy to simply pick and buy baju kurung, which are quite cheap at the 20 pottu on her forehead whereas Siti is in her baju kurung. Another painting shows girls 21 used by secondary schoolgirls for their baju kurung. We can see that some of their baju 22 the room waiting for us, dressed in a pink baju kurung. We exchanged greetings as she 23 The baby was wrapped in a blood-stained baju Melayu and found at the roadside, said 24 is never complete without the traditional baju Melayu and men in Johor will accept 1 Awang and Yusoff, Ibid,

301 ASIALEX Lumpur. Take a look at our choices of baju Melayu cekak musang, the baju Kedah and 26 man can expect to be clad in finely tailored baju Melayu during the festive occasion. These 27 that the tailored baju kurung and baju melayu for my children were not ready yet. 28 will take part in march past here. It is Baju Melayu for men and the Kota Baru kebaya 29 children, he added. Clad in the same white baju melayu that he wore on his wedding, Mohd 30 up in batik-print shirts, baju kurung and baju Melayu. Some were spotted donning the 31 actress Pearly Chua will once again don her baju panjang and twist her hair into Emily Gan 32 blue. Fine floral embroideries sewed on the baju s hemline enhance the feminine essence of Table 2. Concordance display of baju from the MEN Corpus Two observations can be made based on this analysis. First, there are at least four compound loans in MNE that incorporate the word baju baju kebaya, baju Kedah, baju kurung and baju Melayu. Second, there is no evidence in the MEN Corpus to suggest that baju, as a single term, is part of the lexicon of MNE. These two observations seem to indicate some degree of concurrence with the proposition that some loanwords are motivated by the need to compensate for semantic gaps in the recipient language. As the sense of garment, dress, shirt, etc is already adequately expressed using existing English words, there is little need to import baju into MNE. There is however no English words that can accurately express the concepts of traditional Malay or Peranakan wear. Unless one is willing to put up with cumbersome terms like traditional Malay shirt or Peranakan-style blouse with frontal opening, one is likely to import the local terms into one s English and if the local terms are subsequently adopted by other users, they become established as loanwords. Other Malay words that appear to share the same patterns of incorporation into MNE are kain 1 and nasi. 2 In terms of loanblends, an interesting subcategory that I have investigated is a class of derivational blends that are the results of affixation to toponyms. Toponyms are particularly interesting because they are often thought to be some of the earliest words borrowed into new varieties of English (see Schneider 2003b:52). In the case of MNE, affixation to a place name to denote a native or inhabitant of the place or the sense of being characteristic of the place is very common. For instance, suffixation to Kedah, the name of a north-western state in Peninsular Malaysia, results in the blend Kedahan, which is used to denote a native or inhabitant of the state of Kedah (noun) or the sense of relating to or being characteristic of the state of Kedah or its people or language or culture (adjective). Other examples of toponym-derived blends include Penangite, Malaccan, Johorean, Kelantanese, Sabahan and Sarawakian. Concord is also very useful for detecting evidence of ME users attempts at regularising the structure of the English language. Several English verbs, such as discuss, demand and request, are particularly prone to being regularised. To illustrate this point, I shall use the relevant concordance lines of different forms of the verb discuss discuss, discusses, discussing and discussed. A Concord search using the following search parameters: search word: discuss*; but excluding: discussion* yielded 850 lines, too many to be presented here. An analysis of this display revealed 1 Kain, which is a generic Malay name for textiles, does not occur by itself in the MEN Corpus. Instead, kain forms a part of compound loans like kain samping, kain songket and kain tenun. 2 Nasi is cooked rice in Malay, and is also thus defined in the OED. Nasi does not occur in the BNC News. The word occurs 108 times in the MEN Corpus, but never as a single loanword. Instead, it appears as part of compound loans such as nasi ayam, nasi goreng, nasi lemak, etc. 288

302 WORDS IN ASIAN CULTURAL CONTEXTS two distinctly Malaysian usages: discuss (verb) about 1 and discuss (verb) on. Tables 3 and 4 present all instances of these two constructions in the MEN Corpus. An examination of the contexts of discuss (verb) about and discuss (verb) on reveal no obvious difference from the contexts of discuss (verb). Table 3. Concordance lines of discuss (verb) about 1 mufti Datuk Murtadza Ahmad, would discuss about the Black Metal cult which involved 2 from the desire to act and not merely discuss about the social ills that plague the society. 3 in a special session tomorrow to discuss about the news network before its launch the 4 he said. Zefferys said they also discussed about the existence of certain quarters who 5 Syed Hamid said they also discussed about bio-terrorism and the possibility of 6 Omar said the leaders would be discussing about enhancing the climate for foreign Table 4. Concordance lines of discuss (verb) on 1 he said. The IBF are expected to discuss on the current world ranking system at 2 writers, literary figures and culturist to discuss on the identity of Malays by looking at a 3 meet the coaches first next week to discuss on the programmes. After that, we will 4 in this first phase is to get everyone to discuss on a blueprint for motorsports, which my 5 the border with Afghanistan. We also discussed on the need for an Afghan government 6 Dr Mahathir also said that the discussed on the reported underground anticouncil 7 to mid-20s group. They are already discussing on how packages can be customised to The structural congruence between discuss (verb) about and discussion (noun) about, and between discuss (verb) on and discussion (verb) on have led me to hypothesise that the verbal constructions were innovated based on the nominal constructions. In other words, the feature came into existence when certain ME speakers assumed that if one can talk about a discussion about bio-terrorism, it follows that one can also say... they also discussed about bio-terrorism... (see entry 5 of Table 3). The same regularising strategy seems to be at work in the cases of demand (verb) for (which mirrors demand [noun] for ) and request (verb) for (which mirrors request [noun] for ). Conclusion The few instances of lexical borrowing and innovation given here are, of course, insufficient as a basis for the formulation of any conceptual framework, and it is certainly not the objective of this brief overview to endorse any sociolinguistic or language contact theories that have been proposed by scholars in this field (e.g., Platt and Weber 1980, Lowenberg 1984, Schneider 2003b, etc). Rather, the instances cited here illustrate the capacity of the MEN Corpus as a tool to uncover various categories of lexical features that are peculiar to MNE, and this is a pre-requisite for understanding how the English language has evolved in the complex sociolinguistic and language-contact conditions of the region that has come to be known as Peninsular Malaysia. References Awang Sudjai Hairul, and Yusoff Khan Kamus Lengkap. Selangor: Pustaka Zaman. Bell, Allan The language of news media. Oxford: Blackwell. 1 The word verb in parentheses after the word discuss indicates all forms of the verb discuss. 289

303 ASIALEX 2005 Ljung, Magnus Newspaper genres and newspaper English. In Friedrich Ungerer, ed., English media texts past and present: language and textual structure, Amsterdam/Philadelphia: John Benjamins. Lowenberg, Peter H English in the Malay Archipelago: Nativization and its functions in a sociolinguistic area. Ph.D. diss., University of Illinois, Urbana-Champaign. Platt, John, and Heidi Weber English in Singapore and Malaysia: Status: Features: Functions. Kuala Lumpur: Oxford University Press. Schneider, Edgar W. 2003a. The dynamics of new Englishes: from identity construction to dialect birth. Language, 79(2), Schneider, Edgar W. 2003b. Evolutionary patterns of new Englishes and the special case of Malaysian English. Asian Englishes 6(2), Winford, Donald An introduction to contact linguistics. Mass: Blackwell. Abstract The Malayness of the Teochew lexicon Ying Ying Tan National University of Singapore elltyy@nus.edu.sg The dynamics of language contact between two speech communities often presuppose an existing social and historical connection between them. This is precisely the case for Teochew in Singapore. The main characteristic of Teochew in Singapore is its incorporation of a sizeable number of Malay words, appropriated into Teochew, with a systematic allocation of Teochew tones to the Malay words. This therefore suggests a close Teochew-Malay interaction, although the extent and nature of this interaction remains largely undocumented. This paper hopes to provide some reasons for the adoption of Malay words into the Teochew lexicon. This paper will also acknowledge that a historical, geographical, and sociological investigation between the Malay and Teochew communities might provide some interesting insights into the true dynamics between them. A study of the geography of the Malay and Teochew kampongs in the 1930s-1960s will focus especially on the following: Geylang Serai, the trades and social activities the two communities were involved in, the inter-ethnic communication, and various perceptions and attitudes. This paper will further illuminate the Malay-Teochew relation that is first suggested by the linguistic structure of Teochew. What words do Hong Kong students know? A study of lexical input from textbooks Eunice Tang Chinese University of Hong Kong tangeunice@cuhk.edu.hk Abstract In most English as a second language (ESL) or English as a Foreign Language (EFL) 290

304 WORDS IN ASIAN CULTURAL CONTEXTS Asian countries, the principal source of lexical input for acquisition is from textbooks. This paper reports a research project conducted in Hong Kong where English is a compulsory subject in the school curriculum. The research project was designed to i) compile a corpus of English words available for acquisition in the local English textbooks for primary and junior secondary school students; ii) measure the richness of lexical input from the textbooks; and iii) verify the nature of lexical input. The major local English textbooks were processed to generate textbook frequency lists. The lexical variation (type-token) ratio was calculated, using WordSmith Tools (Scott 1998). The textbook word lists were then compared with General Service List (West 1953), VocabProfile (Nation 1986) and Academic Word List (Coxhead 1999) to determine the kind of words that Hong Kong students encounter, after receiving nine years of formal English teaching and learning, at the age of 15. Results have shown that the Hong Kong students are exposed to an impoverished lexical environment. This can be reflected by the lexical variation ratio, nature of vocabulary, and variety of words. Introduction When the close relationship between vocabulary size and English competence was widely recognized in literature (see Richards 1976; Laufer 1986; Carter and McCarthy 1988; Nation 1990; Ellis et al. 1995; Laufer 1997; Wesche & Paribakht 1999; Schmitt 2000; Nation 2001), most work on second language vocabulary teaching also seemed to agree that language learners should first concentrate on the high frequency words of the language. Nation and Newton (1997:239) suggested that the 2,000 high-frequency words of English should receive attention first because without these it is not possible to use English in any normal way. Beyond the 2,000 high frequency words, Nation and Waring (1997) suggested that second language learners who go on to academic study would have a need for general academic vocabulary. If the high frequency words and academic words are what the second language learners need, these words should be readily available to the learners. In most English as a second language (ESL) or English as a Foreign Language (EFL) Asian countries, the principal source of lexical input for acquisition is from textbooks. However, research work in this area is limited. Tang (2002) studied the nature of vocabulary input from the glossary of the College English textbooks in China. No similar study was then found. In this paper, I will report a research project conducted in Hong Kong where English is a compulsory subject in the school curriculum. This purpose of the study was to i) compile a corpus of English words available for acquisition in the local English textbooks for primary and junior secondary school students; ii) measure the richness of lexical input from the textbooks; and iii) verify the nature of lexical input. Methodology To determine the kind of words that Hong Kong students encounter after receiving nine years of formal English teaching and learning, at the age of 15, major local English textbooks were processed to generate textbook frequency lists. Two sets of major English textbooks for the six levels (P1 to P6) of primary education and four sets of major English textbooks for the three levels (S1 to S3) of junior secondary level were selected (see Table 1). At primary level, publishers A and B had one set of textbooks. At secondary level, publishers A and B produced two sets of textbooks (thereafter version 1 and version 2 ). According to the English teachers, version 2 seemed to be easier 291

305 ASIALEX 2005 and simpler in terms of the length of the texts, complexity of the sentences and level of lexical difficulties. Table 1: Distribution of the textbooks 1 Book Level Sub-total Primary PA 1a-b, 2a-b, 3a-b, 4a-b, 5a-b, 6a-b 12 PB 1a-d, 2a-d, 3a-d, 4a-d, 5a-d, 6a-d 24 Secondary SA1 1a-b, 2a-b, 3a-b 6 SA2 1a-b, 2a-b, 3a-b 6 SB1 1a-b, 2a-b, 3a-b 6 SB2 1a-b, 2a-b, 3a-b 6 Total 60 In order to compile the textbooks corpus, all pages of the textbooks were scanned and saved as text file format. Careful spell checking and editing were done before they were transferred to the computer program called WordSmith Tools (Scott 1998) to generate frequency counts. To assess lexical richness provided by the textbooks, the type-token ratio was calculated using the same computer program and was analysed using the measure of lexical variation (LV) adopted by Meara, Lightbown and Halter (1997) and Brown and Sagers (1999). It was assumed that if a high LV was found, it indicated that a rich lexical environment existed. Finally, the textbook word lists were then compared with General Service List 2 (West 1953), VocabProfile 3 (Nation 1986) and Academic Word List 4 (Coxhead 1999) in order to find out the nature of words available for acquisition. The wordlist files were compared to check for the common and different words, using compare two lists from WordSmith Tools 3.0. Findings Textbook word lists A total of sixty word lists with frequency counts were generated from each single textbook. Word lists from the same series were then combined to form another six textbook wordlists. A comparison of the textbooks between publishers A and B showed that their books had 58% to 63% of words in common (see Table 2a). The textbook word lists also revealed that both publishers used the majority of the words in both version 1 and version 2, but version 1 contained more different words than version 2 (see Table 2b). This could probably be affirmed the 1 The textbooks used in this study were: New Welcome to English, Longman Express, Longman Target English, published by Longman and New on Target, New Oxford English, Oxford English Treasure published by Oxford University Press. 2 Nation (2001) claimed that this classic list of 2,000 high frequency words is still the best available list which covers 80% of the running words in spoken and written texts in all kinds of uses of the language. 3 Nation (1996) divided the most frequent words in English into the first, second and third 1,000 word lists. The three lists of words can be found in the computer programme VocabProfile. The first list (VP1) includes the most frequent 1,000 words of English; the second (VP2) includes the second 1,000 most frequent words, and the third (VP3) includes words not in the first 2,000 words of English, but frequent in upper secondary school and university texts from a wide range of subjects. 4 Coxhead (1998) replaced University Word List (UWL) with an Academic Word List (AWL) containing 570 word families. 292

306 WORDS IN ASIAN CULTURAL CONTEXTS observation of the English teachers that version 2 is easier and simpler than version 1. Table 2a: Comparison of textbooks in terms of lexical input (between publishers A and B) Textbook Common words Textbook PA (1,140 / 37.57%) 1,894 PB (1,201 / 39.26%) SA1 (2,559 / 41.64%) 3,587 SB1 (2,438 / 40.47%) SA2 (1,563 / 34.55%) 2,961 SB2 (2,229 / 42.95%) Table 2b: Comparison of textbooks in terms of lexical input (between version 1 and version 2 ) Textbook Common words Textbook SA1 (2,533 / 41.21%) 3,613 SA2 (911 / 20.14%) SB1 (2,579 / 42.81%) 3,446 SB2 (1,744 / 33.60%) Lexical input from textbooks The overall picture of the lexical richness provided by the textbooks was viewed by examining the LV ratio in three different ways: i) the lexical variation of individual textbook; ii) the lexical variation of textbook series at primary and junior secondary levels; iii) the lexical variation of textbook series from primary to junior secondary. At individual textbook level, the LV ratio observed in primary English textbooks was comparatively higher than that in the secondary English textbooks (Table 3). However, the LV ratio was still low. Table 3: Lexical variation of individual textbook Level PA PB Tokens Types LV ratio Tokens Types LV ratio P1 3, , P2 4, , P3 5, ,140 1, P4 9,863 1, ,476 1, P5 11,255 1, ,802 1, P6 14,196 1, ,979 1, Level SA1 SB1 Tokens Types LV ratio Tokens Types LV ratio S1 37,249 2, ,509 3, S2 45,149 3, ,221 3, S3 43,541 3, ,926 3, Level SA2 SB2 Tokens Types LV ratio Tokens Types LV ratio S1 29,173 2, ,364 2, S2 31,518 2, ,242 3, S3 36,159 2, ,837 3, The LV ratio dropped considerably when it was calculated across cumulative 293

307 ASIALEX 2005 levels at primary and junior secondary levels (Table 4). Within 100 running words in the textbooks, there were only 3.47 to 6.24 different words. The LV ratio dropped even further when the whole textbook series were combined (Table 5). There were about 2.79 to 4 different words in a text of 100 words. The LV ratio within the nine years of English education was consistently low, suggesting that learners were exposed to a lexically impoverished environment. It could also suggest that these words reoccurred frequently in individual textbook and across levels and years of study. The repeated exposure to the words creates a favorable condition for successful vocabulary acquisition (Krashen 1982). Table 4: Lexical variation of textbook series (primary and junior secondary) PA PB Tokens Types LV ratio Tokens Types LV ratio P1 P6 48,659 3, ,766 3, SA1 SB1 Tokens Types LV ratio Tokens Types LV ratio S1 S3 125,939 6, ,656 6, SA2 SB2 Tokens Types LV ratio Tokens Types LV ratio S1 S3 96,850 4, ,603 5, Table 5: Lexical variation of textbook series (from primary to junior secondary) PA + SA1 PB + SB1 Tokens Types LV ratio Tokens Types LV ratio P1 S3 174,374 6, ,442 6, PA + SA2 PB + SB2 Tokens Types LV ratio Tokens Types LV ratio P1 S3 145,237 5, ,413 6, Nature of lexical input from textbooks Table 6 below displays the distribution of words in the English textbooks. When the textbook word lists matched with GSL, VP3 and AWL, a comparatively higher percentage was observed in GSL. The result was not surprising as these basic words should be given priority in sequencing. Table 6: The distribution of words in the English textbooks PA PB SA1 SA2 SB1 SB2 No. % No. % No. % No. % No. % No. % GSL VP AWL other Total It was noted that the largest proportion of words fell into the category of other. A quick and brief look of the other words revealed that there were: proper nouns, e.g. Brian, Spanish, Delifrance; 294

308 WORDS IN ASIAN CULTURAL CONTEXTS borrowed words, e.g. shabu shabu, karaoke, croissants; words related to science and technology, e.g. ion, toxic, download, multimedia, websites; spoken expressions, e.g. oops, ok, er; words related to Chinese or local culture, e.g. chopsticks, McMug, minibus; food, e.g. mussels, spaghetti, brownie, stew, durians; words related to the classroom and living environment, e.g. literacy, habitat, indoor, blackboard; animals and birds, e.g. parrots, claws, hare; words related to language learning, e.g. dyslexia, semantic, suffix. When the GSL is considered as the most useful words for all ESL learners, by the age of 15 or after nine years of English education, Hong Kong students learn slightly more than half of the basic words (Table 7). At junior secondary level, they had minimal exposure to the general academic words too. The result implies that these students might encounter problems in daily communication and future academic study. Table 7: The percentage of words in the selected ESL wordlists GSL VP3 AWL PA 50.53% 0.90% 0.80% PB 50.63% 0.73% 0.74% SA % 7.10% 7.94% SA % 4.54% 5.14% SB % 5.58% 5.69% SB % 4.41% 5.05% Discussion and Conclusions This study was designed to measure the richness of lexical input from the textbooks and to verify the nature of lexical input. Results have shown that Hong Kong students are exposed to an impoverished lexical environment. The low LV ratio at each school level and across the nine years of English study suggested that limited variety of words was available for acquisition. This finding is alarming as it might have a strong impact on the vocabulary growth of the students. The low LV ratio also suggested that there was a heavy repetition of words within a short and long period of time, therefore it will provide an opportunity for transforming the receptive vocabulary to a productive one if teachers are aware of the known words to the learners and if appropriate teaching methods are adopted. Since textbook is the principal source of English lexical input, the selection and sequencing of words are of pedagogical importance. However, the textbooks could only cover slightly more than half of the 2,000 most frequent words after nine years of English study. Furthermore, there was limited coverage of academic words which students need in order to move on to their study. The inadequate input of useful words will not allow students to perform effectively with the language in any communicative contexts. The textbooks provide a large number of other words for acquisition. However, the nature of these other words is unknown. It is also not sure about the perception of the textbook writers on the vocabulary needs and vocabulary input for the Hong Kong students. Further and in-depth examination is required to identify patterns and 295

309 ASIALEX 2005 usefulness of these words and the vocabulary demand of the students. References Coxhead, A. (1998), An academic word list. English Language Institute Occasional Publication. No.18. (Wellington: Victoria University of Wellington.) Laufer, B. (1986), Possible changes towards vocabulary acquisition research. International Review of Applied Linguistics. XXXIV/1, pp Laufer, B. (1997), The lexical plight in second language reading. In Coady, J. & T. Huckin. (eds.) Second Language Vocabulary Acquisition. (CUP), pp Nation, I.S.P. (1990), Teaching and Learning Vocabulary. (Rowley: Newbury House). Nation, I.S.P. (2001), Learning vocabulary in another language. (Cambridge: Cambridge University Press.) Nation, P. and J. Newton. (1997), Teaching Vocabulary. In Coady, J. & T. Huckin. (eds.) Second Language Vocabulary Acquisition. (CUP), pp Nation, P. and R. Waring (1997), Vocabulary size, text coverage and word lists. In Schmitt, N and M. McCarthy (eds.), Vocabulary: Description, Acquisition and Pedagogy. (Cambridge: Cambridge University Press.) Richards, J. C. (1976), The role of vocabulary teaching. TESOL Quarterly, Vol.10(1): pp Schmitt, N. (2000), Vocabulary in language teaching (Cambridge; New York: Cambridge University Press). Tang, E. (2002), English Vocabulary Input in the Tertiary Classroom in China, Unpublished thesis, University of Warwick, UK. Wesche M. and T. S. Paribakht (eds.). (1999), Incidental L2 Vocabulary Acquisition: theory, current research, and instructional implications (Cambridge: Cambridge University Press). Nominal, verbal and adjectival structures in Japanese and Turkish Yuanke Tao Guangdong Federation of Social Sciences, China yorke_tao@yahoo.com Abstract Turkish and Japanese both have the basic syntactic structure of SOV. The paper offers some valuable linguistic proofs to reveal their basic differences and similarities, focused on nominal, verbal and adjectival structures. The languages do not have prefixes that determine the grammatical property of compounds, and the difference of their prefixes is of lexical meaning. Since the 17th century, a tendency has developed that some final components of nominal structurse play a role of nominal signal in Turkish and it has abounded nominal suffixes, but nominal suffixes are not found in Japanese. Suffixes are abounded in their verbal structures, but appear in different structures and have separate origins in Japanese and Turkish. Noticeably in Japanese, there are some light verbs played as verbal suffixes, but there is lack of personal suffixes, perfect and participle suffixes like those of Turkish. Turkish has much more adjectival suffixes than verbal suffixes, but Japanese has only two adjectival suffixes. Japanese has specific adjectival suffix for forming connective adjectives, however, Turkish adjectival components do not have such an option of suffix. The critical difference between Japanese and Turkish happens in connection between prior component and post component in compounding 296

310 WORDS IN ASIAN CULTURAL CONTEXTS structures. It might be caused for the difference of certain historical syntactic structures in the development of the two languages. But more information about the historical deviation of suffixes in dictionaries is demanded for further studies. Both Turkish and Japanese have a syntactic structure of SOV, but their identifications of relationship are arguable. Some people tended to think that Turkish and Japanese belong to a same family and supposed they have a proto-relation (Miller 1971). Some other people, such as Hartman and Stork (1973), considered that, Turkish is a western branch of Ural-Altaic Language Family, while Japanese is a branch of Japan-Korean Language Family. And other people just put it as an unsettled problem, as Underhill did in his paper Turkish (1986). To settle the problem, it is needed more comparison of structures at deep level, in addition to historic and textual studies. I ll compare Turkish with Japanese here, focused on nominal, verbal and adjectival compounding structures. Prefixes According to Robert Underhill (1986), except for two prefixal processes involving reduplication, Turkish is exclusively suffixing. But, based on the record of A Turkish English Dictionary (Hony, 1967), there are more than 10 prefixes in Turkish, such as: Ala- in style : ~franga in European style ; **Bade- (prep.) after : ~ttasdik after confirmation ; *Bi- (prefix) without : ~baht unfortunate ; Bilâ- without : ~perva without fear ; Çal-, çala- indicating repeated action of the verb : ~çene talkative, ~kamak shake ; Ke- as, like : ~levvel as at first ; ~mal (like property etc.) perfection, worth, value ; Kema- as, as that which : ~filevvel as formerly, ~fissabιk as it used to be ; Lâ- Arabic neg. particle used in Arabic compounds or phrases : ~yemut immortal ; **li- For, in favour of, on : ~ecli for the purpose of ; mabih- the thing of which there is, cause : ~ülehtiyac what is necessary, ~üliftihar cause of pride ; **min- from, by means of, among ; na-, nâ-, negative prefix : ~durust, dishonest ; pür- full of : ~pürhiddet full of fury ; sui- evil : ~ahiâk immorality, ~hareket evil action. Judged by their semantic characteristics, the cases which are marked with ** should be all taken as prepositions, though in the three only bade- is marked as preposition, compared with the others li-, min- marked above. While bi- should be taken as a prefix according to its sense, though it is labeled as preposition in the dictionary. So, there are total ten prefixes at least in Turkish. Moreover, Turkish has a lot of adjectives which lexical meanings can be strengthened by duplicating first syllables and adding a consonant. For examples (Lewis, 2000): Başka other bambaşka totally different Çabuk quick çarçabuk very quickly Dolu full dopdolu chock-full Koca large koskoca very eminent Mavi blue masmavi very blue Sıkı tight sımsıkı very tight Temiz clean tertemiz absolutely clean Uzun long upuzun extremely long Because the syllables of duplication vary according to the correspondent alliterative syllables of different adjectives, which they duplicate, we do not treat them as prefixes. 297

311 ASIALEX 2005 All prefixes marked in the dictionary have lexical meanings, but are non-grammatical morphemes in compounds, no matter if the compounds contain a grammatical suffixes as mabih-: ~ülehtiyac, ~ülihticac etc. or not, as ke- as, like, ~levvel as at first, ~mal like property, wealth etc. Japanese prefixes express lexical meanings but not grammatical meaning as well. A unique distinctive prefix 御 (o-) or (ogn) needs to be noted here, for it expresses honorific meaning in compounds, such as [o ma su] there is/are and [ogn ka ta] your excellency and so on (Tao, 1999). Differences between Japanese prefixes and Turkish ones are of their amount and lexical meanings. The compounding components of the both can be described in one pattern: lexical prefix + root(s) + grammatical suffix. Nominal structures Languages with some typological similarities may not be the languages of the same family. So, to tell family similarity from typological one in our study, certain historical comparison is significant. Here I d like to introduce some data of 17th century s Turkish. Our study will not be restricted to compounds, but cover nominal structures instead, so as not missing any structures that should be considered as nominal compounds. Based on Gilson s study (1987), which focused on the 17th century s Turkish structures, it has 3 types of constructions combined two or more nouns, which are interdependent. (1) (N1+(n)Iη) + (N2+(s)I). For example, kahvé bahasy the price of Coffee ; halký karé the people s land ; turché tangry Turkish God. N1 is a grammatical possessor of N2, while η and (s)i are phonetic notes indicated the final pronunciations of suffixes in the two compounding components respectively. (2 )(N1) + (N2+(s)I). (s)i again indicates a dependent relation of N2 to N1, which can be used independently and classified as postposition. For example, ne bairac altindé under what Ensign ; dostler arasindé among Friends ; divit yanindé by the Inkhorn Others tend to be transparent, for examples, ayac yóly foot path ; charshanbe Wednesday ; su cuzy a Spring ; ylan balughy Snake-Fish. (s)i signals appear to be a nominal dependency in the two types of nominal structure. (3)(N1) + (N2). It is a combination of non-marked nouns. For example, baziryan hgemiler Merchant ships ; bibér chuval Pepper in sacks ; ing lz bairac English collours. N2 in this type of nominal structure does not have any nominal signals. For more cases, (3 ) (N1a) + (N2): boz dag the snowy Mountain ; kara kulak Black ears ; aksham namaz Prayer at sunset. In these cases, N1a is an adjectival noun. And more than that, (3 ) (N1) + (N2v): tach yatar the Stone lies ; mumsconderen an Extinguisher of the Light. In these cases, N2v is a verbal noun. Noticeably, from the third kind of nominal structures, an influence of basic sentential structure SOV can be seen. 298

312 WORDS IN ASIAN CULTURAL CONTEXTS Since 1700s, a shift of type 3 to type 2, and that of type 2 to type 1 have happened (Grönbech, 1940). The tendency that the final components play a role of nominal signal can be found in modern Turkish. Kornfilt (1997, P ) told us four basic types. A) Noun-noun structures: anne-mother anne-mother maternal grandmother ; B) Adjective-noun structures: koca-old kan-woman hag ; düz-smooth taban-sole flat footed ; C) Numeral-noun structures: kırk-forty ayak-foot centipede ; D) Verb-noun structures: unut-forget ma-neg ben-i i-acc forget-me-not. The last type of nominal structure is a V-O pattern that maintains some trace of early Turkish basic structure, but it has less and less cases in modern Turkish. In a normal sentence of modern Turkish, the accusative morpheme is remained at the end of predicate verb while its accusative object is put ahead of the verb. Kornfilt s (1997) description shows us some typical types of nominal structures, but lacks grammatical suffixes. Actually, modern Turkish have a lot of nominal suffixes applied in derivation and formed nominal compounds (Kreider, 1968), such as -lוk, וntו- and şו etc. (1) -lוk (-lık, -luk, -lük) can change concrete nouns into abstract ones. For example, hוrsוz thief hוrsוzlוk theft ; kan blood kansוzlוk anemia ; para money parasוzוk poverty. The same suffix is used to form nominal compounds derived from adjectives: ağוr heavy ağוrlוk weight ; cahil ignorant cahillik ignorance ; hast ill hastalוk illness. example, can connect a verb root to generate nouns from verbs. For וntו- (2) çוkmak to go out וntוkוç projecting part ; kוrmak to break וntוrוk crumb. The most widely used verbal nouns are, (1) light infinitive (regular infinitive, minus the final k), and (2) nouns formed by adding -ış, -üş or uş to verb stem. For examples: çarpışmak to collide çarpışma collision ; satmak to sell satוş sale ; dolmak to be filled dolma stuffed material ; gitmek to go, leave gidiş departure. Obviously, the category structure of Turkish nominal compounds can be described as a right-head pattern according to the position of their grammatical component. Nominal suffix is not found in Japanese nouns, nor required in its nominal structures. However, a verb or adjective should be changed into its nominal form, to be an object, subject or their central part. And in nominal compounds, the adjective and verb appear in a form different from a part of sentence. Nominal structures may have the following formations described by morphemic sense (Tsujimura, 1996), if phonological borrowing words are not accounted as compounds in Japanese. A) Native compound. It represents the basic patterns of Japanese compounds. N-N: aki-zora autumn sky ; V-V: tatsu stand, yomu read tati-yomi reading while standing ; V-N: nomu drink nomi-mizu drinking water ; A-N: tikai short tika-miti short cut ; N-V: doku melt yuki-doke snow-melting ; N-A: itai ache hara-ita stomachache B) Sino-Japanese compound. Its last component is commonly a nominal morpheme. ki-soku rule ; koo-ri high interest ; satu-zin manslaughter ; kei-koku warning C) Hybrid compounds. The compound has no grammatical suffix. Sino-Japanese + native: dai-dokoro kitchen ; Foreign + foreign: teeburu-manaa table manner D) (Dvandva) compound composed of a pair of roots with opposite meanings. 299

313 ASIALEX 2005 ten sky, ti earth ten-ti sky and earth ; yomu read, kaku write yomi-kaki read and write In the last compound, verbal components are changed in suffix to be a nominal one. E) Reduplication. It can be divided into a) mimetic and b) renyookei reduplications. a)hena-hena weak b)hanasi-o sii-sii tabe-ta (Talk-Acc do-do eat-past) We ate & talked concurrently. The reduplication pattern does not have a suffix to mark grammatical feature. F) Clipping compound. At least one component is a part of another word. seiyoo-siki Western-style yoo-siki; rimooto kontorooru remote control rimo kon Taken grammatical marks as a basic criterion, Japanese nominal structures appear in zero form, while Turkish ones in suffixes. It is noticeable that both Japanese verbs and Turkish verbs must be changed into nominal forms to be components prior to final components in nominal structures. Many Turkish nominal suffixes are formed quite late, and the developmental tendency of suffixes is taken place in SOV syntactic frame. Their position in nominal structures is coordinated to that of verbs at sentence level. Verbal structures Verbal structures in both Japanese and Turkish are abounded with suffixes. Turkish has personal suffixes attached to the end of verbal suffixes in sentences while Japanese has no such personal suffix. Other distinctions need to be described one by one. Gilson (1987) offered 3 types of old Turkish verbal construction with two elements. (1) V + temporal auxiliary *i-: dirsenghiz if you say (aorist + conditional); (ne) olurdy what would it be (aorist + DI-past); aldim idy I had bought (DI-past + DI-past). (2) V + descriptive auxiliary: aurydurmak to stand up ; aly komak to stop one ; buluvirmec to procure, salyvirmec to let go, yakivír light! (3) N + non-descriptive auxiliary: tahsil etmadim I could not get ; rayett idersin (you) will be favourable ; kabul ailemec to receive, makul olsun let be reasonable. Comparing (3) with (1) and (2), we know that their grammatical property is signaled by final auxiliary verb, for the element ahead can be with different grammatical properties. And Kornfilt (1997) divided verbal compounds of modern Turkish into three types. N-V: 1) noun + light verb do (et, eyle, yap) or be (ol): Banyo-bath yap-do (to) take a bath ; alay-mockery et-do to ridicule ; 2) Incorporated noun with verb: yemek-food ye-eat mek-inf. (to) eat (food) ; V-V: kap-snatch tı-past kaç-flec tı-past small, pirate bus ; A-V: çabuk-fast git-go ti-past went fast. The last type is morphologically simple adverbs in fact, though they cannot be moved away from the pre-verbal position. And most adverbs given in his grammar are used before verbs. Other modern cases of verb are from adjectives or nouns (Kreider, 1968): dar narrow daralmak to shrink ; hazוr ready hazוrlamak to prepare kוsa short kוsaltmak to shorten ; imza signature imzalamak to sign Japanese verbal suffixes have three basic types: terminal forms, forms for connecting nouns, and forms for connecting verbs. The first two look similarly with a final syllable containing -u, while the latter ends with an -i or -e. Examples for connecting verbs: a) Iru enter iri-ageru spend freely on ; Oku rise oki-agaru to get up, pick oneself up ; 300

314 WORDS IN ASIAN CULTURAL CONTEXTS b) Miru see mi-ageru look up at ; Kashiru lend kashi-sugiru to lend too much ; c) Katteru buy katte-ageru to buy for one ; Taberu eat tabe-akiru to be tired of eating The verbs ahead change final syllables into -i in a), or drop final syllables so as to end with -i in b) or -e in c) for connecting to the verbs behind. In verbal structures, it is the suffixes of verbs behind that can have inflections, which percolate grammatical properties to the whole compounds. It can be taken as a in both Japanese and Turkish. Moreover, Japanese has a specific kind of light verbs with a final forms of -(o)su ru (Miyamoto, 1999), which works as a verbal suffix attached to the end of nominal or adjectival root, but with weakly substantial and unclear lexical sense. Keibetsu suru to despite ; Gai suru to harm ; Seichoo shi-to (growth do-past) grew up big ; Ryokoo-suru (travel-do) to travel ; Sakushi-shi-ta (poem-writing-do-past) compose poem. The combination is loosened and can be separated into single words in some conditions: Keikoku warning Keikoku (warning) suru (do) keikoku o suru (give warning). Some scholars take hajimeru (begin), kokoromiru (try) and meijiru (order) as light verb as well, because they also show essentially the same properties as the light verb suru. (Matsumoto, 1996, P.64) But they have much weaker ability in combining other words than that of suru, and their lexical meanings are clear when applied as suffixes. Compared with the grammatical suffixes of Turkish verbal structures, Japanese verbal structures lack personal suffixes, perfect and participle suffixes (Lewis 2000). However, Japanese light verb -suru plays a similar role as that of Turkish light words. Adjective structures Adjectival suffixes are abounded in Turkish, such as -msı, -su, -li, -lı, -ane, -cil, -rek, -cek, -ceğiz, -cik, -(i)msi, -(i)mtrak, -si, -ce, -ci etc.(lewis, 2000) Its adjectives are divided into simple adjectives, derived adjectives and participial adjectives. Adjectival compounds concern the latter two and appear in the following kinds (J. Kornfilt, 1997). a) N-A: el-hand i-3.sg. açık-open open-handed ; Eş-same cins-sex el-deradj homosexual ; masa table masa-msı table-like ; çocuk child çocuk-su childish ; hız speed hız-lı rapid In the first case, nominal component is of the third single person, which not found in Japanese. Nominal components drop nominal suffixes in combination; only adjective suffixes are used to mark the grammatical properties of the compounds. b) A-A: som solid som-ut concrete ; büyük big, large büyü-cek biggish küçük small, little küçü-cük tiny, very little ; yakın near yakın-ca somewhat near In the combinations, prior adjectival component connects with the suffix behind without losing it self s suffix. The feature is different from that of Japanese adjectival structures. c) V-A: akmak (to) flow ak-ıcı fluent; fluid ; kıskanmak (to) envy kıskan-ç jealous ; unutmak (to) forget unut-kan forgetful ; çekin (to) withdraw çekin-gen retiring. In the cases, verbal suffixes are omitted, and instead, ended with adjectival suffixes, as -ıcı, -k, -ak, -ik, -u, -ç, -ek, -kan, -gen, -gun, gün etc. Some other adjectival compounds are formed of verb stems unbounded to some pattern in modern Turkish(Kreider, 1968). In regard to grammatical suffixes, Japanese adjectival structures are formed with adjectival suffix -i, -na or negative suffix -nai which contain an adjectival final syllable -i, or with variant forms ni, ku or Ø (null form) used for pause in sentence. Japanese has much less adjectival suffixes than Turkish. And its adjectival 301

315 ASIALEX 2005 structures composed of multi-roots are organized in the same word order as that of sentence. A formational difference between Turkish and Japanese is that, the prior adjectival components in Japanese have to change into certain forms for connecting adjectives, while the prior adjectival components in Turkish do not change their suffixes to do that. Conclusion As a general characteristic shared by the two languages, the three word classes have only suffixes express their grammatical properties. It is described as postpositional determination of compounding, coordinative with that at syntactic level. Noticeably, some suffixes of Turkish as well as the tendency of postpositional determination were formed quite late in history. And that is why I d rather considered it as a coordination of SOV typology, instead of originative one. The most important lexical difference of compounds between Japanese and Turkish happens between the prior components and post ones in compounding. It may be caused for some difference of historical syntactic structures in language development, if some compounding structures are accepted as a record of historical syntactic projections. So, I tend to take the difference as historical deviation, a specific feature of the languages. In order to learn the historical deviations, I hope there are Turkish and Japanese dictionaries that offer information about the origins of various suffixes, especially those used between components of compounds. The suffixes should be systematically studied and treated as single units in the dictionaries. And more explanations could be given as examples of compounding structures, or given in final bracketed notes entry by entry. References: Gilson, Erika Hitzigrath (1987): The Turkish Grammar of Thomas Vaughan: Ottoman-Turkish at the End of the 17th Century according to an English Transkriptionstext (Otto Harrassowitz: Wiesbaden). Grönbech, K. (1940): Der Akzent im Türkischen und Mongolischen, Zeitschrift der Deutschen Morgenlän- dischen Gesellschaft 94: Hony, H.C. ( ): A Turkish-English Dictionary, 2nd edition (Oxford at the Clarendon Press). Komfilt, Jaklin (1997): Turkish: Descriptive Grammar (Cornwall: TJ International Ltd. Padstow). Kreider, Herman (1968): Essentials of Modern Turkish (USA: AMERIKAN NEŞRİYATI BÜROSU). Lewis, Geoffrey (2000): Turkish Grammar, 2nd edition (England: Oxford University Press). Matsumoto, Yo (1996): Complex Predicates in Japanese (CSLI Publications). Miller, R. Andrew (1971): Japanese and the Other Altaic Languages (the University of Chicago Press). Miyamoto, Tadao (1999): The Light Verb Construction in Japanese (John Benjamins B.V.). Tao Yuanke (1999): Applicability of the RHHR: An Investigation from S-V-O Based Languages to an S-O-V Based language, Proceedings of Postgraduate Research Forum on Language and Linguistics 99 (Hong Kong: City University of Hong Kong). Tsujimura, Natsuko (1996): An Introduction to Japanese Linguistics (Blackwell Publishers Ltd.). Underhill, Robert (1986): Turkish, Dan Isaac Slobin and Karl Zimmer (ed.): Studies in Turkish Linguistics (John Benjamins publishing Company). 302

316 WORDS IN ASIAN CULTURAL CONTEXTS The first collocation dictionary for the Japanese language Keiko Terashima Nagoya University Minoru Moriguchi Hiroshima International University Abstract Nihongo Hyougen Katsuyou Jiten, the first collocation dictionary for Japanese language teachers and learners, was published by Kenkyusha in This paper criticizes NHKJ and makes a suggestion for future collocation dictionaries of Japanese, considering the following three points: (1) comparing NHKJ with its English equivalent, Oxford Collocations Dictionary for students of English, in respect of number of headwords and examples, main resources, macro-structures, and micro-structures, from a pedagogical viewpoint; (2) verifying the headwords and collocations in NHKJ with Japanese corpora whose texts have been extracted from newspapers and books; and (3) verifying the collocations in NHKJ with intuition of Japanese native speakers. Introduction The first collocation dictionary for Japanese language teachers and learners, Nihongo Hyougen Katsuyou Jiten (NHKJ), was published by Kenkyusha in Because upper level learners of Japanese needed collocational information and word frequencies as Moriguchi&Terashima (2003) indicates, the publication of this dictionary provides significant progress to Japanese pedagogy, though it comes a bit late compared with its English counterpart. This paper compares NHKJ with its English equivalent, Oxford Collocations Dictionary for Students of English (OCDSE), in respect to the way of selecting headwords, quantity of headwords and examples, main resources, macro-structures and micro-structures. Following the comparison, we show the results of corpora and experimental surveys pertaining to two Japanese words, and explore the extent to which NHKJ reflects the real use of these words by native Japanese speakers. The conclusion makes suggestions for the development of a collocation dictionary that could be used for any language. Overview of Nihongo Hyougen Katsuyou Jiten Most pages in NHKJ are assigned to the main body. Besides the preface, user s guide, and references, no extraneous discourse is included. The headwords were mainly extracted from the 10,000-word list for Level 1 (the highest level) in Japanese Language Proficiency Test: Test Content Specifications (Revised Edition) published by The Japan Foundation and Association of International Education in They include verbs and na-adjectives, the latter being of two types of adjectives in Japanese, and do not include nouns or i-adjectives. The micro-structure consists of four sections: collocates, examples, idioms and compound words. The first two sections appear obligatorily with few exceptions, while the latter two are optional. The collocate section is divided into subsections based mainly upon syntactic behaviour. Each subsection provides the collocated words and 303

317 ASIALEX 2005 colligated postpositions, when necessary, in addition to the headword. The collocate section of a verb entry contains only left collocates, and that of a na-adjective entries only right collocates 1. No information is provided on how to select collocates. Examples are written based on extraction from literary works or newspaper articles, not extracted from a corpus. Comparison with the Oxford Collocations Dictionary for students of English The comparison between NHKJ and OCDSE is shown below. This paper does not address information available on either dictionary s CD-ROM version. Table 1: Comparison between NHKJ and OCDSE NHKJ OCDSE Number of head words 1544 approx Parts of speech verbs and na-adjectives nouns, verbs and adjectives Number of examples over 25,000 over 50,000 Main resource Japanese Language Proficiency British National Corpus Test: Test Content Specifications (Revised Edition) Macro-structure no middle matter no back matter 13-page front matter, 16-page middle matter, and 5-page back matter Collocate verb: left collocates adjective: right collocates both left and right collocates Examples written referring to literary extracted from BNC works or newspaper articles Idioms referred not referred Compound words referred not referred Compared with OCDSE, NHKJ seems less comprehensive, given the small number of headwords and examples, the thinner macro-structure, fewer parts of speech covered, and one-directional aspect of collocates. Its main resource, Vocabulary List for Level 1 in Japanese Language Proficiency Test: Test Content Specifications (Revised Edition), which contains 10,000 words, also seems weak compared with BNC. The strong points of NHKJ are that its resource is specified for Japanese pedagogy whereas BNC has a broader purpose, and that NHKJ addresses idiomatic expressions and compound words. Verifying NHKJ headwords with corpora To verify how much NHKJ reflects real use of words, we explored word frequencies with corpora to ascertain whether or not high frequent words in corpora were included among its headwords. The corpora we used are as follows: - CASTEL/J text database including 48 books, mainly explanatory or educational - Chunichi Shimbun 2001 database 1 Japanese is an SOV language. In Japanese syntax, an adjective is on the left of a noun, and postposition (rather than prepositions) are used. 304

318 WORDS IN ASIAN CULTURAL CONTEXTS - Chunichi Shimbun 2002 database - Chunichi Shimbun 2003 database - Mainichi Newspapers 99 database We used chasen (Windows edition) ( for morphological analysis and Chakoshi ( to extract collocations. The analysis revealed that the following five verbs and five na-adjectives appear frequently in the five corpora. verbs: suru (do) 1, aru (be), dekiru (can do), omou (think), miru (see) na-adjectives: hitsuyou (necessary), kanou (possible), shizen (natural), jiyuu (free), juuyou (important) The ten words above are all included among headwords of NHKJ. Meanwhile, NHKJ does not include low-frequency words even if they are in the above-mentioned word list. It should be noted, on the other hand, that OCDSE does not include highly frequent verbs such as make or do, for the following reasons: Some very common words such as the verbs make and do do not merit entries of their own. This is because these verbs have no real collocations of their own. They themselves are the collocations of lots of nouns, and appear in the entries for those nouns. (p. x, ll.2-6) Verifying NHKJ collocations with corpora To verify the collocations in NHKJ, we explored the collocates of two words and checked whether high frequent collocates were included among the collocates in NHKJ. We used the same corpora as listed above. We explored collocates of the verb omou (think), and the na-adjective hitsuyou (necessary), both of which are among the most high frequent words in the above corpora as shown. The verb omou most frequently behaves in the eight syntactic forms as below: omou (dictionary form); omotte (continuative form); omotta (perfect form); omoimasu (polite form); omowareru (passive or spontaneous form); omoenai (potential negative form); omoimashita (polite perfect form); omowanai (negative form) The collocate section of omou refers only to the dictionary form while the example section provides six forms excepting omoimasu and omoimashita above. 1 Translation words are provided as general examples. Equivalent expressions may change according to the context. 305

319 ASIALEX 2005 The most frequent left collocate to omou is the postposition to, and the second most is ni. To is equivalent to that as conjunction, and ni is an inflectional form of na-adjectives. The frequent left collocations of to omou include -ta to omou (think about a past event; -ta marking past tense), -ka to omou (wonder; -ka forms an interrogative clause), aru to omou (think something exists; aru means exist ), -tai to omou (want to; -tai expresses a desire), -you to omou (be going to; -you expresses an intention), ii to omou (think something is good; ii means good ) and naru to omou (think something becomes something; naru means become ). NHKJ refers to only ii to omou. The most frequent left collocation of ni omou in the newspaper corpora is -you ni omou. You ni is an adverbial form of na-adjective you, which expresses likelihood. Other frequent left collocates include hokori (proud), fushin (doubtful), gimon (questionable), fushigi (mysterious), zannen (sorry), and ikan (regretful) in this order. In the CASTEL/J corpus, the most frequent collocation of ni omou is also -you ni omou, and the second most collocation is fushigi ni omou. Although NHKJ refers to hokori, gimon, and fushigi, its order doest not reflect the frequency order in corpora. In NHKJ, the collocate section of omou provides four subsections: to omou; (i-adjective +) omou; ni omou; and wo omou ( wo is a postposition usually the object). The two most frequent collocates among i-adjectives +omou are ureshiku (gladly) and tsuyoku (strongly) in the newspaper corpora, but NHKJ only refers to ureshiku and provides no collocate refererence to degree such as tsuyoku. The collocate in the last subsection, wo omou, does not occur frequently in corpora. However, this information is significant from a pedagogical viewpoint. The na-adjective hitsuyou (necessary) most frequently takes five syntactic forms as below: hitsuyou-ga [hitsuyou-ha] (used as subject); hitsuyou-da (non-past infinitive); hitsuyou-na (pronominal); hitsuyou-to (with a postposition to ); hitsuyou-ni (adverbial) NHKJ provides four subsections of right collocates to hitsuyou : hitsuyou-na; hitsuyou-wo; hitsuyou-ga; and, hitsuyou-ni. It does not provide subsections for hitsuyou-da or hitsuyou-to, each of which appears once in the example section. HNKJ refers only to right collocates of an adjective and does not have a subsection for hitsuyou-da, which usually occurs at the right end of a clause. However, corpus survey reveals that hituyou-da has a tendency to appear in constructions such as hitsuyou-da to omou (I think something is necessary). Almost 90 percent of the occurrence of hitsuyou-ga is followed by aru (be) or nai (not be). Although NHKJ contains three collocates shoujiru (occur), aru and nai, its order doest not reflect the frequency order in corpora. In any of the above five corpora, hitsuyou-na frequently occurs with no, koto, and mono, each of which forms a noun phrase with an adjective. NHKJ does not refer to these three words in the subsection of hitsuyou-na. Although NHKJ provides twelve words as right collocates to hitsuyou-na, only three collocations are highly frequent in the corpora. 306

320 WORDS IN ASIAN CULTURAL CONTEXTS Frequent collocates of hitsuyou-ni in any corpora above include naru (become), oujiru (respond) and semaru (urge), but NHKJ does not refer to naru. The collocate semaru is usually used as passive semarareru in corpora, and NHKJ provides the passive form both in the collocate and example sections. Meanwhile, the collocate oujiru is usually used in the continuative form oujite in corpora, but NHKJ does not provide the collocation hitsuyou ni oujite (as occasion demands). The compound word section provides hituyou-hin (necessary goods), but this word does not occur very frequently in corpora. Verifying NHKJ collocations with intuition of native speakers To support the result of the corpus survey, we conducted a free-writing experiment. The subjects, native speakers of Japanese in their 20s to 70s, were given a word and asked to make a sentence or sentences including that word. The result of the experiments does not offer additional information to that of the corpus survey; rather, some details are significant from a lexicographical viewpoint as shown below. The words we experimented with were omou, hitsuyou, hakaru (intend), and dokuji (original). Given limited space, below we will discuss only the results of omou and hitsuyou. Concerning omou, we received 48 sentences from 28 subjects. NHKJ covers all of the forms of omou that we received from this experiment and also covers the most frequent collocate, to omou. However, it does not refer to an idiomatic expression omowazu (unintentionally), which is also seen frequently in corpora, or several compound words such as omoidasu (remember), omoitsuzukeru (continue to love) or omoidoorini (as one likes). Concerning hitsuyou, we obtained 53 sentences from 29 subjects. As in the corpus survey, the non-past infinitive form hitsuyou-da was most frequently obtined from our experiment. Compound words found in this experiment are hitsuyou-fukaketsu (indispensable), hitsuyou-saiteigen (minimally required), hitsuyou-juubun (necessary and sufficient), and fu-hitsuyou (unncessary). NHKJ refers to all of these compound words except hitsuyou-juubun, but it refers to hitsuyou-jouken (necessary condition) instead. Towards future collocation dictionaries The above-mentioned surveys reveal that NHKJ does not faithfully reflect collocational information in real use of corpora or native speakers intuition though it has several strong points from a pedagogical viewpoint of Japanese. In addition to this problem, NHKJ has several points to be improved. In the first place, Japanese learners need information on colligations as well as collocations. Although NHKJ provides colligation information in examples, it is hard for learners to find colligations only from examples. Secondly, it should provide information about onomatopoeia, which is a distinctive feature of the Japanese language. Furthermore, some points Komuro (2004) indicates for improvements to OCDSE are also applicable to NHKJ. For example, NHKJ should provide information on cross-references, incorrect usages or synonymous expressions. Although this paper refers mainly to NHKJ, we believe our method of study and comparative analyses would be beneficial to the development of collocation dictionaries of any language. 307

321 ASIALEX 2005 The verb of breaking and its semantic variations in context: a contrastive study of Thai and Japanese Kingkarn Thepkanjana Chulalongkorn University Kingkarn.T@chula.ac.th Satoshi Uehara Tohoku University uehara@intcul.tohoku.ac.jp Abstract The fact that a lexical item has semantic variations when combined with other linguistic elements is a central issue in lexical semantics. A number of researchers claim that a lexical item has one basic meaning, and that other extended meanings are triggered in context by a process whereby the semantic structure of the lexical item is adjusted in certain details so that it is semantically compatible with its neighboring lexical items. This paper aims to examine how this process actually works as it applies to a transitive verb occurring with subject and object arguments. A study of the Thai transitive verb ha k break and its corresponding verb oru in Japanese is presented. Arguably, all seemingly discrete meanings of ha k are interrelated and so are those of oru. The basic meaning of each verb corresponds to the most concrete event and is the most cognitively salient. It consists of a number of facets, which represent different physical resulting states of an entity undergoing an action denoted by either ha k or oru. Two mechanisms are found to derive the extended meanings. First, only some facets of ha k and oru are promoted. Second, ha k and oru are figuratively interpreted. The other objective of this study is to show semantic differences between ha k and oru. It is demonstrated in this paper that so-called corresponding words in different languages, especially verbs, hardly have exactly the same meaning. Introduction The fact that a word form is associated with more than one meaning is one of the central issues in lexical semantics. This phenomenon raises a number of theoretical questions, for example, whether a lexical item in question constitutes a case of polysemy or homonymy; in case of polysemy, whether there is a core semantic element unifying discrete meanings of the word form in question. In accounting for semantic variations of a lexical item in context, it is often claimed that a lexical item has one basic meaning, and that other extended meanings are triggered in context by means of the processes referred to as co-composition by Pustejovsky (1995) and as accommodation by Langacker (1987). This paper aims to examine how this process actually works as it applies to a transitive verb occurring with subject and object arguments. A corpus-based semantic investigation of the Thai and Japanese transitive 308

322 WORDS IN ASIAN CULTURAL CONTEXTS verbs which correspond to break, namely, ha k and oru respectively, is presented as a case study. Ambiguity, polysemy and vagueness The phenomenon in which a word form has more than one meaning is often associated with the task of categorizing the word form in question. This task involves such notions as ambiguity, polysemy, and vagueness. In lexical semantics, the term polysemy is usually defined as opposed to ambiguity or homonymy on the one hand and as opposed to vagueness on the other. Ambiguity is traditionally defined as a case in which two or more distinct meanings are associated with a phonological form. Ambiguity gives rise to distinct lexemes. An example of ambiguity is bank, which means financial institution and land adjoining a body of water. Vagueness refers to a case in which non-distinct meanings are associated with a phonological form. An example of vagueness is aunt, which can refer to a father s sister and a mother s sister. Lexical ambiguity or homonymy involves two lexemes each of which has a distinct meaning. Polysemy involves a single lexeme with distinct meanings. Vagueness involves a lexeme with a single but non-specific and non-distinguished meaning. Thus, polysemy is apparently located halfway between ambiguity and vagueness. Several cognitive linguists, such as Geeraerts (1993), Langacker (1987), Taylor (1995) and Tuggy (1993), argue that the distinction between these cases is not clear-cut. Rather, they are regarded as unfixed points located on a continuum. This paper does not aim to determine whether the association of the transitive verbs under investigation constitutes a case of polysemy, ambiguity or vagueness. Rather, it aims at analyzing how the extended meanings of the verbs emerge in context. Meanings of ha k in Thai Based on the corpus citations of ha k break, it is found that this verb expresses a diversity of meanings. It is noted that the meanings listed below result from a preliminary semantic analysis and are thus tentative. In giving examples of ha k in combination with different direct object arguments, we will not gloss the word ha k to prevent confusion arising from the English gloss. (1) to disjoin or reduce a hard, straight and long entity to pieces by hand with sudden or violent force as in ha k ki ma y, glossed as ha k branch. (2) to deduct (money, expenses) as in ha k kha acha yca ay, glossed as ha k expenses. (3) to harvest, to collect, to pick up (ears of corn) as in ha k kha awpho ot, glossed as ha k corn (4) to take away (marks) in an examination as in ha k khanææn, glossed as ha k marks. (5) to turn away the steering wheel as in ha k phua malay, glossed as ha k steering wheel. (6) to snap fingers as in ha k ni w, glossed as ha k finger. 309

323 ASIALEX 2005 (7) to cut price as in ha k rakhaa, glossed as ha k price. (8) to conquer a town as in ha k m a, glossed as ha k town. (9) to turn (light, sunray) to a different direction; to refract as in ha k sæ æ, glossed as ha k light. (10) to disparage; to slight; to belittle someone as in ha k li am, glossed as ha k angle. (11) to betray someone; to doublecross someone as in ha k la, glossed as ha k back. (12) to discontinue a romantic relationship with someone as in ha k sawa at ha k love. (13) to force oneself as in ha k cay, glossed as ha k heart. (14) to force someone as in ha k kh, glossed as ha k neck. (15) to embarrass someone as in ha k na a, glossed as ha k face. The first meaning of ha k is obviously the most concrete of all the fifteen meanings. It is noted that ha k in Thai does not semantically completely correspond with break in English in that the former is specific with regard to the manner of disjoining an entity and with regard to the physical properties of an affected entity whereas the latter is not. It is apparent that a concrete event of disjoining an entity brings about certain consequences. That is, the broken entity becomes deformed, destroyed, dysfunctional, and smaller in size. In addition, a path formed by a hard, straight and long entity becomes deviated. All of these four consequences which are referred to as semantic facets are incorporated in the basic meaning of ha k. For the basic meaning of ha k to obtain, ha k needs an appropriate syntagmatic context. That is, it must appear with a direct object argument expressing an entity which is concrete, hard, has a potential to be useful and can be perceived to form a path. The other meanings of the verb are arguably extended from the basic meaning in some way by means of accommodation in Langacker s terms or co-composition in Pustejovsky s terms. By this process, the semantic structure of a lexical item, which is in this case ha k and oru, is adjusted in certain details to make it semantically compatible with its syntagmatic context which is specifically the meaning of its direct object argument. The fourteen extended meanings of ha k can be classified into three types as follows. 1.The first type of extended meaning The first type of extended meaning is expressed by ha k occurring with a direct object argument expressing a non-physical entity. This type of extended meaning includes the meanings in (2), (4), (7), (8), (9), (12), and (13). In this case, the meanings of ha k revolve around only one of the four semantic facets because the nonphysical direct object of ha k promotes a semantic facet of ha k whereas demotes some others. For example, the direct object arguments indicating money or expenses, marks in an 310

324 WORDS IN ASIAN CULTURAL CONTEXTS examination, price, as in the meanings in (2), (4) and (7) respectively promote the semantic trait that the size of a broken entity becomes smaller. On the other hand, the direct object argument indicating the light as in the meaning in (9) promotes the semantic trait that its path becomes deviated. 2.The second type of extended meaning The second type of extended meaning includes the meanings in (10), (11), (14), and (15). It is noted that the entities indicated by the direct object arguments of ha k with this type of extended meaning are NOT nonprototypical entities for a breaking action because they are concrete, hard and have a potential to form a path. However, the literal meanings of these expressions are pragmatically odd because the entities denoted by the direct object arguments are not the things which we normally break. The literal meanings of the verb phrases must be interpreted idiomatically. Notice that the processes of accommodation and co-composition do not play as crucial a role as they do in obtaining the extended meanings of the first type. 3.The third type of extended meaning The third type of extended meaning includes those in (3), (5), and (6). These meanings are derived by another kind of semantic extension, namely, inference. It is noted that the literal meanings of the verb phrases are pragmatically possible although they do not occur frequently in discourse. However, the context of situation is the factor which suggests that they must not be interpreted literally. Rather, inferencing must be performed on the literal meanings of these verb phrases. The intended meanings of these phrases are based on either all of the four semantic facets of the physical action of breaking such as in the third meaning of to harvest corn, or are based on only one of the four facets of the physical action of breaking such as in the remaining two meanings. The real-world knowledge and the context of situation play a crucial role in arriving at the intended interpretation. The process of accommodation or co-composition must apply so that the intended meanings of ha k will obtain. Meanings of oru in Japanese In carrying out a linguistic study of a non-english language, it is customary to use English as the metalanguage in expressing the meanings of the non-english data. However, it should be borne in mind that English glosses and translations do not express the exact meanings of the language under investigation as pointed out in the case of ha k in Thai and break in English. The present study is even more complicated because two non-english words which are supposedly corresponding ones are being examined. Again, English is used as the gloss for these two non-english words. If we try to find a Japanese word which corresponds to break in English, which is used as the gloss for ha k in Thai, it is found that there are two candidates which might match break in English, i.e. kowasu and oru. For many people, the word kowasu might come to mind first. However, if we compare ha k and kowasu on the one hand and ha k and oru on the other, it turns out that ha k is more similar to oru than to kowasu. Kowasu is defined as change the shape of something and make it useless or dysfunctional by using force. Oru is defined as apply force to a straight line or a flat object at a point or a line and make two (or more) lines or flat objects. Therefore, oru is chosen as the corresponding word of ha k even though there are still some differences in meaning. However, the same 311

325 ASIALEX 2005 gloss, i.e. break, is used for oru as for ha k for convenient purposes. It should be kept in mind that there are differences in meaning between ha k, oru and break. The following list of meanings of oru results from a preliminary semantic investigation of this word. ( ACC is used for accusative marker in the glosses below.) (1) to apply force to a straight line or a flat object at a point or a line and make two (or more) lines or flat objects as in siitu o hutatu ni oru, glossed as sheet ACC two in oru meaning fold sheets in two. (2) to count (by bending one s fingers) as in yubi o oru, glossed as finger ACC oru (3) to sit down/come to a halt as in hiza o oru, glossed as knee ACC oru (4) to surrender (by bending oneself)/interrupt as in kosi o oru, glossed as waist ACC oru (5) to pluck, break (tree branch), pick up (flower) as in hana o (ta-) oru, glossed as flower ACC (hand-) oru. (6) to destroy, make something dysfunctional as in matti no ziku o oru, glossed as matchstick ACC oru (7) to end one s literary career as in hude o oru, glossed as writing brush ACC oru (8) to make efforts as in hone o oru, glossed as bone ACC oru. (9) to return a phone call as in ori-kaesi denwa-suru, glossed as oru-return phone call (10) to give in as in ga o oru, glossed as self ACC oru (11) to create (folded paper) as in turu o oru, glossed as paper crane ACC oru. The first meaning is the most basic one of all in that it corresponds to the most concrete event and it is the first that comes to mind in a neutral context. The concrete event of folding or applying force to a straight line or a flat object to make two or more lines or flat objects gives rise to many consequences as follows. The affected object may become destroyed and dysfunctional. In addition, the path formed by the affected object will become deviated. These consequences are likely to take place. There may be other consequences that probably take place, i.e., the affected entity may become split into two or more pieces and a new entity may result from an action of folding. The ten extended meanings of oru can be classified into five types as follows. 1. The first type of extended meaning This type of extended meaning, i.e. the meaning in (6), is expressed by oru occurring with an entity with a given FUNCTION as the direct object argument. This extended meaning is based on the semantic facet of oru that an affected entity is destroyed or dysfunctional as a result of the concrete event expressed by oru. (This type is similar to the meaning of ha k in (8) and (12) in Thai.) 2. The second type of extended meaning This type of extended meaning, which is the meaning in (5), i.e. to pluck, break or pick up, is expressed by oru occurring as the direct object argument expressing a PART of an affected entity. This extended meaning of oru is based on the semantic facet of oru that that part is separated from the whole. (This type is similar to the literal meaning of ha k in (2) and (4) in Thai.) 312

326 WORDS IN ASIAN CULTURAL CONTEXTS 3. The third type of extended meaning This type of extended meaning, which is the meaning in (9), i.e., to return a phone call, is expressed by the compound verb ori-kaesu occurring with the (implied) direct object argument expressing the image-schematic PATH of a phone call. This extended meaning revolves around the semantic facet of oru that the direction of the path formed by an affected entity is deviated. (This is similar to the meaning of ha k in (5) and (9) in Thai.) 4. The fourth type of extended meaning This type of extended meaning, which is the meaning in (11), i.e., to create (folded paper), is expressed by oru occurring with the direct object argument that expresses an entity RESULTING from folding paper. This meaning is based on the semantic facet of oru that an entity made of folded paper may be created from the concrete event expressed by oru. 5. The fifth type of extended meaning This type of extended meaning, which includes the meanings in (2), (3), (4), (7), (8), and (10), is derived by INFERENCING on the literal meanings of oru in combination with its direct object arguments. (The literal meanings of (2), (3) and (4) are of the basic type of meaning, while those of (7) and (8) are categorized as the first type of extended meaning.) It requires some real world knowledge to obtain the intended interpretations. For example, an inference enables us to interpret yubi o oru, glossed as finger ACC oru as to count because counting can be done by bending fingers. Another example is kosi o oru, glossed as waist ACC oru, whose intended meaning of this whole phrase is to surrender. This meaning is motivated by the fact that surrendering is often shown by bending oneself down. Semantic variations of a verb as a case of polysemy The meaning of a verb can in principle vary infinitely as it is combined with a different noun argument. Some researchers such as Lakoff (1987) and MacWhinney (1989) consider that all instances of semantic variations are a case of polysemy even though the differences in meaning are so close. MacWhinney discusses the mechanism which gives rise to polysemy. In sentences, some words are in construction with others and these constructions force words to be polysemous. However, not every word in a sentence can impinge on every other word. In order for one word to push another word around, the two words have to be involved in a meaningful relation. This type of polysemy is called pushy polysemy. Polysemic pushing occurs only across what MacWhinney calls valence bridges. For example, in the Thai sentence th ha k khaflacha yca ay rµ µyan Have you deducted expenses?, a valence bridge exists between ha k and khaflacha yca ay. The word khaflacha yca ay expenses pushes or impinges on the verb ha k so that the latter will take on the meaning of deducting. A valence bridge will not be formed unless the verb ha k assumes this reading. In this example, we can say that the noun khaflacha yca ay pushes the verb ha k into a particular polysemic pathway. It is common for nouns which function as the direct object to push the verbs around as seen in the examples so far. This explains why verbs tend to be polysemous 313

327 ASIALEX 2005 than nouns that they are in construction with. However, it is sometimes possible for the verbs to push their noun arguments around. For example, in order for the phrase ha k kra da at break paper to make sense, the noun kra da at must be interpreted as having properties of a prototypically breakable object. Namely, it must be hard enough to allow itself to be broken. If the noun did not assume this semantic property, the valence bridge between the verb and the noun would not be formed and this construction will be semantically odd. It is noted that the case of a verb pushing a noun around does not occur frequently. References Geeraerts, Dirk (1993), Vagueness puzzles, polysemy s vagaries, Cognitive Linguistics. 4 (3): Lakoff, George (1987), Women, fire, and dangerous things. (Chicago: University of Chicago Press). Langacker, Ronald W. (1987), Foundations of cognitive grammar. Volume I. Theoretical prerequisite. (Stanford: Stanford University Press). Pustejovky, James (1995), The Generative Lexicon. (Cambridge: MIT Press). Taylor, John R. (1995), Linguistic categorization. Prototypes in linguistic theory. Second edition (New York: Oxford University Press). Tuggy, David (1993), Ambiguity, polysemy and vagueness, Cognitive Linguistics. 4 (3): MacWhinney, Brian (1989), Competition and lexical categorization, in Roberta Corrgian, Fred Eckman and Michael Noonan (eds), Linguistic categorization (Amsterdam: John Benjamins), pp Abstract Nation-states as a semantic domain Eric C. Thompson National University of Singapore socect@nus.edu.sg This paper addresses two interrelated themes: consideration of nation-states as a semantic domain (in the classic sense of that concept within cognitive anthropology) and how a study of the semantic domain of nation-states can address questions of cross-cultural comparison and difference within semantic domain analysis. With regard to the first issue, the paper outlines the role of the nation-states as a semantic domain in the broader context of the evolution of a globally recognized sense of nationalism from the 18th through 20th centuries. With regard to the second issue, the paper draws on data collected from 932 respondents in five Southeast Asian countries to examine the similarities and differences in their cognitive maps of this domain. Research on other domains (e.g. emotions, kinship terms) has led scholars to argue that the structure of such domains is widely shared and that presumed cultural differences are minimal. The paper argues that differences in the cognitive map of the nation-state domain is more pronounced and provides suggestions as to why this is the case. 314

328 WORDS IN ASIAN CULTURAL CONTEXTS The potential of learner corpora for pedagogical lexicography Yukio Tono Meikai University Abstract There is a growing awareness that a more data-oriented approach can greatly improve the design and content of pedagogical dictionaries. In this paper, I argue that three types of data, general language corpora, research data on dictionary use, and a corpus of learner language, can significantly improve learners dictionaries. I will review my past research in these areas briefly and focus particularly on how learner corpora can be used to better improve the kinds of information provided to learners in pedagogical dictionaries. Research paradigm shift in pedagogical lexicography Lexicography covers a broad range of interdisciplinary areas in linguistics, reference sciences, language teaching and learning, and more recently natural language processing. A common approach in lexicography is to try to apply knowledge and facts found in these disparate areas to the production of dictionaries. These facts are mainly concerned with linguistic observations or theoretical analyses of the system of a language as well as its use. Linguists study languages and inform lexicographers on better or more innovative ways of describing a word, providing usage information, giving illustrative examples and so on. In the last two decades, however, this traditional approach to dictionary-making has been taken over by a more data-oriented approach to lexicography, which is based upon empirical research on language corpora, dictionary users and language learners. This is particularly true in the field of pedagogical lexicography. In this paper, I would like to briefly summarize recent developments in pedagogical lexicography with reference to the three areas mentioned above (language corpora, dictionary users and language learners) and argue that significant improvements in user-friendliness will be achieved by using second language (L2 henceforth) learner corpora to inform the making of learners dictionaries. A data-oriented approach to pedagogical lexicography In a sense, lexicography has always been data oriented. Lexicographers investigate the use of words and phrases by what lexicographers at Merriam-Webster s call reading and marking. The way lexicographers access language use data, however, has drastically changed since computerised corpora became available in the early 1960s. The first fully corpus-based monolingual dictionary was the COBUILD English Dictionary (1987), which was radically different from existing monolingual learners dictionaries and enthusiastically welcomed among linguists and language educators in Japan as well as in the rest of the world. All the other major monolingual learner s dictionaries have more or less followed this trend and lexicographers have been using KWIC (keyword in context) concordances as their primary tool for finding out how a word behaves ever since. After Church and Hanks (1989) introduced the notion of Mutual Information (a measure of the salience of the association between any two 315

329 ASIALEX 2005 words), lexicographers became more interested in identifying statistically salient collocates. Every publisher started having its own corpus (e.g. the Cambridge International Corpus, the Longman Corpus Network) and even its own corpus query system (a set of tools to help lexicographers use corpus data effectively). By 1995, the revised Big 4 (COBUILD, OALD, LDOCE, and CIDE) were all claiming that they were corpus-based. Since then, using corpora for dictionary-making has become standard practice, at least for the major dictionary publishers in the UK. Recently, a more sophisticated approach to using corpora has been proposed, mainly to deal with larger sets of data or more detailed grammatical relations in the text (e.g. the Sketch Engine (Kilgarriff, et al. 2004), the Shogakukan Language Toolbox (Nakamura and Tono 2003) among others). Recent papers presented at Euralex 2004 show that dictionary publishers are shifting their attention from using general mega-corpora to more specialised corpora and from the ordinary use of corpora to more purpose-specific uses. Another important area which underwent a marked shift in the past two decades is research into dictionary use. Up until the 1980s, very little attention was paid to the needs and skills of dictionary users. Reinhart Hartmann was one of the first to enlighten us on the importance of research into dictionary use (Hartmann 1979; 1983). I myself was one of the few researchers who started to conduct experimental studies on dictionary reference skills in the 1980s (Tono 1984, 1986, 1988). There is now a growing body of literature in this field and major works are reviewed in my book in Lexicographica series (Tono 2001). Whilst lexicographers are aware of the importance of user studies, it takes time to apply research findings to the actual production of dictionaries. Some notable successful applications of research findings to dictionary-making include the provision of a menu at the beginning of dictionary entries. With the menu, users can first browse through the various meanings of any given word, which is a feature widely introduced after a study confirmed the fact that the users only look at the beginning of dictionary entries (Tono 1984). Dictionary makers are also taking note of research in dictionary use where the effectiveness of newly introduced organizational devices are put to the test. For example, I conducted an experiment to investigate the effects of Signposts in LDOCE and Guidewords in CIDE and found that the terms used for signposts in LDOCE were more effective in directing users to the right meanings while the terms used for Guidewords were often too abstract to signpost meanings in the dictionary (Tono 1997). More recently, there has been a renewed interest in the role of dictionaries in language learning since electronic dictionaries began growing in popularity in Japan several years ago. The market is constantly increasing in size, and major manufacturers such as Casio, Seiko, Sharp, Canon and Sony all spend time and money on the development of new types of hand-held e-dictionaries. As the number of university and high school students who own pocket e-dictionaries grows rapidly, more research has been conducted on the effects of using pocket e-dictionaries in reading and writing. The dictionary workshop organized by the JACET Lexicography SIG this March was very well attended, where approximately sixty paper presentations were given. The time is ripe for further research on the effects of using pocket e-dictionaries in terms of the medium or interface (paper vs. electronic), L2 vocabulary learning, and dictionary skills training. 316

330 WORDS IN ASIAN CULTURAL CONTEXTS The above two factors, the advent of language corpora and research into dictionary use, have contributed greatly to the improvement of learners dictionaries. A third area which I will now focus on is the study of language learners themselves. Dictionaries serve many different purposes. Pedagogical lexicography is mainly concerned with dictionaries designed to help foreign learners of the language. Language learners as dictionary users need to be investigated more seriously. Pedagogical lexicography should take into account L2 learners learning habits, learning styles, learning strategies and learning processes. There is a large body of research in the field of foreign language learning and second language acquisition (SLA henceforth), but unfortunately very little effort has been made to apply SLA research findings to the study of dictionary-making and on how language learners actually use dictionaries. For the past twenty years, I have been conducting research in all the above three areas. At the beginning of my research career, my primary interest was in the role of dictionary use in language learning. When the COBUILD English Dictionary was first published, I realised the potential of corpus-based research, and this led me to pursue my doctoral research at Lancaster in the 1990s. There I learned about the various branches of corpus-based research and saw examples of corpus applications in different fields. I became convinced that the use of corpora would make a major difference in the field of English language teaching in Japan. Before going to Lancaster, I had collected English essays written by Japanese learners of English as part of a large research project investigating the effects of teacher feedback in L2 writing. I started turning this valuable data into a corpus so that I could more systematically investigate the characteristic features of the writing of Japanese learners of English. Learner corpora and L2 lexicography A learner corpus is a collection of speech or writing by foreign language learners. By looking at the learner performance data, we can find many interesting patterns of use which are quite different from those of native speakers. In many cases, these differences are due to the fact that learners are still in the process of acquiring a language, and they naturally make errors or mistakes. Studying learner errors is not new. The research area called Error Analysis has been around for more than 30 years. What is new is that we can now employ the techniques of corpus linguistics to investigate learner language in a more empirical, data-based fashion. What sorts of information can we extract from learner corpora? How can we apply such findings to pedagogical dictionary-making? Let me describe some of these areas in detail. (a) Source of vocabulary selection for L2 learners One way of analysing learner corpora is to create a wordlist from learner writing and compare it with a comparable wordlist derived from the writing of native speakers. I have been working for NHK (Nihon Hoso Kyokai, Japan Broadcasting Center) in developing a television English conversation program (titled Hyakugo de sutato eikaiwa which means Let s start with 100 basic words in English ). This program is unique in the sense that it is probably the first corpus-based English conversation program on TV. It consists of a hundred lessons based on 100 key vocabulary items which were chosen based on corpus analysis. It is a well-known fact that the high-frequency lexical items in English (or any language) will cover a very high proportion of the words in any text; the most frequent 100 words (lemmas) in English, for example, will cover approximately 70% of words in a spoken corpus (e.g., the 317

331 ASIALEX 2005 spoken component of the British National Corpus). Many of these are core lexical items (verbs, prepositions, personal and wh-pronouns, determiners, adverbs and conjunctions) that play a crucial role in constructing basic English structures (See Lee 2001 for more discussion on core vocabulary). There are relatively few nouns (only six!) and adjectives in the top 100 words. In this TV program, I focus on the most frequent 100 keywords and design the program around a lexical syllabus. As I worked on this program, I became convinced that beginning-level students should study a set of basic core vocabulary again and again in a series of different language tasks. These core vocabulary items are at the heart of English grammar and are rich in meanings and functions, and it takes time to acquire a satisfactory productive and receptive grasp of them. One hundred words might seem too few in number and some people claim that to be functional in English one should know at least the top 2000 words, which would typically cover about 90% of the words in a spoken corpus. Leftover words (i.e. those below the 2000 word level) are said to be mostly those which are affected by particular topics or situations and which can therefore be learned independently from the first 2000 basic items. However, how exactly can we determine the next set of words to learn (after the first 2000)? In an EFL environment like Japan, most L2 input will come from the classroom especially for beginning-level learners. The language spoken and written in the classroom is different from that of everyday conversations encountered by native speakers. It is natural, therefore, that the vocabulary covered in classroom settings will be different from those used in everyday life in Britain, and EFL learners dictionaries should meet the specific communicative needs of L2 learners in terms of vocabulary selection. For such purposes, learner corpora collected from particular L2 learner groups would be most useful. By comparing well-balanced learner corpora with native-speaker corpora, both in spoken and written modes, we can possibly identify a list of words which are significantly more frequently used by L2 learners. These are the candidate words that learners want to express in English. In this way, we could exploit learner corpora to improve the selection of vocabulary for more user-friendly bilingual dictionaries. (b) Identifying L2 learners common errors Recently, monolingual dictionaries such as LDOCE, CALD, and Longman Essential Activator all feature common learner errors as part of the usage information. The primary aim of this information is to give learners information on correct usage based on common errors as shown in the learner corpus data collected by the dictionary publishers. The types of errors highlighted in learners dictionaries may be classified as follows: (a) Lexical choice e.g. Do not say injure someone s health. Say damage someone s health. (LDOCE) e.g. The words not... either are used to add another piece of negative information. Helen didn t enjoy it either. Helen didn t enjoy it too. (CALD) 318

332 WORDS IN ASIAN CULTURAL CONTEXTS (b) Verb forms e.g. You can have problems doing something. Do not use to do (LDOCE) (c) Verb patterns e.g. You propose something to someone: He proposed a possible solution to me. (NOT He proposed me a possible solution.) (LDOCE) (d) Word position e.g. Especially never comes at the start of a sentence: He loves fruit. He especially likes kiwis. (NOT Especially he likes...) (LDOCE) (e) Grammatical/lexical collocation e.g. Be careful to use the correct verb. I have to make a speech. I have to do a speech. (CALD) Whilst such error information is valuable in itself, the way the information is provided in pedagogical dictionaries still needs to be refined. Firstly, the selection of errors is not always appropriate. Some information is too basic for those who would use monolingual dictionaries. There is a trend to provide simple error information in beginners monolingual dictionaries such as Longman Active Study Dictionary (LASD) and Cambridge Learner s Dictionary (CLD), but most of those who would dare to use a monolingual dictionary are likely to be already familiar with such information. The error information should be tuned to the level of learners who would venture to use monolingual dictionaries. Secondly, it is difficult to deal with L1-related errors in general-purpose monolingual dictionaries which are not aimed at particular L1 speakers. For example, in the case of Japanese-speaking learners of English, incorrect sentences such as My house is Kyoto are quite common because of its parallel Japanese sentence Watashi no uchi ha Kyoto desu. This type of error cannot be adequately described in monolingual dictionaries because it is often caused by learners L1 knowledge and error patterns are different from L1 to L1. Thus, this type of L1-related errors should be treated more extensively in bilingual learners dictionaries. (c) Identifying the weak areas of learners: underuse of collocations It is not sufficient to use learner corpora to provide error information only. A more significant application of corpus-based techniques would be to show the gap in performance between native speakers and learners and so encourage learners to perform in a more target-like manner. One typical example would be the pattern of use of grammatical and lexical collocations. Table 1 shows the object noun collocates of the verb make found in the British National Corpus and the Japanese EFL Learner (JEFLL) Corpus. As one can clearly see, Japanese EFL learners tend to use relatively concrete objects such as money, food, friends, and so on. These collocates can be regarded as free combinations with the verb make in the sense to produce. On the other hand, native speakers use the verb make with more abstract nouns such as sense, way, use, decision and so on. Since phrases such as make sense, make a decision, etc. 319

333 ASIALEX 2005 are all highly frequent collocations used by native speakers of English, but constantly underused by Japanese EFL learners, it would be desirable to highlight these differences in dictionaries and to advise learners to use the keyword in a more target-like manner. One way to do this is to allocate more space to the item which needs more attention. In this particular case, one could describe the basic use of the verb make (i.e. the core meaning of produce ) more extensively in a beginner s dictionary and give more space and treatment to the extended and often metaphorical meanings in advanced learner s dictionaries. In so doing, we would be taking into account the gap between native speakers and L2 learners (Tono 2001:203ff). Rank by Freq. BNC JEFLL 1 sense money 2 way food 3 use breakfast 4 decision friends 5 mistake story Figure 1. Object noun collocates of the verb make in BNC and JEFLL Profiling learner language for pedagogical lexicography Finally, I would like to describe my on-going project on profiling learner language and examining its implications for pedagogical lexicography. I am currently working on two large corpora of Japanese EFL learners: one is called the SST Corpus (currently called the NICT JLE Corpus; approximately 2 million words; see Tono, et al. 2002), which is a corpus of 1,200 oral proficiency interviews taken as part of the Standard Speaking Test. Each transcript is based on a 15-minute speaking test. The second corpus is called the JEFLL Corpus (700,000 words; see Tono 2004 for more details), a corpus of free compositions by approximately 8,000 students (in-class, timed essays written without recourse to dictionaries). A project team is now working on the data of these two corpora and will publish the first report sometime toward the end of this year. This will be the first large-scale research project report based on Japanese EFL-learner corpora. There is tremendous potential in the exploitation of these resources for improving EFL syllabuses and materials design. Using analysis tools such as the Sketch Engine or the Shogakukan Language Toolbox, it is now possible to gain an overall picture of learners use of core and specialized vocabularies at various proficiency levels. This will provide very useful input for vocabulary learning theories and for syllabus design. We can identify not only the major error patterns of the learners but also the overuse and underuse of particular words, to which special attention could be drawn by way of usage notes or extra practice in relevant places in textbooks and reference materials. If developmental errors are identified for each proficiency level, dictionaries can then be customized to specifically address the relevant weak points for different levels of users. Electronic dictionaries, in particular, could change their interfaces and even their content according to individual user settings. It would be ideal to have multiple-levels of information in a dictionary, leaving it to end-users to choose the level, amount and type of content they see according to their needs. At the moment, we have very little of this sort of proficiency level-based information, but as relevant corpora grow in size and coverage, the type of customizable dictionary described above should become a reality in the not-too-distant future. Pedagogical dictionaries should deal with 320

334 WORDS IN ASIAN CULTURAL CONTEXTS all the issues I have discussed in this paper and provide the kind of support detailed above. One last point to be made, however, is that proper dictionary training also needs to be given so that learners can learn to access and exploit such information for their own ends and thus become more successful language users. References Church, K. W. and P. Hanks (1989), Word association norms, mutual information and lexicography. Proceedings of the 27th Annual Meeting of ACL, Vancouver, pp Hartmann, R.R.K. (ed.) (1979), Dictionaries and Their Users, Papers from the 1978 BAAL Seminar on Lexicography. Exeter Linguistic Studies, Vol. 1. No.4. University of Exeter. Hartmann, R.R.K. (ed.) (1983), Lexicography: Principles and Practice. London: Academic Press. Lee, D. (2001), Defining core vocabulary and tracking its distribution across spoken and written genres: evidence of a gradience of variation from the British National Corpus. Journal of English Linguistics, Vol. 29(3), pp Nakamura, T. and Y. Tono (2003), Lexical profiling using the Shogakukan Language Toolbox. In Murata, Yamada & Tono (eds.) Dictionaries and Language Learning: How can Dictionaries Help Human & Machine Learning? (ASIALEX 2003 Proceedings), pp Tono, Y. (1984), On the Dictionary User s Reference Skills. Unpublished B.Ed. Dissertation. Tokyo: Tokyo Gakugei University. Tono, Y. (1986), A scientific approach toward lexicography: the user perspective. LEO 15, pp Tono, Y. (1988), Can a dictionary help one read better? On the relationship between E.F.L. learners dictionary reference skills and reading comprehension. In G. James (ed.) Lexicographers and Their Works. Exeter: University of Exeter Press, pp Tono, Y., T. Kaneko, H. Isahara, T. Saiga and E. Izumi (2002), The Standard Speaking Test Corpus. Studies in Lexicography 11 (2), pp Tono, Y. (2001), Research on Dictionary Use in the Context of Foreign Language Learning. (LEXICOGRAPHICA Series Maior 106). Tübingen: Max Niemeyer Verlag. Tono, Y. (2004), Multiple comparisons of IL, L1 and TL corpora: the case of L2 acquisition of verb subcategorization patterns by Japanese learners of English. In G. Aston, S. Bernardini, & D. Stewart (eds.) Corpora and Language Learners. Amsterdam: John Benjamins, pp Abstract Pocket e-dictionaries in Japan: new developments Yukio Tono Miwa Nakamura Shinichi Yagi Meikai University y.tono@meikai.ac.jp This study investigates the comparisons of misinterpretations between handheld electronic dictionary (ED) and printed dictionary (PD), which were analysed from the marketing data of the 2004 Pop Song Translating Competition in Japan. This competition was held by Seiko Instruments Inc (SII)., the pioneer maker of handheld dictionaries in Japan. For the past three years, the use of handheld dictionaries in high school has taken a steep increase, which indicates growing importance and need for handheld dictionaries in schools. As such it is imperative that the dictionary manufacturers seek to understand the needs of this growing market and continually develop user-friendly dictionaries. This comparative study found two plausible results. 321

335 ASIALEX 2005 First, the limited size of ED screens did not pose a significant factor in causing misinterpretations by the user. Second, for PD users, there was a high tendency to skip words more frequently than ED users when they met unknown words while translating. The gap between the importance level indication in English-Japanese (E-J) dictionaries and the vocabulary in English textbooks. Abstract Tsuyako Touno Kwansei Gakuin University touno-tsuyako@mug.biglobe.ne.jp In this study, I will choose some words from English textbooks in Japan and look into the importance level indication of those words in English-Japanese (E-J) dictionaries and discuss (1) what kind of vocabulary should be taught in Japanese junior and senior high schools and (2) how E-J dictionaries have to be improved to meet the needs of Japanese learners of English. 1. Introduction Almost a hundred E-J dictionaries are currently on the market, and a dozen influential E-J dictionaries adopt a notable common practice of giving several thousand important or highly frequent words labels of importance or at which stage of the learning process those words have to be learned. Although each dictionary has its own criteria for labeling those words, quite a few are given different importance labels from dictionary to dictionary. There are also other facts in which vocabulary used in English textbooks are not listed in those influential dictionaries or simple words used in our daily lives are labeled as college level. After revealing these facts, I will compare the importance level indication of words in English textbooks and discuss the two research points shown in the abstract. 2. The importance level indication in learner s E-J dictionaries 2.1 The criteria for the selection of headwords As for the selection of headwords of learner s E-J dictionaries, compilers of those dictionaries have their own criteria for the number of headwords and the level indication for each word. From the criteria of some E-J dictionaries, the following three facts can be understood. (1) Each dictionary specifies its target users and has vocabulary which is necessary for them. Then it defines several thousand headwords as basic 322

336 WORDS IN ASIAN CULTURAL CONTEXTS important words, label them as such and gives them detailed description of their meanings and usages. (2) Each dictionary states what kind of vocabulary is contained as headwords in the dictionary. (3) Some dictionaries state what areas of vocabulary they have tried to contain most. 2.2 The importance level indication of some E-J dictionaries The following table shows the importance level indication of 7 E-J dictionaries. Table 1 Importance level indication of 7 E-J dictionaries names of the dictionaries Y [ ]1300 [ ]1800 [*]2200 for J.H for S.H important words W [ ]900 for J.H [ ]2900 for S.H [*]3400 for U.S. and G.R. [ ]8000 for G.R. G3 [ ]1100 for J.H. [*]3400 for S.H. [ ]5100 for U.S. and G.R. [ ]1300 [ ]1900 [* in red] [* in black] 1900 for state Lexis for J.H. and S.H. for S.H for center exam universities [ ]1100 must [ ]2000 most important [*]5000 important A.F. [ ]2000 most [*]5000 important [ ]8000 next important important 323

337 ASIALEX 2005 [ ]1800 most [*]4600 versity [ ]8100 for general readers K.N.C.7 important exams Prog cf.) J.H. indicates junior high school students / S.H. indicates senior high school students U.S. indicates university students / G.R. indicates general readers From this table, we understand that there are three differences in the criteria of the importance level indication. (1) The way headwords are divided [e.g. Are they labeled according to the stage of learning process or are they labeled just according to their importance?] (2) The number of labels each dictionary has. (3) The number of headwords each label contains. 3. How the words in high school English textbooks are labeled in E-J dictionaries In this section, I will choose some words from the viewpoint of their meaning and compare their level indication among seven E-J dictionaries. 3.1 Daily words I chose 278 daily words from the word lists of the textbooks and looked into their level indication and characteristics. First, the three words, piroshki, wonton and pa were not listed as headwords in some of the dictionaries. It should be noted that the first two words are the names of food. As an example, I will show you a chart of the level indication of daily words in a textbook. Chart 1 The level indication of daily words in a textbook eel * * in red * * heart attack *1 *1 *1 *1 * * lasagna octopus piroshki scone sushi wonton soup *2 *2 *2 *2 *2 meaning of symbols) : listed as a headword / :not listed as a headword *1: listed as a compound of heart / *2: listed only as wonton 324

338 WORDS IN ASIAN CULTURAL CONTEXTS Then I checked the level indication of the 278 words and found that 169 words among them (about 60 %) had a level for university students or general readers. Many of the 169 words were derivatives or compound words and were related to food and cooking, with the percentage of 15 % and 14 % respectively. From these facts, I learned that basic daily words such as toothpaste and kilogram had a level for university students or general readers in the E-J dictionaries, and I think the compilers of the dictionaries should try to rethink the level indication of those words for their necessity in everyday life. 3.2 Words for high-school life I chose 51 words from an Oral Communication textbook and compared their level indication. The words are related to everyday life of high school students. Again, such words as textbook, eraser, picnic, pajamas had a level indication for university students or above. And like the case with daily words, compound words had higher level indication. Then I looked into the situations in which a lot of words with high level indication were used. The result showed that high school textbooks need to include vocabulary for various kinds of activities other than school life, though such vocabulary had high level indication. From these facts, it can be said that the difference in the level indication for the 51 words shows that there is a difference in the number of words for high school students and in the number of words for their school life each dictionary contains. 4. Comparison of vocabulary between English textbooks in Japan and coursebooks for EFL/ESL students Next, I will compare the vocabulary between 16 English textbooks in Japan (henceforth 16 textbooks) and a series of coursebooks with four different levels (henceforth the coursebooks). The reasons for the selection of this series are: (1) the original series Headway received an excellent teaching material award and (2) this series and English textbooks have a similarity in that they both are organized to improve students four skills of English. The following is the original data for the two word lists. 16 textbooks the coursebooks total types total tokens First, I deleted proper names such as the names of people, countries from the lists and compared the percentage of the proper names. 16 textbooks had about 11.2 % and the coursebooks had about 18.4 % of proper names in each list. Then I looked into the differences in both lists. As for morphological aspects, vocabulary in the coursebooks had many derivatives and compound words. It also had many prefixes and suffixes for users of the books to increase their vocabulary. As for semantic aspects, the coursebooks contained many words for daily life (ex. air-conditioned, ATM), and more grammatical words than the 16 textbooks do. From these results, I suggest that English textbooks in Japan should have more derivatives, compound words and affixes for Japanese students of English to increase their vocabulary and they should contain more words for food. 325

339 ASIALEX Suggestions for the improvements of E-J dictionaries Frequency is an important criterion for the selection of headwords and the order of their definitions. Carter (1998:46) raises three problems associated with frequency counts. First, as for the lemmatization problem, E-J dictionaries don t list all the meanings of a homograph, and frequency or historical order is used to arrange homographs as headwords. I suggest that E-J dictionaries should choose and list homographs according to their frequency and the needs of learners. I will use BNC corpus, English textbooks in Japan, TOEIC test and JACET 8000 word list to get the relative frequency of English words. Second, as for inflections and derivatives of words, I think few E-J dictionaries indicate derivatives of words. So as many derivatives of important high-frequency words should be listed as possible to increase learners vocabulary. Third, Stubbs (2001:30-31) defines lexical item as follows; The term lexical item is therefore used to cover a range of individual words and phrases. We should interpret word to suggest that it implies individual words and phrases so that important phrases are listed as run-ons or headwords in E-J dictionaries. I think we should take into account not only the frequency of words but also the needs of learners and the critical judgment of compilers of E-J dictionaries in choosing headwords of an E-J dictionary. 6. Conclusion The importance level indication of headwords in E-J dictionaries differs from dictionary to dictionary. I think we should look into words from various viewpoints (e.g. their place in corpora, familiarity, range of use) and select them as headwords as objectively as possible. References Carter, R Vocabulary 2nd edn London : Routledge Stubbs, M Words and Phrases. Oxford : Blackwell Publishers Inc. English-Japanese dictionary Asano Hiroshi, et al Advanced Favorite English-Japanese Dictionary. Tokyo:Tokyo Shoseki Hanamoto Kingo, et al Lexis English-Japanese Dictionary. Tokyo:Obunsha Inoue Nagayuki, et al Wisdom English-Japanese Dictionary. Tokyo:Sanseido Konishi Tomoshichi, et al Genius English-Japanese Dictionary 3rd edn. Tokyo:Taishukan Shoten Kunihiro Tetsuya, et al Progressive English-Japanese Dictionary 4th edn. Tokyo:Shogakukan Takebayashi Shigeru, et al New College English-Japanese Dictionary 7th edn. Tokyo: Kenkyusha Yagi katsumasa, et al Youth Progressive English-Japanese Dictionary. Tokyo:Shogakukan Word list The Japan Association of College English Teachers (JACET) JACET List of 8000 Basic Words. Tokyo : JACET Coursebooks John and Liz Soars American Headway (Starter, 1, 2, 3). New York:Oxford University Press English textbooks Hanamoto Kingo et al Step English Revised. Tokyo:Obunsha Kaneda Michikazu et al Acorn English Course. Osaka:Keirinkan Matsumura Kenichi et al Genius English Course Revised. Tokyo:Taishukan Shoten 326

340 WORDS IN ASIAN CULTURAL CONTEXTS Morizumi Mamoru et al Exceed English Series. Tokyo:Sanseido / Nakau Minoru et al Quest English Course 2 Revised. Tokyo: Shogaku Tosho Ogata Takao et al Access to English. Tokyo:Kaitakusha Suenaga Kuniaki et al Unicorn English Course. Kyoto: Buneido Yoneyame Asaji et al Clipper English Course 2 Revised. Tokyo:Taishukan Shoten Yoshida Kensaku et al Birdland Oral Communication A 2nd edn Kyoto:Buneido 16 textbooks Hanamoto Kingo et al Step English Revised. Tokyo:Obunsha Hanamoto Kingo et al Step English. Tokyo:Obunsha Hanamoto Kingo et al Step English Revised. Tokyo:Obunsha Hisamura Ken et al All Aboard!. Tokyo:Tokyo Shoseki Hisamura Ken et al All Aboard!. Tokyo:Tokyo Shoseki Ikeda Satoshi et al Vista. Tokyo: Sanseido Jinbo Hisatake et al Hello there! Oral Communication A.Tokyo:Tokyo Shoseki Jinbo Hisatake et al Power On. Tokyo:Tokyo Shoseki Matsumura Kenichi et al Genius English Course. Revised Tokyo:Taishukan Shoten Matsumura Kenichi et al Genius English Course Revised. Tokyo:Taishukan Shoten Morizumi Mamoru et al Exceed. Tokyo:Sanseido Nakau Minoru et al Quest English Course 1 Revised. Tokyo:Shogaku Tosho Sano Masayuki et al Captain. Tokyo:Taishukan Shoten Takanashi Kenkichi et al Daily English Reading. Tokyo:Ikeda Shoten Yoneyama Asaji et al Clipper. Tokyo:Taishukan Shoten Yoshida Kensaku et al Birdland Oral Communication. A. Kyoto:Buneido Reverse indexing and customization future trends in bilingualized dictionaries Abstract Hugo T. Y. Tseng Soochow University hugozeng@scu.edu.tw In spite of all the kudos garnered from both students and teachers, monolingual English learner s dictionaries (MELDs) pose an inherent problem of not addressing target users specific needs. Their bilingualized adaptations, however, enjoy unparalleled advantage over their parent works because they are dictionaries with a focus. Bilingualized dictionaries (BDs), in this sense, combine the strengths of two types of dictionaries the authenticity and reliability of MELDs and the accessibility of native-language equivalents / translations of traditional bilingual dictionaries. However, the BDs are bound to lag behind the MELDs in terms of the currency of the contents. In order to compensate for this irreversible inherent shortcoming, we should engage in making BDs linguistically and culturally more focused, something which MELDs can never emulate. Since the contents are electronically stored, reverse indexing, with proper tagging when translated, is only a click away. This index functions as a mini, self-contained Native Language-English (e.g. Chinese-English) dictionary which directs EFL learners to the idiomatic English expressions with ample illustrations in the dictionary proper. A special appendix on core culture-specific words in the learners mother tongue could also be added. These culture-specific words are among the hardest to translate. With 327

341 ASIALEX 2005 this customized addition endorsed by name-brand BDs, EFL learners will benefit greatly when they want to talk or write about their culture in a foreign language. This paper aims to address these two user-friendly features of the BDs. Introduction The bilingualized dictionary (henceforth BD), of an originally monolingual dictionary, in which the original definitions and explanations are followed by the translations of these into the native language of the target dictionary user group, seems to be gaining in popularity. Take for example the monolingual English learner s dictionaries (henceforth MELD s) adapted for speakers of Chinese. The first such adaptation appeared in 1970, the bilingualization of the OUP s Advanced Learner s Dictionary of Current English, Second Edition (1963). This English-English-Chinese dictionary and its second edition had a monopoly, only to be broken 18 years later when the Longman counterpart entered the competition. The 1990 s saw a decade of duopoly of the Oxford and the Longman English BD s on the Chinese scene. At the turn of the millennium, however, two new competitors Collins COBUILD and Cambridge arrived on the market. Macmillan, the fifth leading brand in pedagogical lexicography, will soon jump on the bandwagon. Dictionary war has apparently spread to the spin-offs of MELD s. In spite of all the kudos garnered from both students and teachers, MELD s pose an inherent problem of not addressing target the users specific needs. They appeal to all students learning English as a second / foreign language, no matter whether they are Chinese, Congolese, Russian, or Brazilian. These bilingualized adaptations enjoy unparalleled advantage over their parent works because they are dictionaries with a focus. BDs, in this sense, combine the strengths of the two types of dictionaries the authenticity and reliability of MELD s and the accessibility of native-language equivalents / translations of traditional bilingual dictionaries. This unique combination of advantages somewhat satisfies the EFL learners wish to be maximally immersed in the English language and at the same time find the mother-tongue equivalents / translations. With the limitations of machine translation not making possible their immediate publication, the BD s are bound to lag behind the MELD s in terms of the currency of the contents. In order to compensate for this irreversible inherent shortcoming, we should engage in making BD s linguistically and culturally more focused, something which MELD s can never emulate. Since the contents are electronically stored, reverse indexing, with proper tagging when translated, is only a click away. This index functions as a mini, self-contained Native Language-English (e.g. Chinese-English) dictionary which directs EFL learners to the idiomatic English expressions with ample illustrations in the dictionary proper. A special appendix on core culture-specific words in the learners mother tongue could also be added. These culture-specific words are among the hardest to translate. With the close collaboration of lexicographers and translators, this customized feature, endorsed by name-brand BD s, will definitely benefit EFL learners when they want to talk or write about their culture in a foreign language. This paper aims to address these two user-friendly features of the BD s. Directionality in bilingual Chinese dictionaries Bidirectional dictionaries seem to be standard practice in European languages. 328

342 WORDS IN ASIAN CULTURAL CONTEXTS Oxford University Press and HarperCollins, among other publishers, have produced a broad spectrum of English-European / European-English dictionaries. Even the four concise dictionaries (French, Spanish, Italian, and German) included as an appendix at the back of The Random House Webster s Unabridged Dictionary are bidirectional. On the Chinese scene, in sharp contrast, monodirectional dictionaries have a long tradition. The reference sections in the bookstores or libraries in Taiwan or the PRC testify to this claim. Bidirectional dictionaries are rare. The first conspicuous exception is Concise English-Chinese Chinese-English Dictionary (1986, with subsequent revised and enlarged editions) jointly published by OUP and the Commercial Press Beijing. Interestingly, although this pocket dictionary overambitiously claims, in its preface, to serve both Chinese-speaking students of English and English-speaking learners of Chinese, it has been far more popular among the Westerners than the Chinese speakers. Nearly two decades later in 2004, Collins Chinese Dictionary (released on the Chinese market as FLTRP COLLINS English-Chinese Chinese-English Dictionary), jointly published by HarperCollins and Foreign Language Teaching and Research Press Beijing, ended the OUP s monopoly. This new pocket dictionary is similar to its rival in size and coverage, but takes a more pragmatic approach to the target user group. Its preface has made clear that it specifically serves Chinese-speaking students of English, and it is corpus-based and rich in several user-friendly features which help students use the words right both grammatically and pragmatically. Interestingly, HarperCollins Chinese partner in this collaboration produced another pocket bidirectional dictionary under its own imprint a month prior to the publication of Collins Chinese Dictionary. Does this move signify that the bidirectional dictionary is emerging as a new lexicographical genre on the Chinese scene? Given a the bigger the better mindset among many Chinese, to what extent this pocket Collins Chinese Dictionary and its competitors will exert influence on Chinese lexicography remains to be seen. Reverse indexing The MELD may be popular among some upper intermediate to advanced EFL learners, but according to the sales statistics I informally obtained from several book dealers in Taiwan and the PRC, its bilingualized adaptation appeals to a much larger market. This serves to explain why the production of the BD is white-hot competitive all the five name-brand MELD s except Macmillan have already had their Chinese BD s (the Macmillan BD is in its final preparation and will come out soon). MELD s are an excellent reference source from which EFL learners learn authentic and reliable English. Their BD s make them more accessible and less intimidating. Besides traditional English-Chinese dictionaries and MELD s, Chinese BD s have emerged as a significant third force in facilitating EFL learning. With a BD at hand, learners gain access to all the merits appertaining to an MELD and a medium-sized English-Chinese dictionary. A prima facie ideal reference book for Chinese EFL students, the BD still lacks one important component the Chinese-English section for target language production. Translation from Chinese to English seems ubiquitous, and it accounts for a significant percentage of language production activities. Rather than suppress or even condemn this natural need, it would be better to actively face the inevitable mother-tongue interference and lead the learner to the right track. A full-fledged Chinese-English section of comparable measure will take up too much space, and, against the backdrop 329

343 ASIALEX 2005 of the history of bidirectional dictionaries, its efficacy may be uncertain. Therefore a simple Chinese-English word list can be appended to the BD for easy retrieval of further information (grammar, usage, examples, etc.) in the dictionary proper. Since dictionary data are all electronically stored, this reverse index is only a click away if key words in Chinese are properly tagged when the MELD is bilingualized. Reverse-indexing the BD sounds simple in theory, but it may not be so in practice. What the headwords should be in the Chinese-English index is a matter of careful deliberation. The Chinese equivalents in the BD should be easier to deal with, although the language used in dictionary definitions certainly needs slight modifications to be granted a headword status in the reverse index. Explanations and parenthetical glosses in the BD are often lengthy, and can be complicated and harder to present in the reverse index. They are necessary products of English culture-specific words which do not have their Chinese equivalents. However, all these complications can be reduced if we make this reverse index a keyword index, as is adopted in some dictionaries of idioms and quotations. The Chinese dictionary definitions, explanations, or examples can be presented in full or in fragments in this index, with the keywords bold-typed and arranged in Pinyin alphabetical order. In this Chinese-English index appended to the BD, Pinyin-ordered bold-typed Chinese keywords in Chinese characters (Pinyin-Romanized here to avoid decoding problems), possibly nested in longer plain-typed phrases, are followed by their English equivalents for easy retrieval, as illustrated by the following sample index: ji-mang di-xia-tou Fa-guo cui-pi bai mian-bao bang ke-xing-de dui (mou-ren) you hao-chu Culture-specific words to duck (one s head) baguette feasible to profit (sb) The BD s inclusion of the latest neologisms is bound to lag behind because it takes years to translate and edit. This seems to be an irreversible weakness. But some BDs did maneuver to append a new words supplement containing those appearing since the publication of their parent MELD s and being considered for inclusion in the next edition. However, instead of making every effort to collect new words, I strongly suggest that the time and effort be spent on further customizing the BD to make it more focused and better serve its users. The Chinese-English reverse index mentioned above is one, and the presentation of Chinese culture-specific words is another. Culture-specific words have always been among the hardest to translate. In interlingual communications, culture is a very interesting topic, but more often than not, Chinese speakers do not know how to properly introduce their culture to foreigners in correct and idiomatic English. The bilingualized Longman Dictionary of Contemporary English (1997) pioneered the introduction of concrete Chinese culture-specific words in four and a half out of sixteen color pages. Dozens of words on Chinese architecture, food, and wushu are illustrated and translated, though some of them are common enough in the West. Most of these terms and expressions in the Longman BD are nothing new; they can easily be found in Chinese-English dictionaries. Furthermore, some English equivalents are wordy explanations only good for language comprehension. For instance, zongzi, a Chinese sticky rice version of the Mexican tamale, is lengthily rendered as glutinous rice wrapped in bamboo leaves. Also, some 330

344 WORDS IN ASIAN CULTURAL CONTEXTS equivalents sound awkward. For instance, langqiao, the Chinese covered bridge, is translated as the collocationally unusual shelter bridge. The number of Chinese culture-specific terms in this Longman BD is quite limited, and some of the English equivalents may not be perfect for language production. Nonetheless, the uniqueness of this dictionary is that it aims to transcend the BD s role as a mere translated MELD; it also tries to be an intercultural conveyor. However, for some unexplained reason, this useful feature has been completely discarded in the latest edition (2004). My suggestion for perfecting the design of the BD, besides the appending of the Chinese-English reverse index, would be the fuller inclusion and better treatment of Chinese culture-specific words. Special dictionaries dedicated to the English rendition of Chinese cultural words can be found on the market. General-purpose Chinese-English dictionaries also cover plenty of such words and expressions. However, what these dictionaries offer is often problematic. The greatest problem is that they appear to copy their English equivalents from each other without painstaking research into their adequacies. From one dictionary to another, we find that explanations for the Chinese zongzi tamale almost always read something like pyramid-shaped dumpling made of glutinous rice wrapped in bamboo or reed leaves. This kind of verbose description is useless in Chinese-English translation. An accurate, idiomatic, and text-insertable equivalent has to be found or even coined to be truly useful (Tseng 2004). Another major problem with those special dictionaries of cultural terms is that the number of entries is too large. The lexicographer has to trim the size and make the entries relevant to Chinese EFL learners life experiences. The culture-specific vocabulary to be entered may include that on education, arts, music, traditional sports and games, food, clothing, architecture, religions and folk beliefs, etc. The selection of headwords will be based on corpus frequency count and native lexicographers intuition. The presentation will be based on semantic fields for more efficient vocabulary learning, instead of alphabetical order. The equivalents will be meticulously researched with empirical evidence (consulting specialized dictionaries and tested against the Google search engine) to guarantee their accuracy and idiomaticity. If no existing equivalents are available, the lexicographer will closely collaborate with the translator on the production of short text-insertable translation equivalents (again tested against the Google search engine to see if similar or even identical expressions are found in English-language texts). The proposed English equivalents, no matter how scientifically accurate and theoretically sound they may be, have yet to be tested by the market. The reason for this resigned pessimism is that sometimes an excellent equivalent may not be appreciated by the user, and publicity marketing seems to play a crucial role in determining if the new term successfully enters the native speaker s lexicon (ibid.). Sample entries of Chinese cultural-specific terms (Pinyin-Romanized here to avoid decoding problems using Chinese characters) with their English equivalents, will look something like the following. Parenthetical glosses add some background information which will not appear in the real dictionary: xie-shi-yan teacher appreciation banquet (The literal thank-teacher banquet or teacher-thanking banquet sounds awkward. Graduation party is very common in the West, but means quite a different event. Teacher appreciation banquet is the closest we can find, and a Google search testifies to this Taiwanese-style graduation party.) fang-niu-ban laissez-faire class (Literally this term means cattle-herding class, denoting 331

345 ASIALEX 2005 gu-qin wen-zai-yu a low-achievement class in the Taiwanese junior high school, in which all the teachers have given up actively educating them. These students are said to be left alone like cattle grazing. The literal cattle-herding class or cattle-grazing class doesn t clearly convey the idea. Laissez-faire class, though collocationally unusual, more or less describes this situation.) guqin, or Chinese seven-stringed zither (This traditional Chinese plucked stringed instrument has been known in the Western music world in its Pinyin guqin, as testified by a Google search. The parenthetical gloss or Chinese seven-stringed zither explains what it denotes to those who see it for the first time.) Taiwan whitebait (This translucent-to-white fish fry is a treasured delicacy on the Taiwanese dinner table. It is not, like the whitebait, of a single species of fish, but belongs to a group of similar fish in their young. The whitebait in the UK is generally much bigger than the one in Taiwan. Taiwan whitebait may have a clear denotation.) Conclusion By adding some features like the two discussed in this paper, the new BD with a linguistic and cultural focus can then address target users specific needs. It has the excellent EFL-oriented tradition of an MELD, the Chinese glosses which lift the linguistic barrier an monolingual English dictionary creates, a self-contained Chinese-English index making it semi-bidirectional, and an appendix containing much sought-after Chinese culture-specific terms with their empirically researched English equivalents. This improved BD customized for the Chinese-speaking market will then be an even better reference book for Chinese EFL students than its original. References Tseng, Hugo T. Y. (2004), Rendering Chinese culture-specific vocabulary into English predicaments and prospects, in Chan Sin-wai (ed.), Translation and bilingual dictionaries (Tübingen: Max Niemeyer Verlag), pp Abstract The implicit meaning of written advertisements Anneke Tupan Petra Christian University, Surabaya- Indonesia atupan@peter.petra.ac.id Some people often do not understand what is being communicated in the advertisements or may have different interpretations of the same advertisement. People who share the same cultural background can fail to understand a certain advertisement within the same culture. Research on a cigarette advertisement is done to find out the implicit expression of culture. Advertisements contain the expression of culture that reflects the way people think, live, and enjoy their life. This shows that learning a language cannot 332

346 WORDS IN ASIAN CULTURAL CONTEXTS be separated from understanding of its culture. Therefore, people should be aware that when copywriters choose a word, they do more than simply name an object because they also convey feelings about what they are describing. People cannot rely only on the explicit words but more importantly they should be able to see what lies behind them. The theory of discourse analysis (Brown and Yule, 1983) and the concept of context of situation (Halliday and Hasan, 1989) are used in the analysis to overcome the problems. The result shows that the intended meanings are meant to educate the young generation, and also aim to encourage people to stand for righteousness, remind them to value friendship and honor loyalty and honesty. In short, the advertisements under study use words that imply the value of being creative, smart, having multiple interpretations, and being a trendsetter of the time. Copywriter and advertisement Language is a system of communication, a medium for thoughts, a vehicle of literary expression, a social institution, a matter of political controversy and a catalyst for nation building (O Grady et al., 1996:1). This proves that language plays an important role in human life as a means of communication. Verbal communication as a written communication in the form of advertisements has a wide range of influence towards society. One of the reasons is because they are highly structured by the copywriters. They consider themselves as word mechanics skilled professionals with a fondness for words. They must be imaginative, and still, their writing must be succinct and eye catching. They are good writers, even though some of their vocabularies and structures may be somewhat unconventional. After all, their entire job is to attract the readers. If their copy is grammatically perfect but lacks fire and vitality, readers will not be interested and the product will not sell. In other words, advertisement is a means of communication that aims to attract the readers, inform the readers, interest the readers and persuade them to think about the product. Advertisements can be classified into three categories namely compound, complex and sophisticated advertisements (Dyer, 1982:88). Compound advertisements contain information, but they rely on pictures to do the persuading, while the facts are left to the copy. For example, in some fashion advertisements in women s magazines, the name of the store, the address, the telephone number, the brand name of the garment might be listed alongside the drawing or photo of it. If there is a headline or a caption, it is of a semi-technical nature such as New Spring Collection from Guess. The word new might be regarded as persuasive rather than informative but the real persuasion comes in the picture, which is usually featuring attractive-looking models or an artistic modernist line drawing of a pencil slim and elegant woman. In a compound advertisement, the advertiser is obviously hoping that the reader will associate the product with the total impression garment with model. Complex advertisements usually concentrate on the presentation of luxury and status; the background takes over, the product merges into it. The visual and verbal imagery evoke the status feelings associated with money, wealth, elegance, luxury, and the public display of these things. For example, in most advertising for expensive consumer durable carpets, bedroom suites, et cetera it is the image created by the combination of commentary and photography that is selling point, not the specific carpet or site. At first glance, it is sometimes hard to see what precisely is being offered because the product is buried in the total image created. 333

347 ASIALEX 2005 Sophisticated advertisements are an extension of the complex advertisement. They usually explore the hidden or subconscious feelings; subtle associations are made between product and situations. Sophisticated advertisements imply what the copywriter s intended meaning is. Different people read and interpret advertising text in different ways. This is the type of advertisement that is observed since it challenges any readers to bring out the hidden meaning through a discourse analysis research. The research focuses on the implicit meaning of cigarette written advertisements namely A Mild for some reasons. Firstly, on any cigarette products there is always a warning printed at the bottom of the package about the danger of smoking for people s health. Therefore, the copywriters do not invite the readers to buy the products directly but they imply the persuasion to buy them in such a way that nobody is aware that they are actually cigarette advertisements. What makes the discussion more interesting is that readers may have different interpretations of the same advertisement. For the copywriters, nobody s interpretation is right or wrong as far as they attract readers attention so that the products are sold out. The next reason for the unique form of the advertisement is the copywriter s creativity in designing the advertisements. The lay-reader cannot understand them readily and they even think that they are a political campaign instead of cigarette advertisements. This is because the copywriters always make use of current issues as the vehicle for their advertisements. They wrap the selling message product in moralistic and smart message wrappers. It makes people need to think harder to understand what the copywriters really want to communicate. This condition unconsciously drives everybody s attention toward not only the implicit meaning of the advertisements but the products also. These facts make the written advertisements of A Mild worth discussing in order to bring out the copywriter s implied message in the advertisements through a discourse analysis study. When talking about advertisements or attempting to analyze them, most people tend to assume that advertisements are vehicles for the communication of exaggerated publicity and that they may be transparent or invisible carriers at that. People tend to take for granted that what is on the screen or page is what the advertisement means and they measure advertisement against some assumed reality which could replace the unreal images which constitute most advertisements. For example, the images of men and women in advertisements are usually considered to be mythical rather than real, and also stereotyped. This kind of criticism usually gets bogged down in arguments about the extent to which such images are true or false. It also seeks to replace distorted images with representations of people and situations as they really are. It assumes that there is a simple and better reality to replace the stereotypes and myths and ignores the fact that advertisements are in themselves a kind of reality which have an effect. In this sense, advertisements are not secondary to real life or copied or derived from it. Advertisements are what some critics call specific representational practices and produce meanings that cannot be found in reality. There is no simple reality to replace the falseness of advertisements, and there are no simple alternatives to stereotypes. In order to gain better understanding of the role that advertising plays in the society, people need to ask how advertising organizes and constructs reality, how ideology and meanings are produced within the advertisement discourse and why some images are the way they are, or how they could have been constructed. Most semiotic/structural studies of advertising text distinguish between their outward manifestation and inner mechanisms the codes and conventions that organize 334

348 WORDS IN ASIAN CULTURAL CONTEXTS and release the meanings of a text in the process of viewing or reading. Such codes are what make meaning possible. Texts result from the dynamic interplay of various semiotic, aesthetic, social and ideological processes within them that also operate in the culture outside them. The audience member is involved in the work of the text and the production of its meaning; his or her own knowledge, social position and ideological perspective is brought to bear on the process of the construction of meaning. As Judith Williamson argues: Advertisements must take into account not only the inherent qualities and attributes of the products they are trying to sell, but also the way in which they can make those properties mean something to us.. Advertisements are selling us something besides consumer goods; in providing us with a structure in which we, and those goods are interchangeable, they are selling ourselves. (1978:13) Advertisements do not simply manipulate people; they create structures of meaning which sell commodities not for themselves as useful objects but in terms of readers as social beings in their different social relationships. Products are given exchange value : advertisements translate statements about objects into statements about types of consumer and human relationships. For example, Williamson refers to an advertisement for diamonds ( A diamond is forever ) in which they are linked to eternal love: the diamond means something not in its own terms as a rock but in human terms as a sign. A diamond cannot buy love, but in the advertisement it is the diamond that is made to generate love and comes to mean love. And once this initial connection has been made, readers almost automatically accept the object for the feeling. People and objects can become interchangeable as in, for example, the slogan The Pepsi generation, or The Marlboro country. It is in this sense that advertisements should be seen as structures that function by transforming an object into something that is given meaning in terms of people. The meaning of one thing is transferred to or made interchangeable with another quality, whose value attaches itself to the product. Context of situation Contexts can be understood as the particular environments in which communication, texts and meaning making occur, and in which they function as meaningful (Schirato and Yell, 2000:109). Halliday proposes a model for analyzing text/context relations, which accounts for the interdependent relationship between particular social contexts and the kinds of textual practice that take place in these contexts. He says that context can be separated into three aspects or dimensions namely Field, Tenor and Mode. Field refers specifically to the social action of the text, what the language is being used to do. The field of discourse refers to what is happening, to the nature of social action that is taking place: what is it that the participants are engaged in? (Halliday, 1989:12). Fields are the discursive and institutional domains where cultural and textual practice occur (Luke, 1996:327). In order to analyze the context of situation, people need to ask some questions such as: What is the social activity performed by the text, or the social occasion in which the text is embedded? What is the subject matter of the text? What discourses are involved, and what institutions? What rules and procedures associated with the field are implicated through the text? How are these being negotiated? What institutional values and hierarchies are implicated through the text? 335

349 ASIALEX 2005 Meaning making is a form not just of social action but of interaction. When people make meanings they address them to some other a listener, reader, or audience, actual or implied: sometimes they simply address meanings to themselves. Communicative acts are not just simple exchanges of information, but are about the allocation, negotiation, acceptance or rejection of a variety of social relations. Tenor refers to this interactive dimension of communication. Here is Halliday s definition of tenor: The tenor of discourse refers to who is taking part, to the nature of participants, their statuses and roles: what kinds of role relationship obtain among the participants, including permanent and temporary relationships of one kind or another, both the types of speech role that they are taking on in the dialogue and the whole cluster of socially significant relationships in which they are involved? (Halliday & Hasan, 1989:12) To analyze tenor relations in more detail, the categories suggested by Poynton (1985, 76-77) are used. She identifies three aspects, which constitute tenor namely social distance (contact), emotion or attitude (affect) and power. Contact or social distance, refers to the positioning of the participants as socially intimate or distant. While affect refers to the kinds of attitudes and emotions that are being displayed as part of the interaction and power refers to the extent to which the participants are positioned as equal or unequal. Mode concerns the formation of the communication as a text, including the code/s and the medium/s involved. Halliday offers the following definition of mode: The mode of discourse refers to what part the language is playing, what it is that the participants are expecting language to do for them in that situation: the symbolic organization of the text, the status that it has, and its function in the context, including the channel (is it spoken or written or some combination of the two?) (Halliday & Hasan, 1989:12) The concepts of code and medium show that texts are material forms as well as meaning- making practices. The code and the medium make possible, and also limit, the range of meanings. They provide a diversity of materials (medium) and ways of shaping these materials (codes). Methods of study The source of data is A Mild written advertisements in Indonesian version collected from newspapers, magazines, billboards, fliers, stickers issued in 2003 and Advertisements published in 2003 are under a thematic advertisement of BUKAN BASA-BASI (Not a Lip Service) while the ones published in 2004 have not only a different theme namely KALAU BENDA BISA NGOMONG (If Things Can Talk) but also pictures to symbolize something. The data are analyzed by using the concept of context of situation suggested by Halliday & Hasan (1989). There are twelve advertisements chosen randomly and they are given the English version to help the non-native readers to understand them. The following is how the data are classified for the analysis. Figure 1: A Mild Verbal Advertisements 2003 NO 1 2 ADVERTISEMENTS Jadi Tua Itu Pasti Jadi Dewasa Itu Pilihan (Growing old is a certain becoming mature is a choice) Jangan Tunda Besok Apa Yang Bisa Dikerjain Lusa (Don t postpone till tomorrow what you can do the day after) 336

350 WORDS IN ASIAN CULTURAL CONTEXTS Ringan Sama Dijinjing Berat Elo Yang Pikul (Share equally in carrying light burdens but you take the heavy ones) Teman Sejati Selalu Berbagi Emang Pacar Bisa Dibagi? (True friends always share but can a girl friend be shared? Gue Berpikir Karena Itu Gue Tambah Bingung (I am thinking therefore I am getting more confused) Jenius Ada Batasnya Bodoh Nggak Terbatas (There is a limit to being genius but not for becoming stupid) Figure 2: A Mild Verbal-Visual Advertisements 2004 NO SYMBOL ADVERTISEMENTS 7 An Armchair A Wok A Bed A Television Cards A Crown Findings and discussions Kalau Nggak Dibersihin Kutu Busuknya Nggak Bakalan Pergi (If it is not cleaned up the bugs will not go) Jangan Cuma Bisa Manas-Manasin Doang (Be not only capable to provoke) Rawat Baik-Baik atau Gue Obral Rahasia Elu (Take good care or I will open your secrets) Nggak Usah Protes Kalo Cuma Bisa Nonton (Don t protest if you can only watch) Ngobrol Dong Biar Ketemu! (Let s talk to find a solution) Gak Heran Kalo Jadi Inceran (No wonder everybody is going for it) The results of the analysis are put in a contextual configuration. In analyzing the 2003 advertisements, the interpretation of the advertisements rely on the verbal version only while for interpreting the ones issued in 2004, the inserted pictures help readers to come to the interpretation since they specify the participants and symbolize the implicit meaning of the advertisements. Figure3: Contextual Configuration of A Mild Written Advertisement NO FIELD TENOR MODE Raising awareness Being a hard working people Respecting Cooperation Respecting a relationship Educating the society Educating the society Warning the society Warning the political parties Warning the opposition party Warning the political intruders Encouraging a leader of a political party Educating the society Between politicians Seniors to youngsters Authority to youngsters Authority to youngsters Between authorities Politicians to society Authority to society Authority to politicians Between political parties Authority to political parties Authority to a political leader Authority to politicians Billboards, fliers Fliers, stickers Billboards, stickers Billboards, magazines Magazines, newspapers Billboards, newspapers Billboards, newspapers Billboards, magazines Billboards, stickers Billboards, newspapers Billboards, magazines, newspapers, stickers Billboards, magazines The field of discourse for the advertisements is similar. The first data are issued when the society is in the process of learning to have a great concern towards political situation since they are published before the general election campaign. The second data are issued during the general election campaign to vote for the leading political party in one step and for President and Vice president in the second step. Therefore the subject matter is always about the latest political issues. Concerning the tenor of discourse, the copywriters play different roles in each of the advertisements. Sometimes they act as if they were politicians, or senior citizens, or authorities, or political parties. Whatever role they create, the implicit meaning is to 337

351 ASIALEX 2005 educate the society particularly the young generation. The target of the advertisements is obviously young people. This can be seen from the use of informal language style and the Jakartanese dialect which is the icon of the way young people communicate in this country. Well educated youngsters are expected to be the future leaders who will bring this country to prosperity. The mode of discourse is in such written forms as billboards, magazines, newspapers, stickers and fliers. They are meant as didactic written documents in the form of media that can be widely read by society. When the lay-readers don t understand what the copywriters want to say, they can at least talk about it or even think harder to come to an interpretation and this reveals different interpretations of the same advertisement. Even though the lay-readers eventually come to different interpretations, they still point to the same core, which is didactic. Conclusion From the discussion, it can be concluded that the implicit meanings of the advertisements are meant to educate the young generation, and aim to encourage people to stand for righteousness, remind them to value friendship and honor loyalty and honesty. The advertisements under study use words that imply the value of being creative, smart, having multiple interpretations, and being a trendsetter of the time. References Brown, Gillian and George Yule (1983), Discourse Analysis (New York: Cambridge University Press). Dyer, G (1982), Advertising as Communication (England: Clays Ltd. and Company). Goddard, Angela (1998), The Language of Advertising (New York: Routledge). Halliday, M.A.K. and Ruqaiya Hasan (1989), Language, Context, and Text: Aspects of Language in a Social-SemioticPperspective (Hongkong: Oxford University Press). Luke, A (1996), Genres of power? Literacy education and the production of capital, in Literacy in Society, eds. R. Hasan and G. Williams (London & New York: Longman). O Grady, M. Dobrovolsky, and F. Katamba (1996), Contemporary Linguistics: An Introduction (England: Copp Clark Pitman Ltd). Schirato, Tony and Susan Yell (2000), Communication and Cultural Literacy: An Introduction (Sydney: Allen & Unwin). Williamson, Judith (1978), Decoding Advertisements: Ideology and Meaning in Advertising (London: Marion Boyars). Abstract How good are graded English readers for ESL/EFL students? Toshihiko Uemura Siebold University of Nagasaki yandi@sun.ac.jp This paper discusses the validity and usefulness of graded English readers for ESL/FEL students in comparison with authentic English texts. This corpus-based study reveals to what extent the ESL/EFL texts of controlled vocabulary and grammar codes are similar 338

352 WORDS IN ASIAN CULTURAL CONTEXTS to the genuine English ones written by contemporary American writers, and what lexical and/or grammatical differences there are between these English data. 1. Aim of this paper Graded English readers such as Cambridge English Readers, Oxford Bookworms, and Penguin Readers are well-known to ESL/EFL professionals and specialists. How useful are they for ESL/EFL students? Are there any significant differences between these English readers and contemporary books for native speakers of English? This paper focuses on these questions by employing the corpus-based analysis of English texts. 2. Procedure First, two types of English text data, full texts of six graded English readers published by the above mentioned publishers and 113 authentic English texts, which appeared on the New York Times Book Review Section, First Chapters, were converted into the plain text file format. (Throughout this study, they are referred to as 6GR and 113NYT, respectively.) Then, 6GR and 113NYT text files were input into Mike Scott s English text analyzing software, WordSmith Tools Version 4. (Hereafter, referred to as WordSmith.) WordSmith output was examined and analyzed in this paper. 3. A basic comparison When English texts are input into the WordSmith, and its Wordlist function is activated, it produces two types of wordlists (a frequency-based wordlist and an alphabetical wordlist), and a table of statistical information. All numerical data in Table 1 and Figure 1 below were taken from the 6GR and 113NYT statistical tables. The WordSmith defines tokens as all words in the texts used for the wordlist, and types as the distinct words in the wordlist. As Table 1 indicates, 113NYT is larger than 6GR in terms of the number of tokens and types. Although all 6GR texts are in the same 3000-word level texts, the total numbers of types vary from less than 2,500 to over 3,500 types. Table 1. Tokens and types in the six English texts 6 G R title tokens types Gibbon-West(1998) 30,880 3,136 Hancock(2000) 28,405 2,638 Huxley-Cartledge(1999) 35,796 3,585 Maley(1999) 27,919 3,767 Rodat-Kehl(2000) 27,112 2,414 Waugh-West(1998) 29,010 3,136 total 179,122 8, NYT 252,562 21, NYT is composed of 113 English texts every eight to ten of which appeared on the New York Times website per week, and were gathered from November 28, 2004 to March 3, Most of these texts were taken from the first chapter of recently-published humanities or social science books. Figure 1 illustrates the distribution of tokens in terms of their number of letter. In both 339

353 ASIALEX 2005 data sets, the peak of total token distribution comes within the range of three to four letters. 60,000 50,000 40,000 No. of Tokens 30,000 20,000 6 GR 113NYT 10, N-letter Word Figure 1. N-letter word The patterns of n-letter type distribution in Figure 1 are very similar and related. There exists a high correlation between 6GR and 113NYT. (r=.994. The correlation is significant at the 0.01 level.) 4. JACET8000 as the stop list On the basis of the frequency of occurrence, all types in the 6GR and 113 statistical tables were grouped into four categories: frequency levels 1 to 9 times, over 10 to 99 times, over 100 to 499 times and over 500 times, then the total number of types in each category was counted. Figure 2 shows the type distribution of the four categories at the default setting, and Figure 3 the type distribution with the JACET8000 as the stop list. When the JACET 8000 is used as the stop list for WordSmith, this software eliminates all types and tokens in the text data which are identical with those in the JACET 8000, and only the remaining types and tokens are tabulated for its statistical table. Again, there were high correlations in both cases: the correlation coefficient of the default setting is r=.986, and that with the JACET8000 is r=.997. (In the former default case, the correlation is significant at the 0.01 level, and in the latter case with the JACET8000, at the 0.05 level.) The results of high correlations seem mainly due to the fact that all 6GR and many 113NYT books belong to the same genre of fiction. 340

354 WORDS IN ASIAN CULTURAL CONTEXTS No. of Types before after 0 1 to 9 over 10 over 100 over 500 Frequency of Occurrence Figure 2. 6GR with and without the JACET8000 It is evident from Figures 2 and 3 that after the use of the JACET8000 as the stop list, the total number of types whose frequency of occurrence is over 10 to 99 times becomes smaller, and that the differences in the number of types over 100 to 499 times and over-500 times in the two data become almost negligible. However, a close observation reveals that there are minor differences No. of Types before after to 9 over 10 over 100 over 500 Frequency of Occurrence Figure NYT with and without the JACET8000 Among types whose frequency of occurrence was over 10 to 99 times, those found more than 10 times in 6GR and less than 10 times in 113NYT are: PARATROOPER (43, 1), MEDIC (21, 1), HULLO (18, 3), MUMMY (11, 1), PORRIDGE (10, 1), FERTILIZING (10, 1). Those of over 10 times occurrences in 113NYT and less than 10 times in 6GR are: SURF (40, 2), MARVEL (21, 4), BEAVER (20, 3), RIVET (18, 2), SIDEWALK (16, 7), BLOND (16, 12), STEAMBOAT (14, 1), LIFEGUARD (14, 2), AUNTY (14, 2), HUFF (12, 1), TATTOO (11, 3), YEAH (11, 7), RANCH (10, 5), FORTH (10, 9), EXPAT (10, 2), DORM (10, 3). (The digit after the comma in parentheses indicates the number of texts in which the types in question are actually found.) Except HULLO (three out of six 6GR texts), PARATROOPER, MEDIC, MUMMY, PORRIDGE, and FERTILIZING were found only in 341

355 ASIALEX 2005 one of the 6GR texts. Their relatively high frequency of occurrence in a single text suggests they are topic-related words unique to the English reader in question. Similarly, the 113NYT types of over 10 times frequency of occurrence were limited in certain First Chapters texts. As Table 2 suggests, while WALKED and BOYS in 6GR occurred more than 100 times, they appeared less than 50 times in 113NYT. Conversely, FRENCH, BECAME, SAYS and YORK occurred more than 100 times in 113NYT, they were found less than 50 times in 6GR. Table 2. Types of wide gap of occurrences forms 6GR 113NYT WALKED 118, 6 (48, 29) BOYS 104, 5 (37, 22) FRENCH (18, 3) 118, 24 BECAME (30, 6) 116, 6 SAYS (22, 5) 112, 38 YORK (4, 2) 107, 35 In Table 2, the types of less than 50 times frequency of occurrence are marked, and the digit after the comma in each cell indicates the number of texts in which the types in question are actually found. In 6GR, FRENCH and YORK were less frequent words, and the number of texts they appeared in was three and two, respectively. Although found in all 6GR, BECAME was less frequent in comparison with its occurrence in 113NYT texts. WALKED and BOYS were found less than 50 times in 113NYT. These wide gaps of type occurrences seem to reflect what each text is describing. 5. Verb Collocation Patterns The default 6GR and 113NYT wordlists were lemmatized. Then the lemmatized forms were counted in which the total number of frequency of occurrence exceeded 500 times. There were 56 in 6GR, and 68 in 113NYT. Among these types, there were 11 verbs in 6GR and 8 in 113NYT, which are listed in Table 3. (Marked in the table are the cells whose frequency of occurrence was less than 500 times.) Table 3. Verbs of over 500 times occurrences types 6 GR 113NYT BE HAVE SAY GO DO THINK 727 (414) LOOK 689 (425) KNOW SEE 612 (450) COME 575 (483) GET 536 (415) MAKE (486) 631 LIKE (472)

356 WORDS IN ASIAN CULTURAL CONTEXTS The WordSmith command of Concord analyzes the collocation of the input text data. Every collocation found more than five times in the text is listed as a Cluster by the WordSmith Concord. All lemmatized verb forms in Table 3 were examined, and their Clusters were listed in Table 4. (Lemmatized verb forms with a total number of occurrence less than 500 times were marked in the table.) Table 4. Collocation Patterns types 6GR 113NYT HAVE HAVE+TO 121 HAVE+TO 105 (HAD+TO 37) (HAD+TO 57) SAY SAY+THAT-cl 40 SAY+THAT-cl 25 GO GOING+TO 69 GOING+TO 45 BE ABLE+TO 30 ABLE+TO 28 THINK THINK+THAT-cl 86 THINK+THAT-cl 18 LOOK LOOK+AT 121 LOOK+AT 36 KNOW KNOW+Wh-word 39 DON T+KNOW 64 (DIDN T+KNOW 6) KNOW+Wh-word 40 DON T+KNOW 32 (DIDN T+KNOW 25) SEE TO+SEE+Pron 45 COULD+SEE 12 COULD+SEE 24 COME HAD+COME 14 HAD+COME 19 CAME+TO 26 GET TO+GET 24 (none) MAKE TO+MAKE 34 TO+MAKE 42 LIKE WOULD+LIKE 39 WOULD+LIKE 5 The patterns of verb collocation found frequent in both data were (1) auxiliary+verb: DON T+KNOW, COULD+SEE and WOULD+LIKE, (2) semi-modals: HAVE+TO, BE+ GOING+TO and BE+ABLE+TO, (3) those that constitute compliment structures: SAY and THINK with THAT-cl, and KNOW with Wh-word, and (4) other collocations: LOOK+AT and TO+MAKE. While these were very common both in 6GR and 113NYT data, TO+SEE+Pron, TO+GET and CAME+TO seem less so. The former two collocations were found frequent only in 6GR, and the latter one frequent only in 113NYT. Among graded grammar charts devised by the three publishers of English readers, Penguin Readers Grading (PRG) is the most thorough and complete. Most of the dominant collocation patterns in Table 4 were found in the first three grading levels: GO+TO and WOULD+LIKE+Noun in Level 1 (Beginner 300 words), HAVE+TO, THINK+THAT-cl, LOOK+AT, KNOW+THAT-cl and WOULD+LIKE+TO+Verb in Level 2 (Elementary 600 words), and BE+ABLE+TO in Level 3 (Pre-intermediate 1200 words). However, no syntactic structures described in PRG levels 4 to 6 were found equally as frequent as these. Therefore, it remains uncertain how many graded readers ESL/EFL learners should read to get accustomed to other lexical and syntactic patterns listed in the latter half of the Penguin chart. 6. Concluding remarks A detailed examination and analysis of the WordSmith data reveals there are many similarities between the graded English texts (6GR) and the authentic texts (113NYT) so far as their distributions of types and tokens are concerned. Also found similar were many verb collocations. Yet, the corpus data used for this research were not large enough to extract all PRG major lexical and/or syntactic patterns. 343

357 ASIALEX 2005 References Biber, D; Conrad, S. and Leech, G. (2002). Longman student grammar of spoken and written English. Harlow: Pearson Education Limited. Cambridge University Press Grammatical grading & level chart, available online. (accessed: 5 March 2005) (accessed: 5 March 2005) JACET Committee of Revising Basic Words ed. (2003), JACET List of 8000 Basic Words (Tokyo: The Japan Association of College English Teachers) Oxford University Press Bookworm Syllabus, available online. (accessed: 5 March 2005) Penguin readers grading, available online. (accessed: 5 March 2005) Scott, Mike (2004) WordSmith tools version 4. available online. (accessed: 5 March 2005) Appendix: Six graded readers Gibbons, Stella retold by West, Clare (1998), Cold comfort farm (Oxford: Oxford University Press). Hancock, Penny (2000), A love for life (Cambridge: Cambridge University Press). Huxley, Aldous retold by Cartledge, H. A. (1999), Brave new world (Essex: Pearson Education Ltd.). Maley, Alan (1999), He knows too much (Cambridge: Cambridge University Press). Rodat, Robert retold by Kehl, Jacqueline (2000), Saving Private Ryan (Essex: Pearson Education Ltd.). Waugh, Evelyn retold by West, Clare (1998), Decline and fall (Oxford: Oxford University Press). Abstract Genesis and function of words with astronomical and religious implications in Indus Valley civilization N.S.Valluvan nsvalluvanart@yahoo.co.in The syllables AH, MU, VAN define the primordial god ahmuvan, who also functions as a cosmic man in terms of identity construction for a natural man from pre-history to the present and ahmuvan provides the link to the process of cultural continuity from pre-history to the present and parallels murukan, who is the god of Tamil language. The word ahmuvan defines the micro-macro correspondence in terms of art, myth, astronomy and architecture in Indian cultural and religious context and ahmuvan as a lord of space and time has played a key role in ritual chronometry. The festivals that are being conducted for god murukan in Asia justify his relationship to ahmuvan and the words AHMUVAN and MURUKAN have stood as a definition of religious and cultural symbolism in time and space over several millenniums from Indus valley to Tamil nadu in southern India. 344

358 WORDS IN ASIAN CULTURAL CONTEXTS AHMUVAN 1. AHMUVAN AS SEEN IN AN INDUS TABLET 2. AHMUVAN AS WRITTEN IN THE INDUS TABLET The Indus tablet (G.R.Hunter, 1934) shows an elongated anthropomorphic figure (No:1) with a trident shaped protuberance in the head surrounded by a loop from head to toe from both sides of the figure and pipal tree leaves attached to 13 smaller loops that come out from the bigger loop. The reverse side of the tablet is inscribed with the Indus sign sequence No: (No:2)* and comprises the signs for the Constellation Draco, Lord of space and time and Sky, which can be interpreted as THE CONSTELLATION DRACO IS ESTABLISHED BY THE LORD OF SPACE AND TIME IN THE SKY and can be read from right to left as AHMUVAN. This Deity finds his parallel with the God MURUKAN of the Tamil tradition of southern India. 1.1 AHMUVAN stood for the time period during which the constellation Draco held the key position in the northern night sky from around B. C. onwards. Draco and Orion are the partners in a great celestial SEE -SAW that is being performed over several thousands of years and Ahmuvan must have been worshipped and venerated as a primordial God of the Indus valley civilization. 1.2 The Draco sign is read as AH In B.C. Draco marked the north meridian when the Orion marked the south meridian (Graham Hancock, Santha Faiia, 1998) and Draco surrounded the pole of the ecliptic in 6000 B.C. It is circumpolar at higher altitudes and rotates slowly around the north pole of the sky. 2.1 The notion of unmoving pole star around 3000 B.C. refers to the ALPHA DRACONIS of the Draco constellation in Indus valley and the GAMMA DRACONIS of the same is named as the Zenith-star since it almost lies in the Zenith of Greenwich. The Draco had dominated the northern night sky during several thousands of years and must have influenced the minds of the ancients. They should have venerated and 345

359 ASIALEX 2005 worshipped it as a powerful deity with awe and wonder since it occupied a large area of the northern night sky Tamil Alphabet AH Tamil Vattezhuthu AH Indus Sign AH The Tamil alphabet which is read as AH metamorphosed from an earlier script, Tamil Vattezhuthu that is also read as AH has its origin from the Indus sign denoting the Draco constellation read as AH too (No:3). 2.3 The unmoving north pole star of the vedic times is the Alpha Draconis and the Greek God HERCULES displayed the Draco in his shield. Draco was associated with the Egyptian deity ISIS HATHOR and pictured as a cow and held a powerful position in myth and astronomy in China. So the constellation Draco had occupied an important position all over the ancient world. 2.4 AH also stands for the primal expression sounded by the human beings by opening the mouth naturally and may have inspired the way for linguistic communication and invention of language. 2.5 ALPHA DRACONIS was 0.6 degree away from the heavenly pole in 2780 B.C. and this period corresponds to the Indus valley civilization. Approximately the time period from B.C. to 2500 B.C. can be understood as the Draco phase of the Indus valley civilization and possibly the Ursa Major phase followed the Draco phase AHVAN The Indus sign sequence No: can be interpreted as THE CONSTELLATION DRACO IS IN THE SKY (No: 4)* which is read as AHVAN. And the sign of Draco is self-evident in the northern night sky. 2.7 The Kanaga sign is read as MU 5. Kanaga Sign Indus Sign Indus Rock Art Sign The structural relationship between the Kanaga sign, Indus sign and Indus rock 346

360 WORDS IN ASIAN CULTURAL CONTEXTS art sign (No:5) shows a remarkable resemblance. All these variations represent a symbol for space and time and a source of life energy, which is none other than the sun itself.3.1 The Kanaga sign is a religious symbol of the DOGON people of West Africa and is worn by the male members as masks during the funerary dance rituals to satisfy the spirit of a dead person. The Kanaga sign stands for the primordial energy and acts as a mediator between earth and heaven. In Indus rock art this Kanaga sign is identified as KADAVUL, the name for God in Tamil (S. Gurumurthy, 1999) and it is also a symbol of COSMIC PURUSHA, which is drawn in the exteriors of houses as an auspicious symbol in southern India. 3.2 The Indus anthropomorphic figure is represented by the Kanaga sign and can be interpreted as the symbol for a primordial God and a Lord of space and time with an astronomical meaning primarily denoting the three positions of the sun with its north, east, and south directions over the yearly motions. This sign may explain various time periods and also other celestial positions.3.3 The syllable MU, which forms the root of the word MURUKAN, is the name of the God for the Tamil language and is explained as a Lord of space and time. MU also forms the root for the Tamil word MOONDRU, which is the name for the numeral denoting three. The Lord Murukan is also called KANTHAVEL. 3.4 The root syllables MU, KA both refer to Murukan and Sun. The Egyptian word KA denotes the spirit or soul and the Kanaga sign represents the pivotal three positions occupied by the sun during the equinoxes and the solstices in the sky MUVAN The Indus sign sequence No: , can be interpreted as THE LORD OF THE SPACE AND TIME IS IN THE SKY (No: 6)* which is read as MUVAN. 3.6 The sky sign is read as VAN The U or inverted U is the symbol for the sky in Indian art symbolism (K. C. Aryan, 1981). When the U is affixed with four strokes in the top, it becomes the Indus sky sign. The four strokes stand for the four directions north, south, east and west. The U is read as VAN as the same denotes the sky in Tamil. 4.1 The primordial God of Indus valley AHMUVAN, stands inside a bigger loop embedded with 13 smaller loops with pipal tree leaves attached to it possibly denoting 13 time periods as found in the astronomical calculations of the Mayan. The bigger loop stands for the sky dome or the celestial arc of the god as found attached to the Indian Gods in temples. It is termed as THIRUVATCHI in the Tamil agama tradition. 4.2 The syllable AN denotes the God in ancient Sumerian culture and the suffix AN is common in naming male persons in the tamil tradition of southern India. Hence AHMUVAN can be called a deity of the sky and heaven. 4.3 * Numerous Indus signs are present in Southern India as rock art signs (Refer to I.Mahadevan, 1977 for Indus signs, sign sequences and S.Gurumurthi, 1999 for Indus rock art signs) and read the sign sequences in the order of right to left. 347

361 ASIALEX 2005 References Aryan, K. C., Basis of decorative element in Indian Art. (Rekha Prakashan) New Delhi. P. 16. Hunter, G. R., The script of Harappa and Mohenjo-Daro and its connection with other scripts. (Munshiram Manoharlal Publishers Ltd.) New Delhi. Plate. X X X X, No: 42. Mahadevan, Iravatham, Indus script texts, concordance and tables. (ASI, No: 77) New Delhi. Parpola, Asko, Deciphering the Indus script (Cambridge University Press) Cambridge. Gurumurthy, S., Deciphering the Indus script from graffiti on ancient Indian pottery. (University of Madras) Chennai. P Tilak, B. G., The Orion or researches into the antiquity of the Vedas. (Messers Tilak Bros) Poona. Plunket, Emmeline M., Ancient calendars and constellations. (John Murray) London. Roy, S. B., Prehistoric lunar astronomy B. C. (Institute of Chronology) New Delhi. Ghosh, Ekendranath, Studies on rigvedic deities astronomical and meteorological. (Cosmo Publications) New Delhi. Subbarayappa B. V., Sarma K. V., Indian astronomy a source book. (Nehru Centre) Bombay. Krupp, E. C., Echoes of the ancient skies the astronomy of lost civilizations. (Oxford University Press) New York. Aveni, Anthony, Stairways to the stars sky watching in three great ancient cultures. (John Wiley & Sons, Inc.) New York. Temple, Robert K. G., The Sirius mystery new scientific evidence for alien contacts 5000 years ago. (Century) London. Hancock, Graham, Faiia, Santha, Heaven s mirror quest for the lost civilization. (Three Rivers Press) New York. P Ashfaque, Syed M., Primitive astronomy in the Indus civilization. (Wisconsin Archaeological Reports, Volume 2, No: 2) Wisconsin. Sidharth. B. G., The celestial key to the vedas discovering the origins of the world s oldest civilization. (Inner Traditions) Rochester, Vermont. Heras, Henry, Studies in proto Indo Mediterranean culture, Vol: 1. (Studies in Indian History of the Indian Historical research Institute, St. Xaviers College, No: 19) Bombay. Wheeler, Sir Mortimer, The Indus civilization. (The Syndicate of the Cambridge University Press) London. Zvelebil, Kamil V., Tiru Murugan. (International Institute of Tamil Studies) Madras. Zvelebil, Kamil V., Tamil traditions on Subramanya Murugan. (Institute of Asian Studies) Madras. Clothey, Fred W., The many faces of Murukan - The history and meaning of a South Indian God. (Mouton Publishers) The Hague, The Netherlands. Clothey, Fred W., Rhythm and intent ritual studies from South India. (Blacki & Son Publishers Pvt. Ltd.) Bombay. Frawley, David, Gods, sages and kings vedic secrets of ancient civilization. (Motilal Banaridas Publishers) Delhi. Ganapathi Sthapati, V., Vaastu purusha mandala. (Dakshina Publishing House) Chennai. Snod Grass, Adrian, Architecture, time and eternity, Volume 1, 2. (Aditya Prakashan) Delhi. 348

362 WORDS IN ASIAN CULTURAL CONTEXTS Occidentalism in Singapore s Anglophone fiction: Victorian wor(l)ds lost in C.M. Woon s The Advocate s Devil Tamara S. Wagner Nanyang Technological University, Singapore TSWagner@ntu.edu.sg Abstract Responding to the need for a reassessment of the transformations that (neo-) orientalist and occidentalist clichés undergo in recent Singaporean novels, this paper proposes a new approach to the literary significance of these developments. Taking C.M. Woon s The Advocate s Devil (2002) as an illustrative example, it shows that Singaporean novels are particularly interested in the shifting ideologies that frame Anglophone, or English-language, writing. Yet, Woon s novel does significantly more than that: as a historical novel, it thrives on its revaluation of English, showing its function in colonial Singapore to be much more complicated than conventional postcolonial critiques would like it to be. Peranakans, Eurasians, local enclaves of various minority groups, and their languages feed on and into a diverse pool that is cunningly revealed as a source of crime stories echoing with the voice of Somerset Maugham and which also exposes stratifications of ethnicities as deeply conflicted. The main protagonist finds himself lost for words and his world lost when confronted with a girlfriend from China spouting pan-chinese ideologies. As it examines the novel s twofold engagement with the currents of Singapore s fluid languages and literature(s), this rereading suggests the need for a new theory of occidentalism that helps us to reassess such re-representations of literary and other traditions. Literary intertextuality the body in question What is perhaps most striking about C.M. Woon s first novel The Advocate s Devil, published in 2002 in Singapore, is its style: the ways in which it stands out among recent locally published novels about Singapore, past or present. It is not only that the stories that comprise the book are episodic in nature, picaresque almost, even as they follow the, albeit somewhat static, central character from his return to colonial Singapore as an estranged homeland to his self-conscious awareness of the end, and ends, of colonialism, but that they redeploy the language as well as the plots, popular types, and, with a certain self-irony, also the clichés of the nineteenth-century classic English novel. By no means limited to intertextual references, this reworking engenders a tongue-in-cheek replication of a colonial exotic that draws specifically on William Somerset Maugham s so-called Eastern stories. This does not mean that Woon simply imitates or straightforwardly dismantles Maugham and the colonial world of Maugham s fiction, which has rather tautologically been termed quintessential Maugham, so much so that colonial Malaya is firmly established in the British popular imagination as Maugham country (Holden, 1996, 95). But neither does the retrospectively much exoticised world of Woon s works serve only as a backdrop. Self-irony runs through the narrative, and the play with language makes it an intriguing reworking of literary traditions, even as the breakdown of the traditional novel genre 349

363 ASIALEX 2005 looms over the seemingly tranquil colonial city as a metaphorical vehicle for a much larger breakage. Orientalist clichés are therefore not so much duplicated as cunningly redeployed, juxtaposed with their occidentalist counterparts, to make both sets of discourses problematic, and most importantly, to turn them into a self-reflexive literary work. As it closely examines the novel s twofold engagement with the currents of Singapore s fluid languages and literature(s), the following reading of Woon s so far rarely acknowledged work of fiction seeks to illuminate its intriguing introduction of both the colonial, anecdotal, story and the genre of the detective novel into the postcolonial Singaporean novel. The body in question that figures as the title of the first chapter (or episode) strikingly accentuates this venture into crime fiction as a foray into new fields, away from a mere postcolonial or nationalist revision of either the colonial past or its legacies for post-independence nation building. Surely now that these narratives have been written, and rewritten, repeatedly over the last decades, the fictionalisation of such and, more significantly, other new histories can be accomplished with more confidence, self-referentiality, self-irony even. Woon s novel indeed stands among a number of recent fictional works in Singapore that have embarked exactly on this project of literary reworking. As I have shown in some detail elsewhere, there is growing interest in thrillers, crime novels, and even science fiction stories with a local setting that serves as more than just local flavour (Wagner, forthcoming). Rex Shelley s use of the spy-novel, for example, shows how it can integrate themes of minority culture, hybridity, and diaspora in a political thriller set in the recent past, and I shall come back to his series of novels as a good parallel to Woon s writing on more levels than one. By thus setting The Advocate s Devil amidst recent literary developments in Singapore, this paper furthermore aims to highlight the significance of an ongoing self-referential use of occidentalism and orientalism as discourses not only to be dismantled, but to be, in the very process of this dismantling, made into fiction. As a provocative point-of-entry into the novel s creative reuse of colonial fiction, this reading starts out by pivoting on the two meanings, or sets of meaning, of body in the first story. A euphemism of the dead body that never materialises as the story turns out to be a comically demystified crime mystery, a detective story fascinatingly constructed on wrong clues, its title moreover points to the wrong body from the very beginning: Madeline. There s no mistake about it. That s her name in the papers all right. And that photo. (Woon, 2002, 9) Instead of a body, we have only got a name and a picture that set off memories of 1930s Singapore as a colonial world of exoticised crime stories. The first person narrator, Dennis Chiang can zip through its landscapes, passing by huts on stilts and [l]ittle brown figures (Woon, 2002, 16), imbibing an atmosphere that has become exotic to him, and indulging, with his customary self-irony, in conjuring up his own story: I had been precipitated in the midst of a murder-mystery, zipping around Singapore town with the legendary Clarence d Almeida, who even now was hot on the trail of the killer. Who needs fiction? (Woon, 2002, 32) But the body s desirability is, rather predictably perhaps, only on one level that of the alluring young woman: one of those fortunate alliances of East and West, a handsome delicate creature who inherited the best features from both her parents (Woon, 2002, 12), and as such a perfect embodiment of Singapore s mixed heritage and a mirror-version of Dennis s own cultural hybridity. Her missing husband s body, however, turns out not be the dead body desired by the lawyer s assistants playing detective either. Instead, he turns up disguised as a Malay cook, forced into cultural cross-dressing by colonial strictures that 350

364 WORDS IN ASIAN CULTURAL CONTEXTS are thus shown to be as limiting to the coloniser as to the colonised. In a sarcastic evocation of colonial clichés, the exposed tuan re-emerges under the cook s trappings; his use of imported jargon and local languages pointedly expressing his bitterness about ethnic demarcations: Do you think that jobs for an orang puteh grow on trees like coconuts? [ ] I tried selling things in the street. But some pompous ass of a Police Inspector rounded me up and gave me a lecture on the white man s burden. Not done for a tuan to go flogging stuff on the five-foot-way. What will the natives think? Can t let down the side. Silly old goat! (Woon, 2002, 40) Kipling s white man s burden clashes with local references and Malay words, and it is important to note that the novel, like the colonial anecdotal collection of stories with which it plays as a genre, includes a glossary of Malay and Chinese words. The image of the single white street-seller rounded up by one pompous inspector moreover mocks the ludicrous invocation of the white man s dignity, as it is put when, in an inversion of the tuan s cross-dressing, a Eurasian and a Peranakan pose as British officials to trick an Anglophile sultan s son out of his custody rights to his half-english daughter: He invited us into one of the adjoining reception rooms, but I thought that the white man s dignity would be better upheld if we stood around haughtily in the reception hall. (Woon, 2002, 88) Ethnicity as fluid, in fact, runs through the stories, linking them together, to swamp colonial and postcolonial constrictions in their fusions, or confusions, of race, culture, language, and elusive concepts of ethnic communities. In other words, the body in question in the opening chapter embodies the wrong leads that guide (or misguide) the would-be detectives, and the reader, through the story. Alec Russell, or Iskandar tellingly rather a grand name for a jaga stands exposed with an arm [of] that curious off-grey pink colour that we call white (Woon, 2002, 28) in The Body in Question. Similarly, in A Prince Among Men, Dennis Chiang and Ralph Smallwood successfully imitate caricatures of the colonial and the colonised emulative occidentalist, both complete with walrus moustache and spectacles or monocle, the very spitting image of a tuan besar and his faithful minion in appropriately eclectic surroundings, a style that one local architect dubbed Anglo-Marzipan : My fellow pupil Ralph Smallwood sat next to me, resplendent in white tutup jacket, trousers, shoes, gloves and solar topi. He had a monocle scrunched up in is right eye and a walrus moustache, the very caricature of an Imperial official. (Woon, 2002, 89, 80, 87) But this is of course more than just camouflage or caricature. Throughout the novel, ethnic and linguistic clues are exposed as misleading, and this is surely significant, especially as all of the main protagonists are cultural and, in colonial terms, racial hybrids. Alec s easy merging into local life is partly facilitated by his family s local roots, a settling down that is rumoured to have run more deeply than is publicly acknowledged: His family had been in the Straits Settlements for nearly half a century, and it was said (not very loudly in his hearing though) that his dark hair and dark eyes were a little local colour introduced as a result of a trifling dalliance on the part of old man Russell in his forgotten youth. (Woon, 2002, 13) The ambiguities of colonial conceptualisations of hybridity, in fact, become part of the detective stories. In pointed contrast to Alec Russell, Ralph Smallwood, for example, could never get into the Singapore Club or the Tanglin Club because of this, the nigger in the woodpile as they so eloquently put it. (Woon, 2002, 87) The ways in which Woon works colonial racism into the first person narrator s self-ironic, bitterly nostalgic, memories of colonial 351

365 ASIALEX 2005 Singapore importantly do not shirk the violence ingrained in the languages of racial slotting, while the nostalgic charm of the narrative arises from quite another hybridity: Ralph s grandmother was Chinese, though this wasn t noticeable at a casual glance. [ ] D Almeida, on the other hand, was quite clearly coloured; a sort of warm cocoa colour. In fact, in chambers we had the whole spectrum, from Ralph who could pass off quite easily as a tuan, all the way to George Singham, who when he grinned on a dark night disappeared completely, leaving his smile hanging like the Cheshire Cat s. (Woon, 2002, 87-88) Literary hybridity, intertextuality, the presence of Alice in Wonderland in the retrospective account of the former Straits Settlements, filtered through an, in a sense, post-postcolonial, self-ironically occidentalist, reworking of Maugham strikingly show that the body in question throughout the novel and the new historical novel of Singapore at large is also a body of words and a body of literature. Working on and through a colonial and a corresponding postcolonial exotic, to borrow from Graham Huggan (2001, passim), The Advocate s Devil can be seen to transform (neo-)orientalist and occidentalist clichés. It is indeed self-consciously interested in the shifting ideologies that frame contemporary Anglophone, or English-language, writing in Singapore. If it stands firmly in the tradition of the classic English novel, it also negotiates different ideas, and ideals, of tradition, including a tradition of occidentalist revision. As such, it does significantly more than simply replicating either emulative or revisionist occidentalism: as a historical novel, it thrives on its revaluation of the English language and its literature(s), showing its legacies to be much more complicated than conventional postcolonial critiques would perhaps like them to be. Words and worlds lost: Anglo-Marzipan constructions A self-reflexive revisioning of colonial histories, The Advocate s Devil is thus an intertextual crime story that seeks to find new terrain for such revisionist projects (and, as I have pointed out, with a certain self-irony), but it is moreover also part of a growing number of local novels that focus specifically on the region s minority communities. The Straits Chinese, or Peranakans, Eurasians, local enclaves of various minority groups, and their languages feed on and into a diverse pool that is cunningly revealed as a source of crime stories, just as their cultural/ethnic hybridity is used as misleading clues. As has similarly been remarked with reference to Rex Shelley s series of historical novels about the Eurasian communities of Southeast Asia, the described minority cultures become the main thrust as well as the theme of the story, not simply a background. In her suggestively titled article on Shelley s first novel, Rex Shelley s The Shrimp People: What Manner of Beast is it? Patricia Wong even maintains that the book may at first sight appear like a litany of names, for it is certainly replete with Portuguese-Eurasian, Dutch-Eurasian, Anglo-Eurasian names. The litany forms a registry of Eurasian society, primarily the Eurasian society in Singapore, but also to a much lesser extent in the novel the parallel communities in Malacca, Seremban and Penang. (Wong, 1998, 45) Peter Wicks has similarly focused on what he terms Eurasian Images of Singapore and their naming: Shelley comes from the numerically small Eurasian community, and it is the distinctive historical experience of this minority, also known colloquially as mesticos, serani, or gerago, that richly frames his fiction. (Wicks, 2002, 377) Wong, however, importantly distinguishes between accidental and incidental functions of ethnicity in the novel. The Shrimp People sets out to provide a 352

366 WORDS IN ASIAN CULTURAL CONTEXTS panoramic account of the Eurasian community, yet towards the middle of the story, it changes into a spy thriller, and as Wong argues convincingly, ethnicity is the reason the novel mutates (Wong, 1998, passim). Being Eurasian singles out the central character, Bertha Rodrigues, as an ideal spy in early post-war Singapore and, it dramatically turns out, a double-agent, committed to do all she could to prevent the formation of Malaysia [and] the merger of Singapore and Malaysia (Shelley, 1991, 370). But what is more and much more disturbing another Eurasian, Wee Andy, symbolically has to die to prevent exposure. Eurasians embody the meeting of East and West, yet there is always a certain ambiguity that harbours self-ironic disruptions of neat categorisations. In short, the celebration of the region s history and the role of the Eurasian community is much more than the tale [ ] woven around a small minority in Southeast Asia: mixtures of the East and West known as Eurasians the introduction promises (Shelley, 1991, 5). With his Peranakan first person narrator, Woon s The Advocate s Devil can be said to do something very similar for Singapore s Straits Chinese heritage, even as it pairs it with a strong emphasis on an equally prevalent Eurasian presence: My father s grandfather came to the Straits Settlements shortly after Raffles had conned the Sultan of Johore into ceding Singapore to the British. (Woon, 2002, 44) They became Babas, marrying, as Dennis proudly emphasises, Nonyas from established Peranakan families, and, he suggests with more ambiguity perhaps, were loyal to the British. The Queen s (later King s), or Straits, Chinese formed the Peranakan (Malay for locally born) culture unique to the region, which had originated in intermarriages between Chinese merchants and Malay or Burmese women over the centuries. They displayed an emulative and creative interest in British culture, and their absorption of the English language and literary tradition has in many ways shaped the role of literature in the region and the development of ongoing debates about Asian values and the Westernisation introduced by colonial enterprises. Their writing formed a crucial stepping-stone of English (or English-language) literature in the region. Their emulative occidentalism importantly makes problematic the revisionist occidentalism that runs through the novel as an ambiguously felt counter-narrative: Throughout all the years I had lived in England I had believed in the Empire and what it stood for. I was a Baba, a King s Chinese. One year back in the Colony had frayed my faith in the ideals of Empire. As I walked through the jungle I realised that no matter how good I was, I would never be good enough. Never would I be accepted as an equal by the whites in the Colony. That realisation left me dismayed and not a little confused emotionally. The only world I knew was that of the white man. The other world, the one to which Siew Chin and the Towkay belonged, was totally alien to me. I felt instinctively that I should always be a stranger there; I would never be accepted as an equal either. (Woon, 2002, 284) The complexities of the region s cultural (and literary) heritage are already under stress. Symptomatically fatherless, Dennis feels estranged from his distant and rather Victorian uncle (Woon, 2002, 23), although the Chinese-educated towkay from Malaya does not offer an alternative father figure either. On the contrary, he embodies a world that is totally alien just as his desirable daughter remains inscrutable and maddeningly unreadable (Woon, 2002, 250-1). Conversely, she regards him like some strange laboratory specimen. [ ] The concept of a non-chinese-speaking Chinese was evidently a contradiction in terms to her. (Woon, 2002, 192) So if the decline of the great Peranakan families links the novel to Victorian literary classics about genteel poverty (the words are Dennis s own) sometimes precipitated by a spendthrift son (a Baba black sheep, one might say) (Woon, 2002, 46) Dennis finds 353

367 ASIALEX 2005 himself lost for words and his world lost when confronted with a would-be girlfriend from China spouting pan-chinese ideologies. With a critique of contemporary language policies that link language to race in a problematic concept of the mother tongue or second first language in Singapore as an important undercurrent, the novel shows the sizeable presence of minority communities in the former Straits Settlements to be not only multilingual and primarily English-speaking, but well versed in European (literary) traditions. The lingua franca of the Colony, as Dennis puts it, may be bazaar Malay, or rather, in some cases bizarre Malay (Woon, 2002, 212), but the language that really works among the protagonists (as opposed to this working language) is English. Even more pointedly, presumably ethnically fitting languages are exposed to be as skin-deep as the racial camouflage that drives enticing detective plots. Dennis s brief desire to chase roots in order to impress the towkay is quickly exposed as a red herring: Aha, said George, Siew Chin s father. Now it all becomes clear. Cherchez la femme, as somebody or other once said. Who was it, now? Dumas, I think, I replied. Right, spot on, Dumas it was. Look for the lady. For the sake of a lady I d learn Swahili if necessary. (Woon, 2002, 263) Perhaps wise d Almeida has the last words after all when he warns Dennis against joining the towkay and conforming to his values: A man s worth isn t determined by the language that he speaks. (Woon, 2002, 265) References Holden, Philip (1996), Orienting masculinity, orienting nation: W. Somerset Maugham s exotic fiction (Westport: Greenwood Press). Huggan, Graham (2001), The Post-Colonial Exotic: Marketing the Margins (London and New York: Routledge). Shelley, Rex (1991), The shrimp people (Singapore: Times Books International). Wagner, Tamara S. (forthcoming) Singapore s new thrillers: boldly going beyond the ethnographic map, ARIEL: a review of international English literature. Wicks, Peter (2002), Eurasian images of Singapore in the fiction of Rex Shelley, in Mohammad A. Quayum and Peter Wicks (eds), Singaporean literature in English: a critical reader (Serdang: Universiti Putra Malaysia Press), pp Wong, Patricia (1998), Rex Shelley s The Shrimp People: what manner of beast is it? in Kirpal Singh (ed), Interlogue: studies in Singaporean literature: fiction (Singapore: Ethos Books), pp Woon, C.M. (2002) The advocate s devil (Singapore: Times Books International). Multilingualism and the alienation effect in Kuo Pao Kun s theatre Abstract Shengyu Wang National University of Singapore wsychina@263.net This paper discusses the dramatic effect of Kuo Pao Kun s multilingual plays with reference to Brecht and John Gay. It argues that Kuo s theatre, rooted in the multiracial and multilingual context of Singapore, facilitates the Brechtian alienation effect on one 354

368 WORDS IN ASIAN CULTURAL CONTEXTS hand and extends it to multiple dimensions on the other. Kuo Pao Kun ( ), doyen of the twentieth century Singapore theatre, represents postmodern cultural multiplicity with traceable inheritance of Bertolt Brecht ( ). Archives show that Kuo became acquainted with Brechtian theories when he studied in the National Institute of Dramatic Art (NIDA) of Australia in the early 1960 s, and he subsequently translated into Chinese and directed The Caucasian Chalk Circle. 1 His first full-length play, Hey, Wake up! (1968), resembles Brechtian theatre in form. 2 The majority of Kuo s plays created from 1984 onwards are highly experimental and can be termed as, in the Brechtian definition, Non-Aristotelian Drama, in which mimesis is not the means of representation, nor is catharsis the goal for the audience to attain. 3 What Kuo seems to strive for on the stage parallels Bertolt Brecht s theories of the epic theatre as against the traditional dramatic theatre. Corresponding to the elimination of the fourth wall, Kuo adopts Brechtian stage set-up, for instance, the three-sided stage in Mama Looking for Her Cat (1988), the house under construction in Lao Jiu (1993), and the scaffold in Geylang People in the Net (1997). Instead of presenting a linear development of plot, Kuo offers narratives in montage, curves and jumps, 4 as in Mama and Lao Jiu. When viewing such plays, one gets the feeling of being turned into a critical reader, an observer who is made to face something and to take decisions. 5 Similar to Brecht s use of music, Kuo incorporates a great range of symbolic physical movements mimicking those in traditional Chinese opera and martial arts. Play within a play, puppet show and masks are also among the assets of Brechtian theatre employed by Kuo. Such features alienate the spectator from the incidents in the play and work toward the Brechtian concept of Verfremdungseffekt, which can be roughly translated into English as the alienation effect (hereafter referred to as A-effect). To reiterate, many elements in Kuo s theatre create the A-effect in the audience. Based upon the multiracial and multilingual context of Singapore, Kuo Pao Kun, nonetheless, departs from the Brechtian devices of achieving the A-effect by taking advantage of the linguistic variety in the city-state. As a result, he extends the A-effect to multiaspect. Advocating an open culture as a way out for Singaporeans, who are cultural orphans segmented from their parental cultures, 6 Kuo epitomizes his ideals in his multilingual plays. Mama, OZeroO1 (1991), Geylang People in the Net and Sunset Rise (1999) are particularly produced as multilingual, involving speeches in English, Mandarin Chinese, Malay, Tamil and such Chinese dialects as Hokkein, Cantonese and Teochew. The fact that different languages are spoken on the stage either with or without mutual comprehensibility among the speakers brings in more facets to the A-effect. On the audience s side, people of different ethnic identity watching the performance further complicates the issue. In this paper, I will explore how Kuo develops and extends the A-effect in his multilingual theatre in the postcolonial and postmodern context of Singapore from three viewpoints, namely, social critique, generic contradiction and intercultural communication. Multilingualism as a means of social critique Kuo Pao Kun s multilingual plays reflect his social critique by the engagement of different languages. Consistent in tone with his other theatrical works, they are all concerned with the state of being of Singaporeans positioned in a unique social and 355

369 ASIALEX 2005 cultural context. Mama talks about the generation gap between an uneducated mother and her well-nurtured children, and the conflict is represented through the collision of languages. Throughout the play Mama speaks nothing but Hokkein and her dialect gradually alienates herself from her children, who have given up their Hokkein for English and Mandarin. The postcolonial issue of loss and memory of one s mother tongue is a focal point in the play. The language one speaks may be responsible for shaping one s worldview and therefore constructing the self. As Ngugi points out, through language people not only describe the world, but also understand themselves. Besides, language and culture are inseparable, and therefore the loss of the former results in the loss of the latter. 7 Mama s Hokkein is not just a language being suppressed it marks the marginalization of the culture belonging to the early Chinese immigrants. A view of the stage version of Mama impresses the audience that Mama, speaking an old dialect and most of the time crawling on the floor groping about, is both confusing and confused. The confusion of Mama incarnates the dilemma of traditional culture. On the other hand, her children s pursuit of English and Mandarin, the two dominant languages of political and economic power, 8 embody their submission to the trend of personal and social development. What Kuo seems to critique is that this kind of development is often at the sacrifice of moral values, for instance, filial piety. Unlike Kopi Diam (1986), another of Kuo s play presenting clashes between older and younger generations, Mama achieves the end of social critique via collision of various languages spoken by the dramatic personae. Other multilingual plays by Kuo such as OZeroO1 and Geylang People in the Net address social and cultural issues in a similar way. Structurally, in Mama there is a developmental line indicating the gradual process in which the younger generation detach themselves from their mother language and consequently mother culture. Evidently all the devices employed are concerned with language. In the beginning scene, Mama s children play games, sing lullaby, and enjoy the Hare and Tortoise tale all in Hokkein; later on, they are presented as shifting their linguistic priority to English and Mandarin: they compete in tongue twisters such as Peter Piper and shí hé sì, play word translation games and write letters in the two languages. In the translation game, they are said in the stage direction to know both the English and Mandarin terms but fail to know the proper Hokkein rendition. (Original Italics) 9 In the Mandarin version of Mama, when the children are writing letters to their siblings studying overseas, Kuo has the direction that they are interpreting the content of the letters between English, Mandarin and Hokkein, the last case of which is just for Mama, but every one of them is having trouble with Hokkein. As a result, there is either a wrong version or distorted version. (My English translation) 10 Here the status of Hokkein as perceived by the younger generation is revealed in a telling manner. These instances clearly show Kuo s awareness of the dying out of a dialect as symbolizing that of a tradition. The suspicion of the prevailing social and cultural trend in Kuo s multilingual theatre finds its predecessor in Brecht, who is a successor of John Gay ( ). Both The Beggar s Opera (1728) and Die Dreigroschenoper (The Threepenny Opera, 1928) imply political resistance, the former against mercantilism and the latter upcoming Nazi influence. Dianne Dugaw observes that in Gay The punning language, self-conscious generic contradictions, and mimicking intertextualities expose the extent to which currency and interest define this world, 356

370 WORDS IN ASIAN CULTURAL CONTEXTS whose corrupt public policies and institutions troublingly intersect with and parody the affectional relationships and private life. 11 What Dugaw says does not fail to do justice to Gay s plays. In Scene XIII, Act I of The Beggar s Opera Macheath s pledge of love for Polly may illustrate Dugaw s observation: Is there any Power, any Force that could tear me from thee? You might sooner tear a Pension out of the Hands of a Courtier, a Fee from a Lawyer, a pretty Woman from a Looking-glass, or any Woman from Quadrille. But to tear me from thee is impossible! 12 The mock-heroic, trivializing use of parallelism subverts the discourse itself. None of the analogies drawn here are of a reliable nature, and therefore the pledge just becomes flubdub. John Richardson also comments that The Beggar s Opera achieves its resistance by manipulating genre, skewing language, and defying expectation. 13 In an age when the audience s generic expectation is defined, The Beggar s Opera stands out as an experiment, and the linguistic expressions in it are highly referential. One metaphor from The Threepenny Opera may be appropriate to denote the deviation of Gayian and Brechtian theatre, where the A-effect is expected to take place, from the Aristotelian: See the shark with teeth like razors, All can read his open face. And Macheath has got a knife, but Not in such an obvious place. 14 The passage can imply the subtlety of theatrical form in Gay and Brecht, in which nothing is self-evident. Form is manipulated to reveal the content, language being a chief vehicle in such a practice. Kuo seems to have inherited the emphasis on language from the two playwrights, yet he works with several languages instead of one. Plays of social critique written by the three of them all depart from what Brecht terms as einfuehlung (empathy) 15 and alienate the audience into serious pondering. Multilingualism and generic contradiction According to Dianne Dugaw, The Beggar s Opera is a play intent on undermining the generic expectations of tragedy and comedy. 16 In the same tradition, Kuo s multilingual plays integrate tragic and comic elements in an intricate manner. To a large extent, Mama, Geylang People in the Net and Sunset Rise are not comic, since thematically they touch upon grave issues like clashes between old and new cultures, mercantilism in the age of information and old age problem. However, they consistently provoke laughter when being staged. The use of various languages contributes significantly to the comic effects. The well-praised scene in which Mama and the old Tamil man talk about their respective missing cat in Hokkein and Tamil can serve as a case in point. Despite the absence of comprehensibility between the two persons languages, they are shown to make each other understood by speaking, gesturing and mimicking cats. It is true that their exaggerating actions are fun to watch, but most possibly the laughter comes form those among the audience who understand either Hokkein or Tamil or both. To hear a familiar language spoken to one who does not speak it can create comic effects in a serious situation. Such comic effects may correspond to what Brecht describes as to disrupt the gaze and open the symbolic eye of the spectator. 16 At this point occurs the A-effect, which works in multi-dimension in Kuo s theatre: the audience understanding different portion of the languages being 357

371 ASIALEX 2005 spoken experience different degrees of alienation. On the other hand, those who capture nothing from the conversation may be totally confused. In his two later multilingual plays, Geylang People in the Net and Sunset Rise, Kuo seems to make a conscious effort to reduce confusion by interweaving translation in the communication. Irony appears to be an effective means to intrigue generic contradiction in Kuo s theatre, as it does in John Gay and Brecht. In Geylang People in the Net, a nostalgic memoir of the multiracial heritage in the area, Kuo deploys ironies among all the scenes with alternating narration mainly in English, Mandarin and Malay. For instance, a Geylang prostitute, Judy, mumbles about her trade in a monologue like: We very clean. You very welcome. Our sitting room clean, corridor clean. My room small but also very clean. And we very, very clean because we wash after every serving My boss say very nice, he say we must offer our services with our unique spirit of excellence clean, courteous and comprehensive. The 3Cs of Geylang S! S for what? S for sex lor! 17 The passage goes against people s expectation of a narrative about a prostitute s life, for which complaint of suffering is the norm. The seemingly comic complacency and self-elevation expressed in substandard English may be carrying a tragic note. However, the quotation of jargons offer our services with our unique spirit of excellence and 3Cs against a background of base language definitely invites a good laugh from the audience. Not altogether coincidentally, the following scene in The Beggar s Opera also plays with the idea of register in the sociolinguistic sense: Lucy: If you are determin d, Madam, to raise a Disturbance in the Prison, I shall be oblig d to send for the Turnkey to show you the Door, I am sorry, Madam, you force me to be so ill-bred. Polly: Give me leave to tell you, Madam; These forward Airs don t become you in the least, Madam. And my Duty, Madam, obliges me to stay with my Husband, Madam. 18 In their grapple for the so-called love of Macheath, Lucy and Polly, the two women who are clearly not tailored for the way in which they are talking, sound buffoonish in a presumably trying situation. Their repeated use of Madam for each other is particularly comic. Similarly, in The Good Woman of Setzuan, Brecht makes the heroine, Shen Te, express her exultation after she falls in love with Yang Sun as: How wonderful to see Setzuan in the early morning!... In the early morning, the old city looks like a heap of rubbish! Nice, though, with all its little lights. And the sky, so pink, so transparent, before the dust comes and muddies it! What a lot you miss if you never see your city rising from its slumbers like an honest old craftsman pumping his lungs full of air and reaching for his tools, as the poet says! 19 The lyric tone here may not be appropriate for a woman who has ever been compelled to sell herself for a living. The paradoxical images of a wonderful morning and a heap of rubbish and pink, transparent sky with muddy dust undermine Shen Te s superficially happy mood. What can be inferred from the parallel ironies quoted above is that although Kuo, Gay and Brecht manipulate language at different scales, they all challenge the audience s generic expectations. Certain distorted understanding of expressions in one language also creates the generic contradiction. For example, in persuading people to accept the idea of building The Great Geylang Sex and Food Megamall, John, a real estate developer, quotes Confucius saying Shí, sè, xìng yě, which roughly means Food and sex are human nature, and thus arouses various distorted translations of it: 358

372 WORDS IN ASIAN CULTURAL CONTEXTS Pat: Oh! Food! Colour! Sex! Also! Lian: Shí, sè, xìng yě means Food and colour make sex wild. Judy: No. Shí, sè, xìng yě means If you eat food with artificial colour, you become very wild. Dick: No, no, no. Shí, sè, xìng yě means If you eat food with artificial colour you become sexually very wild. Tian: No, no, no. Shí, sè, xìng yě means Shí, sè, xìng yě. You see this man Shise, his surname Ye, so Shí, sè, xìng yě. 20 Through the game of language play the situation turns to be comic. However, to a reasonably well-educated Mandarin speaker, all the nonsensical contribution to the conversation may appear deplorable since it obliquely reflects the deteriorating status of Mandarin in the country. It could go so far as to turn for the tragic. Yet for a non-mandarin speaker, what is laboriously presented in the lines hardly makes any difference. In short, multilingualism in Kuo s theatre subverts the audience s generic expectations and thus complicates the A-effect. Multilingualism and open culture Multilingual play is a unique form of theatre developed out of the special social and cultural context of Singapore, where different races co-exist under the ideology of multiculturalism. Facing the impact of globalization and information technology, Kuo sees the urge to progress beyond multiculturalism to an open culture, which contemplates a transcendence of the individual from race and tradition-bound communities to embrace a diverse global community. 21 His ideal of transcending ethnic boundaries culminates in Sunset Rise, a multilingual play set in an old folks home with Chinese, Malay and Indian personae not only living in racial harmony, but also integrating themselves into each others culture in depth. Compared with his other plays using different languages, Sunset Rise allows the largest ratio of comprehensibility. This is owing to a close, if not line by line, translation of dialogues performed by the dramatic personae who are designed to be at least bilingual. Such a phenomenon may indicate Kuo s enhanced awareness of the key role that translation plays in the strife for the open culture. It is well acknowledged that translation not only means the transmission of language but of culture. Eugene A. Nida and Peter Newmark have elaborated on the importance of translation in cross-cultural communication. 22 In a recent paper, Newmark points out that no global communication would be possible without translation. 23 Translation can help transcend barriers and bridge cultures up. In Scene 3 of Sunset Rise, two former teachers residing in the old folks home, Bala and Lao Zhang, come to an exultant status of mind when they realize that both can recite the To be or not to be soliloquy in Hamlet: Bala: To be, or not to be that is the question. Whether tis nobler in the mind to suffer The slings and arrows of outrageous fortune Or take arms against the sea of troubles, And by opposing end them? Zhang: Shēngcún háishì huĭmiè, zhè zhēnshì yígè zhídé sīkăo de wèntí. Mòrán rěnshòu qīrén mìngyùn de jiànshāng qiāngtiăo, Háishì báqĭdāo, xiàngzhe wúbiān dàhăibānde mónàn bόdòuqù, yìliăobăiliăo! 24 The two speakers, one Indian who used to teach English literature, the other Chinese who also taught, find their common ground in Shakespeare. Their respective 359

373 ASIALEX 2005 recitation of the English and Mandarin version of the same passage is seen as a force to draw them to each other. Thus comes the stage direction The two are ecstatic, they shake hands. They start reciting simultaneously. Upon finishing, they embrace. (Original italics) 25 Viewed from another angle, an Indian man reciting Shakespeare in English and Chinese man doing it in Mandarin may embody the postmodern perception of cultural multiplicity. Hamlet re-implanted in such a situation could in a certain sense deconstruct the canonical play and therefore alter the power relation between English, its varieties and other languages. To revert to the central issue of this paper, Kuo s multilingual plays with obvious effort in putting meanings across ethnicity via translation, for example, Geylang People in the Net and Sunset Rise, opens up a new dimension of the A-effect. With the dramatic personae removing linguistic barriers themselves, verbal communication is made easy. Non-naturalistic acting such as shouting, repeating and exaggerating gestures are reduced, so the plays may appear less comic and as a result, with a weaker A-effect. However, from the point of view of the audience, watching the plays becomes a more self-conscious process, for they need to pay some extra effort to the deciphering of the translation, which at least involves two languages. In other words, the audience need to be more mentally engaged in order to follow the rendition. For Singaporeans, who are mostly bilingual, it is only natural for them to compare the source and target languages and form their own evaluation of the translated version. In this case, the A-effect, instead of being reduced, is enhanced. In conclusion, from social, linguistic and cultural perspectives, Kuo Pao Kun s multilingual plays can be rendered as to be able to facilitate the A-effect, however, in a multiplied sense. Notes: I would like to thank Theatre Practice and Theatre Works for their courtesy to allow me to view the stage versions of Kuo Pao Kun s plays, Theatre Practice for the generosity to provide me the two unpublished scripts, and Professor John Richardson of ELL, NUS for his perceptive and gracious supervision. 1. C.J.W.L.Wee and Lee Chee Keng, Breaking Through Walls and Visioning Beyond Kuo Pao Kun Beyond the Margins,Introduction to Two Plays by Kuo Pao Kun, eds. C.J.W.L.Wee and Lee Chee Keng (Singapore: SNP Editions, 2003 ), P Two Plays by Kuo Pao Kun, P Elizabeth Wright, Postmodern Brecht: A Re-Presentation (London and New York: Routledge, 1989 ), P Bertolt Brecht, The Modern Theatre is the Epic Theatre, in Modern Theories of Drama : a Selection of Writings on Drama and Theatre, , ed. George W. Brandt (Oxford : Clarendon Press ; New York : Oxford University Press, 1998) P Ibid 6. Kuo Pao Kun, Contemplating an Open Culture: Transcending Multiracialism, in Singapore: Re-engineering Success, eds. Arun Mahizhnan and Lee Teo Yum (Singapore: IPS, 1998), P Ngugi wa Thiongo, Decolonising the Mind: The Politics of Language in African Literature (Portsmouth, NH: Heinemann, 1981) P In his Contemplating an Open Culture: Transcending Multiracialism, Kuo observes that the Singaporean Chinese, who are 75 percent of the population, can be split between the Chinese-educated who are economically stronger, and the English-educated who hold the political power. See Singapore: Re-engineering Success, P Kuo Pao Kun, Mama Looking for Her Cat, in Images at the Margins: A Collection of Kuo Pao Kun s Plays (Singapore : Times Books International, 2000 ) P Kuo Pao Kun, Xun Zhao Xiao Mao de Mama, in Bian Yuan Yi Xiang: Kuo Pao Kun Xi Ju Zuo Pin 360

374 WORDS IN ASIAN CULTURAL CONTEXTS Ji ( ). (Singapore: shi bao chu ban she,1995) P Dianne Dugaw, Deep Play : John Gay and the Invention of Modernity (Newark: University of Delaware Press, London: Associated University Press, 2001), P John Gay, The Beggar s Opera, in John Gay: Dramatic Works, Volume II, ed. John Fuller (Oxford: Clarendon Press, 1983), P John Richardson, John Gay, The Beggar s Opera and Forms of Resistance, in Eighteenth-Century Life, 24:3 (2000): Deep Play : John Gay and the Invention of Modernity, P Brecht Sourcebook, eds. Carol Martin and Henry Bial (London and New York: Routledge, 2000) P Postmodern Brecht: A Re-Presentation, P Kuo Pao Kun, Geylang People in the Net, unpublished script, Scene John Gay: Dramatic Works, Volume II, P Bertolt Brecht, The Good Woman of Setzuan, in Parables for the Theatre: The Good Woman of Setzuan and The Caucasian Chalk Circle, trans. & ed. Eric Bently (London: Penguin, 1965) P Kuo Pao Kun, Geylang People in the Net, unpublished script, Scene Singapore: Re-engineering Success, P The understanding that translation means cross-cultural communication can be found in Eugene A. Nida s Language, Culture and Translating (Shanghai: Shanghai Foreign Languages Education Press, 1993) and Peter Newmark s A textbook of Translation (New York : Prentice-Hall International,1988) 23. Peter Newmark, No Global Communication Without Translation, in Translation Today: Trends and Perspectives, eds. Gunilla Anderman and Margaret Rogers (Clevedon,Buffalo,N.Y.: Multilingual Matters,2003) P Kuo Pao Kun, Sunset Rise, unpublished script, Scene Ibid Relational network notation and the intelligent web Jonathan J Webster City University of Hong Kong ctjjw@cityu.edu.hk Ian C Chow City University of Hong Kong ian.c.chow@student.cityu.edu.hk Abstract In this paper, we discuss the implementation of a network form of representation of linguistic information based on Relational Network Theory called Relational Network Notation. Frame Logic (FLogic), an ontology modeling language, together with Ontobroker, a database engine, are used to implement RNN. The ontology engineering environment also includes an inferencing capability which may be used to extend the knowledge base. The simplicity of Relational Network Notation plus its applicability to handling all kinds of information suggests it may have a role to play in improving access to information on the web. Relational Network Theory Elaborating on the nature of language as an information system, Lamb [4] describes it as a network of interrelationships. This network is composed only of nodes and their connections. Lamb makes no distinction between linguistic information and non-linguistic information; all information is stored and processed the same way, as connections in a vast network. Information processing consists of (a) the transmission of activation along pathways defined by the network, and (b) changes 361

375 ASIALEX 2005 in connection strengths. Mental networks (made up of nections) form the basis for our representation of reality. The model proposed by the theory is constantly being tested against the findings of neuroscience [6]. Lamb s [5] Relational Network Theory provides the conceptual basis for implementing an interface ontology extending from the lexicogrammar as ground to meaning and context above. The boundary between lexicogrammar (wordings) and semantics (meanings) is a stratal one. Interface ontologies maintain this stratal distinction between conceptual and lexicogrammatical information while providing the means for mapping between them an abstract semantic organization underlying our use of grammar and lexis that is motivated on essentially linguistic grounds and that acts as a complex interface between lexicogrammatical resources and higher-level strata in the linguistic system [1]. On the one hand, the conceptual or sememic level provides the motivational covering or context for each choice that the grammar provides, while on the other hand, the lexicogrammar serves as a resource for both understanding and articulating semiotic constructs at higher strata of meaning and context. Relational Network Notation The focus of this paper is on showing how relational networks can be represented using Relational Network Notation (RNN). RNN is based on Relational Network Theory, and offers a simple yet powerful means for representing lexicogrammatical, semantic and sememic information. Figure 1 illustrates the network diagram for the lexeme father, in which there are multiple sememes represented by the same lexeme father. In other words, father can refer to the male parent or to a clergyman in the Roman Catholic or Anglican churches. This network diagram illustrates an OR node extending from a single lexemic unit upwards to several sememic units. Conversely, in the case of synonymy (e.g. father-daddy, big-large, hard-difficult) there is more than one lexeme connected to a single sememe via another node. Ontology Modeling Figure 1. Relational network diagram for father We are using an ontology modeling language called Frame Logic (FLogic), along with various tools which together comprise a kind of ontology engineering environment to facilitate implementation of relational networks. FLogic is a database 362

376 WORDS IN ASIAN CULTURAL CONTEXTS logic which accounts in a clear and declarative fashion for most of the structural aspects of object-oriented and frame-based languages, including such features as object identity, complex objects, inheritance, methods, etc. [3]. The first step in designing an ontology involves conceptualizing the problem domain in terms of concepts and relations. A node, the most basic concept, points to other nodes. The arc(s) between nodes is/are described by the method to@, which takes up to three parameters: a threshold value, an identifier, and a sequence. The threshold value is the number of arcs pointing from the node (where 1 = OR; > 1 = AND). The identifier is the parameter uniquely associated with that node. If the lines are Ordered, then a sequence value (1,2 n) is provided; if the lines are Unordered, then no sequence value is assigned. This is stated in FLogic with method overloading [8] as follows: node[to@(threshold,identifier)=>node]. node[to@(threshold,identifier,sequence)=>node]. The parameterized method to@ describes the arc between nodes and takes either two or three parameters, depending on whether the arcs are Ordered or not. Instances the actual linguistic data corresponding to the information shown in Figure 1, may be input into the knowledge base in FLogic as follows: instance1: node[to@(1,male)->>{instance1,instance2}]. instance2: node[to@(1,father2)->>{instance4, instance5}]. instance3: node[to@(1,father1)->>instance4]. instance4 :node[to@(1,father)->>{instance2,instance3}]. instance5: node[to@(1,daddy)->> instance2]. instance6: node[to@(4,father,1)->>instance8; to@(4,father,2)->>instance9; to@(4,father,3)->>instance10; to@(4,father,4)->>instance11]. instance7: node[to@(4,daddy,1)->>instance12; to@(4,daddy,2)->>instance13; to@(4,daddy,3)->>instance14; to@(4,daddy,4)->>instance15]. instance8: node[to@(1,f)->>instance6]. instance9: node[to@(1,a)->> instance6]. instance10: node[to@(1,th)->> instance6]. instance11: node[to@(1,er)->> instance6]. instance12: node[to@(1,d)->> instance7]. instance13: node[to@(1,æ)->> instance7]. instance14: node[to@(1,d)->> instance7]. instance15: node[to@(1,i)->> instance7]. Axioms allow one to infer new knowledge (and thereby increase the size of the knowledge base). One can write axioms, for example, which take the linguistic information about lexemes/words and sememes/concepts (both represented as a network of relationships) and further identify lexical relations such as synonomy, etc. This is illustrated in the following FLogic statements: FORALL X,Z,Y,A,B Z[sem_parent->>A] <- X[to@(1,Z)->>Y] AND Z:sememe AND Y:node[to@(1,A)->>X]. 363

377 ASIALEX 2005 which may be re-stated as Z is sememe-as-parent of A if node X has the value Y for the method to@ with the parameter Z, AND Z is a sememe, AND Y is a node which has the value X for the method to@ with the parameter A. FORALL X,Y,Z X[synonym->>Y] <- Z[sem_parent->>X] AND Z[sem_parent->>Y] AND Z:sememe AND X:lexeme AND Y:lexeme. which may be re-stated as X is a synonym of Y if Z is sememe-as-parent of X and Z is sememe-as-parent of Y. Finally, queries might be likened to activation of the network because they traverse the network to retrieve all the possible answers. In the following queries (see Figure 2), we retrieve all X and Y where X is sememe-as-parent of Y; the second query retrieves all X and Y where X and Y are related by synonomy. Note that in both instances, we are querying information not entered as facts, but rather based on what has been inferred by the application of these axioms to a knowledge base about nections only. Figure 2 shows the results of the query in Ontobroker. Populating RNN using WordNet Figure 2. Displaying query results in Ontobroker Recall that we defined that RNN is based on Relational Network Theory which offers a simple yet powerful means for representing lexicogrammatical, semantic and sememic information. Lexical entries are thus indispensable to our description of linguistic phenomena. Lexical data in WordNet is taken as a resource from which to populate the knowledge base. In WordNet, words (lemmas) are organized into synonym sets (synsets), each representing one concept and there are different relations linking the synsets. This organization of lexical data provides a clear specification of lexical relations which makes it ideally suited to representation in RNN. An FLogic WordNet (FWN) is constructed on the basis of a corresponding implementation of WordNet2.0 in Prolog ( Over 200,000 lemmas with indication of linguistic properties and the relations between synsets are translated into FLogic formalism, facilitating reuse. New concepts are inferred in terms of the information represented in RNN. Simple mapping to RNN can be achieved by additional axioms. Automatic mapping of lemmas into RNN involves the following steps: i) unification of identical concepts, i.e. classes referring to the same concept should share 364

378 WORDS IN ASIAN CULTURAL CONTEXTS the same object type. ii) application of mapping axioms to populate the relational network with instances from FWN. The axioms are illustrated as follows: //Unification of Identical concepts FORALL X,Y Y:sememe <- X:synsetId[gloss->Y]. which may be re-stated as Y is a sememe if synsetid X (from WordNet) has the value Y for the method gloss. FORALL X,Y Y:lexeme <- X:synsetId[haslemma->>Y]. which may be re-stated as Y is a lexeme if synsetid X (from WordNet) has the value Y for the method haslemma. //Mapping Axioms for populating RNN FORALL A,B,S,L,G,O A:node[to@(1,G)->>(B:node) ] <- S:synsetId[offset->O; gloss->g; haslemma->>l] AND G:sememe AND concat( node_,o,a) AND concat( node_,l,b. which may be re-stated as node A is connected to node B with method to@ with G as identifier if G is the gloss, L the lemmas and O the offset of Synset S; A is the concatenation of node_ and O; B is the concatenation of node_ and L; And G is a sememe. FORALL A,B,L,G B:node[to@(1,(L:lexeme))->>A] <- A:node[to@(1,G)->>B] AND G:sememe AND concat( node_,l,b). which may be re-stated as node B is connected to node A with method to@ with lexeme L as identifier if A is connected to B with method to@ with sememe G as identifier; And B is the concatenation of node_ and L. Figure 3 and 4 depicts the instantiation of lemma from WordNet to RNN and the query results through Ontobroker respectively. Figure 3. The instantiation of father from FWN to RNN 365

379 ASIALEX 2005 Figure 4. Displaying query results population from FWN 1 The Semantic Web and Beyond The simplicity of RNN plus its applicability to all kinds of information storage suggests it may have a role to play in improving access to information on the web. Most attempts at creating a semantic web have focussed on Resource Description Framework (RDF) as the means for converting the vast collection of web resources, both documents and data, into machine-readable form. If RDF holds the promise of creating a semantic web, then I would suggest RNN opens up the possibility of an intelligent web. RNN can describe anything that can be described in RDF. Like RDF, nodes in RNN can be URI s which point to resources on the SemWeb. More importantly, RNN is driven by Relational Network Theory and incorporates developments in the theory which have been shown to not only describe but also explain linguistic phenomena in a neurologically plausible manner. Conclusion Because neurological plausibility is the goal, Relational Network Theory is constantly being checked against the findings of neuroscience. As the theory develops, along these lines so too is its corresponding notation. Subsequent versions of the notation will incorporate those network properties motivated by the linguistic evidence and tested against the neurological evidence. Given its theoretical basis, Relational Network Notation may be the key to achieving a truly intelligent web. References Bateman, John A The theoretical status of ontologies in natural language processing. In Proceedings of the workshop on text representation and domain modelling ideas from linguistics and AI. Berlin. Ford, Paul August 2009: How Google beat Amazon and Ebay to the semantic web. Retrieved Oct , from 1 Lexeme daddy is manually added to the FLogic WordNet for consistent demonstration. In WordNet2.0 library, daddy is an entry in the synset of an informal term for a father which is the hyponym synset of a male parent (also used as a term of address to your father). 366

380 WORDS IN ASIAN CULTURAL CONTEXTS Kifer, Michael, Georg Lausen & James Wu Logical foundations of object oriented and frame based languages. Journal of The Association of Computing Machinery (JACM) Lamb, Sydney M Discussion with Sydney M. Lamb. In Discussing language, ed. by Herman Parret. Mouton. Lamb, Sydney M. & J. Webster (ed.) Language and Reality. Continuum Books, London. Lamb, Sydney M Neurological evidence for relational networks. Lecture presented at the Cognitive Functional Linguistics Workshop. Nanjing Normal University, China. Maedche, Alexander D Ontology learning for the semantic Web Kluwer Academic Publishers. London. Ontoprise GmbH How to write F-Logic programs: A tutorial for the language F-Logic. Retrieved Oct from W3C World Wide Web Consortium RDF primer. Retrieved Oct , from WordNet Retrieved Oct 20, 2004, from Word combinations represented in illustrative examples in English learners dictionaries Hai Xu Guangdong University of Foreign Studies xuhai1101@yahoo.com.c Abstract This paper primarily addresses two issues: (1) What types of word combinations are usually represented in illustrative examples? (2) What principles underlie the inclusion of word combinations in illustrative examples? The study adopted a quantitative approach to examining the types of multiword combinations as shown in illustrative examples in the Big Five, NCED3 and two specialized dictionaries of collocations (BBI2 and OCDSE1). The data were further compared with the BNC and the CLEC. It was found that free and/or semi-fixed lexical collocations, which have a high frequency in natural discourse, dominate the word combinations in illustrative examples in the English learners dictionaries. We argue that learners dictionaries should give more weight to grammatical collocations in their illustrative examples. Finally, we propose that the Principle of Salience and the Principle of User-friendliness should be observed in the inclusion of word combinations in illustrative examples. Introduction The role of illustrative examples in showing word combinations is well recognized in the literature (Drysdale 1987; Cowie 1989). Yet, it still remains unknown what types of word combinations are generally represented in illustrative examples in English learners dictionaries. What principles underlie the inclusion of word 367

381 ASIALEX 2005 combinations in illustrative examples? This paper attempts to address these issues by adopting a quantitative approach. Types of word combinations Before the investigation, let s consider the typology of word combinations. Howarth (1996) proposes a practical classification of word combinations. In his view, two principal criteria are central to the taxonomy: the semantic characteristics of the whole or parts and restrictions on commutability (p.32). He also contends that word combinations (composite units, in his term) cannot be discretely divided. Instead, they show in a continuum scale. His typology, which he calls continuum model, is shown below: free collocations free Transparent Opaque restricted collocations pure idioms Figure 1. Howarth s Classification of Word Combinations restricted The diagram shows how the two principal criteria that he mentions take into effect. Horizontally, the parameter of semantic transparency is used to differentiate word combinations into free collocations, restricted collocations and pure idioms. Vertically, the variable of structural invariability is at work. At one end stand free collocations, which are structurally free as well as being semantically transparent; at the other end there lie pure idioms, which are structurally restricted in addition to being semantically opaque. And straddled in between are restricted collocations. There are no sharply-cut boundaries among the three categories of word combinations. Howarth argues that in empirical studies, the following four types of composite units deserve special attention: free collocation, restricted collocation, figurative idiom and pure idiom. The present study adopted Howarth s typology of word combinations. I also accommodated Benson et al s scheme (1986), which differentiate lexical collocations from grammatical ones. Method 368

382 WORDS IN ASIAN CULTURAL CONTEXTS For investigating the employment of word combinations in dictionary examples, I sampled some entries from the Big Five English learners dictionaries and the New English-Chinese Dictionary (NECD3). The corresponding illustrative examples in those entries were then extracted for investigation. Meanwhile, the data sampled from the Big Five and from NECD3 were compared with the corresponding data from the BBI Combinatory Dictionary of English (BBI2) and from the Oxford Collocations Dictionary for Students of English (OCDSE1). Finally, against the word combinatory patterns as shown in the British National Corpus (BNC) and the Chinese Learner English Corpus (CLEC), I discussed the implications of the results. I used two criteria in the entry sampling. First, the selected entry words should cover all the ranges of word frequency. Second, they should cover the major parts-of-speech. To meet the two criteria, a noun, a verb, and an adjective respectively ranking in the 3000 most frequent words, in the word frequency from No. 3,001 to No. 5,000, and in the word frequency beyond the 5,000th were selected: 369

383 ASIALEX 2005 Frequency Headwords ranking in the 3000 most frequent words Headwords ranking from No. 3,001 to No. 5,000 Headwords ranking beyond the 5,000th Word Class Noun history glory superiority Verb save cure rally Adjective available dreadful foremost Results Table 1. Sampled Headwords for Present Study Based on my sample data, illustrative examples in the Big Five show some general tendencies in the inclusion of word combinations: Firstly, free collocations dominate illustrative examples, whereas the other types of fixed lexical combinations, such as restricted collocations, figurative idioms and pure idioms, are not common. In my sample, there are 422 illustrative examples, but only 76 relatively fixed collocations or idioms can be found. And in the entry of dreadful, all the lexical combinations shown in the illustrative examples are free. Secondly, the lexical collocations in the sample examples (185 in total) have a higher frequency than the grammatical collocations (39 in total). In other words, examples in English learners dictionaries frequently illustrate free lexical collocations. Thirdly, the number of collocations in illustrative example varies according to the frequency of a headword. The relevant one-way ANOVA test supports this observation, as p (=.003) is smaller than.05. The reason might be that a frequently-used word is often a heavy-duty word and hence more likely to collocate with other words. Fourthly, the number of collocations in illustrative examples does not have a correlation with the part-of-speech, although the data seems to indicate that nouns contain more collocations (85 in total) than verbs (77 in total) and adjectives (62 in total). The relevant one-way ANOVA test indicates that the differences among the three categories are not statistically significant (p =.891 >.05). Lastly, the number of identical collocations included in illustrative examples in each of the five English learners dictionaries is not high. Among the 422 collocations, only 47 collocations overlap in more than three dictionaries. Compared with BBI2 and OCDSE1 (the two combinatory dictionaries), the gap between the Big Five and OCDSE1 is wider than with BBI2. This indicates that OCDSE1 has given a more detailed description of collocations than BBI2. Also noticeably, the following collocations, which the Big Five have not included, enter both the combinatory dictionaries: military glory, win glory, bring glory to, achieve superiority, enjoy superiority, clear superiority, numerical superiority, (so-called) racial superiority. We may wonder whether the above collocations have a high frequency in the BNC as well. If the case is true, then the Big Five should be criticized for failing to record some frequent collocations. If not, we can say that only the specialized combinatory dictionaries will give such a detailed treatment. I used the WASPBENCH to concordance the BNC. The results indicate that among the 63 collocations that the Big Five have not recorded, 20 (around 32%) are insignificant in terms of salience, 26 (around 41%) collocations have a salience lower than 8, and only 17 (around 27%) have a salience higher than 8. That is, the Big Five fail to record only 27% high frequent collocations, and they should not be blamed for an oversight of inclusion. 370

384 WORDS IN ASIAN CULTURAL CONTEXTS As for the above eight collocations that have been included in both of the combinatory dictionaries, only two high-salience collocations (numerical superiority, (so-called) racial superiority) have not entered the English learners dictionaries. So, in general, the Big Five have a smaller coverage of word combinations than the combinatory dictionaries. This might be attributed to the fact that the former are not specialized combinatory dictionaries. Let s further consider whether those collocations as shared by the Big Five have a high frequency in the BNC as well. Again I used the WASPBENCH to concordance the BNC and obtained the salience of those collocations. The statistics shows that most of those collocations (accounting for about 70%) have a high salience in the BNC. Only 30% of the collocations have a salience which is lower than 1. The results correspond with our expectations that most collocations recorded in illustrative examples in the Big Five should be frequently-used in natural discourse. Let s shift our attention to NECD3 the English-Chinese dictionary specifically designed for Chinese speakers. It encompasses some combinations that none of the Big Five learners dictionaries includes: a case history, oral history, unprecedented in history, save one s eyes, save the post, save the game, available means, available candidate, available nitrogen/water, dreadful scene, dreadful disrepair, rally one s courage, rally one s energy. Are those lexical combinations common in natural discourse as well? The salience of those combinations as shown in the BNC gives a negative answer. Except for the first two, all of them are insignificant. To put it in another way, NECD3 wastes its space by recording some word combinations that are infrequently used in natural discourse. Learners dictionaries should be oriented towards a particular group of dictionary users. Naturally, we may wonder whether the Big Five and NECD3 have solved some particular collocational problems of Chinese learners of English. Before addressing this issue, let s clarify what lexical combinations are overused and what are underused by Chinese learners. I did the relevant concordances in the CLEC. The results are shown in Table 2. Overused Collocations (freq.) Underused Collocations (freq.) history ~of (73); [as subject] maths, etc.,~(24); in~(19) [modified by adj.] natural/chequered/social~(0); [modified by n.] case/art~(0); trace/rewrite/write ~(0) save ~(the/much/ fresh) water (127); ~dying patients (7) ~on cost/bill (0); ~soul/ marriage/forest (0); ~from death/extinction/closure, etc. (0); God~(0) available ~water (3); ~war (2); ~way (1) [modified by adv.] readily/currently/widely/freely~(0); be made ~(0); sb. ~(0) glory one night s ~(1) ~of God/lord (0); moment/ blaze of ~(0); crowning ~(0) cure ~environment (1) ~sb. of (1); [modified by adv.] completely/miraculously ~(0) dreadful [as modifier] ~world (1) / strain ~mistake/thing/thought (0); absolutely ~(0); (1) / reputation (1) / bouncing (1); how~ (2) look ~(0) superiority man s ~(3); great ~of (1); sense of ~(1); concepts of ~(1) air ~(0); demonstrate ~(0); belief in ~(0); numerical/racial/innate, etc. assumption of ~(0) 371

385 ASIALEX 2005 rally / ~support/troops (0); ~round (0); supporters ~ (0) foremost first and ~(3) [as modifier] ~scholar/exponent, etc. (0); ~amongst (0) ; ~in mind (0) Table 2. Overused and Underused Collocations in CLEC Table 2 indicates that Chinese learners have encountered at least three problems in using collocations. Firstly, they often use some unidiomatic expressions. Such expressions as available war, one night s glory, cure environment, dreadful reputation, would sound very odd to a native speaker. These expressions are freely constructed by learners and are apparently the result of negative transfer of their mother tongue (Chinese). Hence, Chinese learners generally have superficial knowledge of prefabricated word combinations. Secondly, they are not aware of the collocation range that a headword can be used. For instance, Chinese learners have no problem in using history when it refers to a discipline or when it is used in such constructions as in history, history of, but they have no idea that history can collocate with some words of relatively formal style, such as chequered, trace. Thirdly, compared with native speakers writings, the recurrent collocations exhibited in learners composition have a very limited coverage. Many idiomatical collocations have not become their productive vocabulary. Therefore, an ideal dictionary for Chinese speakers should take into account those special needs when its illustrative examples are selected. Thus, the Big Five more satisfactorily handle collocations in their illustrative examples than NECD3 does. As presented above, most collocations represented in the sample illustrative examples in the Big Five are typical ones which frequently occur in discourse, whereas the word combinations listed in the sample examples in NECD3 are often infrequent ones. The Big Five often use sentence examples as well as the techniques of the bold typeset, the skeleton example and the gloss to illustrate collocations, while in the sample data, NECD3 often uses phrasal examples. However, none of the six learners dictionaries has worked out a solution to remind users of their common errors in collocational usage. And this kind of prompts is very helpful for learners who have difficulty in inferencing the information from illustrative examples. Discussion and Implications The frequent employment of free and/or semi-fixed collocations instead of figurative or pure idioms in illustrative examples can be attributed to: (1) the space constraint, (2) the format of a learners dictionary entry and (3) the frequency of collocations in actual discourse. The meaning of a figurative or pure idiom is always opaque and need quite a lot of space to illuminate. And an illustrative example cannot afford such space. In addition, a learners dictionary often sets up a special section of phrases and idioms in its entry. Phrases and/or idioms are usually grouped at the end of an entry. More importantly, collocations have a higher occurrence in discourse than idioms. According to one statistics, collocations constitute around 40% of discourse (Cowie, 1998: 10). Thus, figurative or pure idioms in illustrative examples should be cautiously used in learners dictionaries. The learners dictionaries include much more lexical collocations in their illustrative examples than grammatical collocations. This practice is, to some extent, unsound, because learners, particularly Chinese students, often have difficulty in using 372

386 WORDS IN ASIAN CULTURAL CONTEXTS the particle in a phrase. So, learners dictionaries should give more weight to grammatical collocations in their illustrative examples. Unlike a combinatory dictionary, a learners dictionary cannot handle collocations in its illustrative examples in a lengthy way. But highly frequent collocations should be focused on in illustrative examples in a learners dictionary. As many English learners cannot competently use word combinations, they need special help in this respect. An illustrative example containing typical collocations supplies them a model for imitation. But how do pedagogical lexicographers know that a special group of users have specific difficulties in using some collocations? One possible solution is to compare a learners corpus with a native speakers corpus, single out the overused and underused collocations, and then highlight those collocations in illustrative examples. Such kind of comparison usually reveals more information about learners knowledge of collocations than the lexicographer s assumption does. On the basis of above discussion, two principles governing the representation of collocational information in illustrative examples can be summarized: the Principle of Salience and the Principle of User-friendliness. Collocations included in an illustrative example should have high salience (or statistically significant re-occurrence) in discourse. The availability of corpora and concordance tools will help the lexicographer to observe recurrent word combinations. Another type of collocations that deserves lexicographers special attention is the collocations that users have potential difficulties in using. The collocational errors committed by them convey a useful message to lexicographers. Those problematical collocations should be focused on in illustrative examples. On the other hand, in order to catch users attention, lexicographers need to consider the techniques that are effective in presenting those collocations in illustrative examples. Bold typeset, skeleton examples, glosses and counter examples are some commonly used techniques of presentation. References Benson, M., E. Benson and R. Ilson (1997), The BBI combinatory dictionary of English (Amsterdam-Philadelphia: John Benjamins Publishing Company). Cowie. A.P. (1989), The language of examples in English learners dictionaries, in G. James (ed.), Lexicographers and their works (Exeter: University of Exeter Press), pp Cowie, A. P.(1998), Phraseology: theory, analysis, and applications (Oxford: Oxford University Press). Drysdale, P.D.(1987), The role of examples in a learner s dictionary, in A.P. Cowie (ed.), The dictionary and the language learner. Papers from the EURALEX Seminar at the University of Leed, 1-3 April 1985 (Tǖbingen: Max Niemeyer), pp Gillard, P. et al. (2003), Cambridge advanced learner s dictionary (Cambridge: Cambridge University Press). Howarth, P. (1996), Phraseology in English academic writing: some implications for language learning and dictionary making (Tübingen: Max Niemeyer Verlag). Lea, D. et al. (2002), Oxford collocations dictionary for students of English (Oxford: Oxford University Press). Rundell, M. et al. (2002), Macmillan English dictionary for advanced learners (Oxford: Macmillan Education). Sinclair, J.M. et al. (2003), Collins COBUILD advanced learner s English dictionary (Glasgow: HarperCollins Publishers). Summers, D. et al. (2003), Longman dictionary of contemporary English (London: Pearson Education Ltd). 373

387 ASIALEX 2005 Wehmeier, S. et al. (2000), Oxford advanced learner s dictionaries of current English (Oxford: Oxford University Press). Wu, Y. et al. (2000), Xin yinghan cidian (New English-Chinese dictionary) (Shanghai: Shanghai Translating Press). Transform from a productive approach to a process-oriented approach in online English class Xiaoyu Xu Northwestern Polytechnical University, Xi an, China xiaoyuxu2003@yahoo.com.cn Xue ai Zhao Northwestern Polytechnical University, Xi an, China xazhao@nwpu.edu.cn Abstract Web-based learning has turned out to be a trend in the English Course study in the universities of China nowadays. With the maturing of the construction of the online teaching materials, people s focus has turned to the pedagogy of the online teaching. In the traditional classroom, most of the teachers would adopt the Product Approach, pay much attention to the end of the teaching and neglect the nurturing of the practical skills of language learning. However, when the language study is changed to the web-facilited study, the Product Approach can no longer meet the demands of the online teaching, for the latter requires much more monitoring and assessing of the study progress of the students. Therefore, a transform of the pedagogy is a necessity to promote the language teaching effect. In this paper, we will observe the application of the Process-oriented Approach in the online English class. This method, first commonly used in the English writing class, emphasizes on the process of the language study, and creates an environment for the students to participate in the virtual classroom activities, including the task-based learning, webquest learning, theme-based learning, etc. While the teacher employs all these teaching methods, they should also notice combining the fomulative evaluation with the summative evaluation, and thus the efficiency of learning language online will be assessed based on the procedural steps made in the study, either in daily performance or in the exams. Introduction In the epoch of 21st century, it seems everything changes and rolls on quickly. With the world being facilitated with modern technology, things become facile, which may be observed especially in the English classroom. Traditionally, teachers have to use the chalk-and-talk way to give lessons and the students passively listen to the lecture, but now, since there is the VOD(Video-On-Demand) technique and the Internet, the students become much more involved in the class and they can easily click the button and find the materials they are required to study and those they want to study. Thus, the learning process tends to be a student s self-directed and personalized learning style. 374

388 WORDS IN ASIAN CULTURAL CONTEXTS Background for the Transform With the changes in the English classroom listed above, we need to notice there has been a change in the aim of teaching, i.e. from a productive approach to a process-oriented approach. The above two both refer to the attitudes we may hold towards teaching and the way we evaluate our teaching. The former one shows that a classroom is like a factory and the students are regarded as raw materials or semi-products. The teacher s job is to instill knowledge into the students and when a student mechanically remembers all the rules and vocabulary, we say we have made a good product. Whether the teaching is successful or not totally depends on the number of the products, but not the quality of a product. The proficiency of Chinese students English level is decided simply from the paperwork in one final exam. The practical language skills are neglected and the rote learning is a way most students and teachers use to pursue a high exam score. Therefore, the evaluation of the teaching result is much more important than the monitoring, diagnosing and adjusting during the teaching. Nobody cares for their abilities in listening, speaking, reading, writing and translation, let alone the intercultural communication ability. Teachers are complaining; students are complaining; the whole society is not satisfied with the result of the English teaching. However, the process-oriented approach is quite on the opposite. This kind of approach first appeared in the English writing class. Teachers observed that if they played a guiding role in the writing and encouraged the students to work on their own paper and share their opinions on the papers of each other, instead of letting the teacher become the judge and give the marks, it would be much helpful to improve the students learning ability and writing level. Owing to this, the classroom activities are focused on the students performance in the daily class activities, and communication as well as cooperation with other students and with the teacher tends to be more important. The teacher will make a judgement of the study of the student mainly on the involvement in the whole learning process, such as the preparatory work of students before the class, their performance in the class, their homework and the small tests in phases, and also the performance in the final exam. This kind of teaching approach has won the popularity inside the university and is gradually being employed in the other fields of language teaching, and one field that benefits most from this approach is the online-learning English class. Why should it be in the online-learning class? Compared to the traditional classroom teaching, online classroom learning requires the process-oriented approach more urgently. In the traditional classroom, although the students are regarded as the final products, there is still a teacher going around checking over the study of the students and making judgements on the performance of the students. But in the online class, it s quite hard to monitor the students. The teacher can only make evaluations based on the online homework, online testing and online forums. With a simple way to make the judgements on the study performance, the students true learning process cannot be wholly observed and modulated; the activities online will not be able to reflect the real problems in the students study for the teacher. Therefore, a process-oriented approach being employed in the online-learning class becomes necessary and important. 375

389 ASIALEX 2005 Necessity of the process-oriented approach in the Chinese college English reform The time for the Chinese government to pick up English teaching may be dated back to the 1860s, and the trend for English language study becomes hotter and hotter with the globalization of the world from the end of the last century. When the century turned to a new leaf, Chinese people see two important things concerning both their physical and their spiritual life: one is China joining in the World Trade Organization (WTO) and the other is the successful bidding for the 2008 Olympics. These two events have cast great influence on the Chinese social life and more and more people are engaging in the work. Both our central government and civilians realize a more important role of English being played in the country, for enhancing the boosting of economy internally and for promoting a better understanding and communication with people externally. Thus, to encourage the country-wide English study, Chinese government puts great efforts and sums of money to change the learning environment and learning materials. In the year of 2004, the government chose 180 key universities and some non-key universities to take the College English Reform and installed the multimedia network on the campus for College English Course study. Therefore, the students can have easy access to the computer and browse the course online, which may afford them a better language learning environment and help the students acquire the decent language materials online and thus they can produce decent language. What is inside a process-oriented English classroom? In the Reform, the process-oriented approach is firstly adopted in the design for the curriculum. In the traditional classroom, we only had one kind of class and one way of assessment of the teaching effect the summative evaluation, whereas in the modern English classroom, to strengthen the teaching effect, we facilitate the online classroom with the other two kinds of class one is face-to-face teaching and the other is small-class tutoring. Each type of the three classes plays different roles in monitoring and administrating the learning process, and focusing on the different language skills to make sure the students can be fully developed into a learner with comprehensive skills (see Illustration 1 below). From the illustration below, we may see different roles played by the classes in the process-oriented teaching: in the online English class, i.e. the web course class, students try to practise the skills of listening and speaking; in the face-to-face teaching, the purpose in learning is to study the skills of reading, writing and translating; in the small class, the teacher adopts a tutoring way and the main function of the class is to play an assistant role to the web course class, but here the students will find the teacher has changed the role from being the lecturer to the language learning advisor. What s more, we should bear it in mind that though both the face-to-face class and the small tutoring class offer the students the chance to meet the teacher and listen to the lecture or tutoring, they play their roles only in this transitional phase of the reform, and the web course has already taken a dominant role in the College English curriculum. 376

390 WORDS IN ASIAN CULTURAL CONTEXTS Web course learning Selflearning Facetoface Teach g Multimedia Class Skills of listening & Speaking Skills of reading & writing & translating Small class Tutoring Comprehensive skills Intercultural communicative Skills Fully-developed knowledge and personality Illustration 1:Classes that form into the process-oriented teaching Elements in the process-oriented approach in the online English course Besides the curriculum change for the College English course, there are other elements which may play great function to ensure the quality of the teaching. The first one is the way of interaction in the class, which plays a very important role in the process-oriented approach, for if there was no interaction, the monitoring would not be possibly carried through the learning activities. In the traditional way of teaching, there seemed to be no interaction between the teacher and the students. The teacher usually had a chalk in hand and a blackboard behind, whereas the students passively listened to the teacher with little chance to talk and to express their opinions in English. Their creative power and initiatives stayed in a dormant state and needed to be wakened up (see Illustration 1 below). Illustration 2:Traditional Teaching Mode In contrast, in the online class, the teacher pays much attention on the way of interaction and calls for a better way of communication between the teacher and the 377

391 ASIALEX 2005 students. (see Illustration 3 below). Illustration 3:New Teaching Model From the illustration above, you may see there exist kinds of interactions among the teacher, the students and the computer. In the different interactions during the process of learning, we may use different ways of evaluations to ensure the efficiency of the online learning. This is also a very important element in the online learning process (see Illustration 3 below). Items Mode of interaction Teacher computer Student computer Teacher computer student Student computer student Distance interaction Activities in learning lecturing Self-learning Forums and Bulleting Board System Cooperative learning Information exchanging Learning areas Reading, writing, translating Listening, speaking Problem diagnosing Theme-based webquest Fetching further learning materials Ways of evaluation online homework checking Computer record and evaluation Online phase tests and final tests; Peer learning and evaluation Illustration 4:Evaluations in the Online English Class In these evaluations, teachers will take every learning activity into consideration and give a comment on the student s performance, which may include daily performance ( i.e. class activity involvement, preparation for the class, reviewing of the knowledge, the handing-in of homework as well as the test scores), all of which have already been taken into the database in the computer. With a good design for the curriculum, a variety of activities online and the evaluations throughout the whole learning process, we can see there will be a better learning environment for the students and assist them to exert their potential to the full. Conclusion When we face this changing world, we have to quicken our steps and catch up with the trend. As it is said: There is nothing that will never change, but only change itself. Therefore, as the teachers, we might closely notice the changes in English teaching, and make adjustments to them. 378

392 WORDS IN ASIAN CULTURAL CONTEXTS References Jack C. Richards and Theodore S. Rodgers (2000), Approaches and methods in language teaching (Cambridge: Cambridge University Press). Bernard Spolsky (2000), Conditions for second language learning (Shanghai: Shanghai: Foreign Language Publishing House). Jing, Wang and Ansheng, Wang (1992), Cognitive psychology (Peiking: Peiking Publishing House) Long, M.H Process and product in ESL program evaluation. TESOL Quarterly 18(3): Xiaoyu, Xu, Jian, Wang and Jun, Lei (2004), The standards for the evaluation of interactions in the English webcourse teaching, Foreign Language Education 25 (14-15) Abstract Students preferences of Web-based EFL dictionaries Shigeru Yamada Waseda University EFL dictionaries are now available not only on paper but also in hand-held machines, on CD-ROMs, and on the Internet. The electronic media have made for new search methods which were not possible with paper dictionaries. The ability to make full use of such products and methods will enhance the study of English. Especially, web-based versions of dictionaries, free of charge, are welcome additions to language-learning resources for those students who have easy access to the Internet. Under appropriate class circumstances, web-based dictionaries can be useful reference and teaching materials. I used to have my students buy a copy of a paper EFL dictionary to be used in class but have switched to web-based ones. I have used the web dictionaries by Cambridge, Longman, and Oxford to familiarize students with the content and use of state-of-the-art English dictionaries. I conducted a number of exercises and activities for these purposes and to gauge the students preferences and ability to use the dictionaries. The dictionaries differ in purpose, content, and structure, and the students tastes and ability to use them differ. In this paper, I point out the merits and shortcomings of the three web-based EFL dictionaries, based on their use by the students. 379

393 ASIALEX 2005 Lexical concerns about neologism in Chinese lexicography: a cognitive approach to the motivated structure of new words and expressions 1. Introduction Yihua Zhang Guangdong University of Foreign Studies, China Since the early 1980s, the reform and open policy bring about a lot of changes in China, the advanced technology and sophisticated equipment well up unceasingly to the Chinese society along with exotic culture and styles of working and living from western countries. This inevitably leads to unceasing appearance of new words and new terminology, which can be found in large numbers in every communication media. Especially with the coming of the cyber age, there has been an acceleration in the spread of new words. In the Chinese-based Internet along with other communication media, at least thousands of new words involving all walks of life come forth annually, and about of them become fixed. In 20 years time, Chinese will have accumulated a great quantity of new words including such a number of loan words as well as lettered words, so that Chinese writing style seems to have experience a great change. Let s look at the following examples: APEC 記者招待會後, 我約了 STV 的記者和一群 MBA MPA 研究生朋友, 討論中國加入 WTO 後 IT 業對 GDP 的影響 讀 MBA 的張小姐本來想去.COM 當 CEO, 但覺得 IT 業風險大, 轉而想去 Nike 公司 讀 MPA 的李先生卻認為加入 WTO 後政府職能將大有改變 隨後大家相約關掉 BP 機, 也不上 Internet 的 QQ 和 BBS 聊天, 而是去了 KTV 唱卡拉 OK "(cited from the People s Daily, the 20 April) This is a piece of news report quoted by a professor to criticize the abuse of lettered words in Chinese newspapers. Even it is wholly turned into English; an English native speaker can hardly catch the meaning without difficulty. This phenomenon is common occurrence. When I randomly picked two official newspapers from my bookshelf and go through it quickly, the following lettered words leapt to my eyes: Guang Ming Daily (31 March) :NEC, CMMI, Erwin Neher, DNA, RNA, HOTHEAD, HDTV, CDT, LCD, TFT, GDP, TCL, MP3, IT, PC, CD, DVD, PRGR, COMBO, CNGI, DRW, CD-ROM, DVD-ROM, PSV, IPv, Show NET, QQ, PDA, CPU, XL, VOIP, RFID; the mixed lettered words: U 盤 " E 考通 " T 計畫 "etc. Nan Fang Daily (26 March) :3+X, KTF, KT, MPL, ARD, CPI, ST, CCTV, GDTY, OTC, NBA, CBA, initialized economy, pay-per-click, Google, Monster.com, COTTON US, CD; the mixed lettered words: X 光 " B 超 " A 股 " B 股 " AA 制 " 卡拉 OK" IT 行業 " J.K 羅琳." The above-cited lettered words involve general names, technical terms as well as commodity names, company names, people names, and trademark etc. In additional to these, there are lots of other imported words. It goes even further on the website; when you browse a Chinese website at random you can see many foreign words mixed with Chinese characters. 380

394 WORDS IN ASIAN CULTURAL CONTEXTS Some years ago, this was only seen in Hong Kong, known as an interlanguage area. And then, why and what has given rise to such an interlanguage in a purely Chinese populated area? There are, I suppose, two reasons: First, due to the Internet, information about new products and new events, once occurred abroad, spreads quickly within mainland China before it is naturalized. Second, the young professionals, knowing more or less some foreign language, are likely to use new foreign words directly for the sake of avoiding mismatching of Chinese equivalents. 2. Distribution Characteristics of new words The new terms wholly fall into the content words, most of them are nouns and verbs, none of function words is found. As for the word-formation, compound words (including acronyms and Initialisms) are by far more numerous than simple words. According to the Electronic Dictionary of Contemporary Chinese Neologisms, new words have the following distribution: Items Dist. number Percent Items Dist. number Percent Items Dist. number Percent daily life social field economics politics culture science/ technology education agriculture sanitation military sports law business industry To our surprise, the most important part of new words focus on everyday life and social field instead of business and industry, this suggests that new words do have influence upon our lives. Related study indicates that the population distribution is regular. The numbers of users are inversely proportional to their ages, but directly proportional to their knowledge. Netizens take the most active part in the creation and use of neologism. 3. Different attitudes towards neologisms Although the neologism has always existed, the number of new words is never so important and their spreading is never so fast as in today s cyber-age, thus the argument about it gets more intense. Since dictionaries hold ultimate authority in their users mind, Lexicographers and linguists are very prudential about new concepts and their designations. In the past ten years, 3-5 academic conferences have been held annually with respect to Chinese lexicography, the neologism is always one of the main topics for discussion. I remember that, in the early 1990s conference, talking about the inclusion of neologism in dictionaries, the voice of disagreement was far louder than that of agreement with regard to lettered words and overseas Chinese words. They believed that lettered words were an abuse of loanwords and that they were profane the purity of the Chinese language. Some scholars, with a wait-and-see attitude, thought that time will prove everything about them. With the popularity of the loan words in daily life, those who had voiced disapproval became sparse by the late 1990s. What was left is a concern about how to standardize the new words and guide their use in dictionary-making, especially about the exotic loanwords and lettered words. In the new century, exotic loanwords and lettered words have become a common occurrence in Chinese media as is mentioned at the beginning of the paper, it is a flood of them. How should dictionaries confront all such problems? Chinalex held an 381

395 ASIALEX 2005 academic conference in Chengdu at the end of Sept. 2002, the main topic focused exclusively on new words and expressions, as well as the compilation of neologism dictionaries. The dominant opinion focused on how instead of whether such special lexical items should be dealt with in dictionaries. While lexicographers hold a positive view towards neologism, there exist different voices in the social and cultural circles. To solve this problem, the Language Application Administration of Ministry of Education has initiated a Web Questionnaire Survey on Hotspot Issue of Language Application so as to solicit societal opinions about neologism, lettered words and cyber-language, etc. In 2004, however, a so-called Chinese Defense Campaign was launched through China s media. Shi Fang, a journalist of People s Daily Online, issued an article titled Pidgin English should be eliminated, and Loan Words be Standardized 1; Liu Hanjun wrote an article published in China Daily (April 20, 2004): Do Respect Our Mother tongue, in which he even cited La derrière Classe (The Last Lesson) by Alphonse Daudet, so as to arouse people s cherishment of our own language; Also, in his article Safeguard Our Mother Tongue 2, Lin Zhipo pointed out that the harm to the Chinese Language is proceeding from three aspects: a) inadequate selection of entry-words for dictionaries; b) Hong Kong and Taiwan dialects; c) pidgin English. However, most scholars are rather optimistic toward such neologism. Mr. Wang Huidi from the Chinese Language Society of HK expressed his own view as a response to the above arguments. In his article put up on the Huayuqiao Website3, he said, the occurrence of lettered words in Mandarin is a progress for Chinese, which manifests the self-confidence of our mother tongue. As for the overuse of such words, persuasion is better than prohibition. In reality, adequate use of lettered words may enrich Chinese expression. For example, MP3 is popularly used? But who knows its Chinese name? Some Chinese scholars are willing to develop a national standard for LINUX system so as to compete with Microsoft, but is there anyone who knows how to call LINUX in Chinese? Everybody knows B 超 [B chao] (B-ultrosonography), IP 電話 [IP dianhua] (IP phone), IC 卡 [IC ka](ip card). Even farmers in china use IP 電話 for chatting, B 超 for medical diagnosis, and IC 卡 for shopping. but seldom would they question what the Chinese equivalents of these B, IP and IC are. Another example is NBA match, the Chinese equivalent 美國全國籃球聯賽賽場 (US National Basketball Association Match) is neither cognitively economic nor acceptable to the audiences as well. 4.Research on Neologism and Neologism Dictionaries The development of neologism has always been the concern of lexicographers. According to the statistics of the Index Database of academic papers on Chinese Lexicography over 20th Century, there were only 45 papers related to neologism from1900 to 1993; while the statistics of CKK shows that the number of articles about neology published from 1994 to the present has greatly increased: 382

396 WORDS IN ASIAN CULTURAL CONTEXTS Items up to the present Neologism Number Annual average Number Annual average Related to neologism Neologism as topic Lettered words Number Annual average Number Annual average Related to neologism Neologism as topic Imported words Number Annual average Number Annual average Related to Imported words Imported words as topic Total From the above table, the number of research papers on new words, lettered words and imported words in Chinese have speedily increased, especially those on lettered words, there are ten times increase. This proves from one aspect that the imported and lettered words have been steadily on the increase in Chinese, a phenomenon that increasingly draws the attention of scholars and the public. The growth of research papers on this issue is in proportion to that of neologism. On April , the First International Conference on Chinese Lexicology was held in Wuhan University. Lots of lexicographers attended this conference and one of the most important topics is Chinese neologism. Dozens of scholars talk about it. Sun Maosong, prof. of Tsinghua University, jointly with other scholars launched a special investigation into the formation of modern Chinese lettered words. They conclude that seven types of components contribute to their formation in a dozen ways. Zhou Hongbo, senior editor of the Commercial Press, puts forward that neologism dictionaries are urgently required to recommend the well-formed new words to the public and provide necessary guidance for their use. Kang Shingyong, prof. of Yantai Normal College, undertook a systematic study on Chinese neologism based on a special corpus, including: (1) family trend of word formation; (2) overall situation of new word family; (3) configuration of new word family; (4) syntactic features of new word family; (5) motivation of new word familization. Guo Xi, prof from Jinan University, provides a general principle to set a standard for lettered words use, that is: being scientific, compatible and flexible. He also suggests that the lettered words be pronounced on the basis of IPA, and be written in accordance with the orthography of the Chinese pinyin and the Romanization (Pinyin) of Chinese, trying to normalize the use of lettered words. The neologism research achievements should naturally find their place in the dictionary. Since 1987, about 47 Chinese neologism dictionaries, including those of lettered words, have been published. In terms of the selection of entry words, some of them mainly include the words first found in the early 20th century, without contemporary characteristics; many others are limited to a rather small range of vocabulary, e.g. covering a certain style of text; still others include only several hundred words. But those published in the 21st century take more care about new things. For example, the Xinhua Neologism Dictionary (by Zhou Hongbo) includes exclusively the words coined since 1990s. The Comprehensive Neologism Dictionary (by Kang 383

397 ASIALEX 2005 Shiyong) includes the words (many of them are lettered words) that appeared in The Neologism Dictionary (by Shen Mengying), has two parts. The first one includes new words beginning with the Chinese characters; the other one includes new words beginning with letters and Arabic numerals. In 2002, the Contemporary Chinese Dictionary, the most authoritative and well-formed dictionary in China, and put out its supplementary edition which has adopted a large number of new words, such as: 三個代表 [sange daibiao] (Three Representatives); 一國兩制 [yiguo liangzhi] (One country, two systems); 納米技術 [nami jishu] (nanotechnology); 藍牙 [lanya] (bluetooth); 寬頻 [kuandai] (broadband); 克隆 [kelong] (clone); 大盤 [dapan] (large cap); 建倉 [jiancang] (open a position); 補倉 [bucang] (cover short position); 套牢 [taolao] (hung up); 前衛 [qianwei] (very modern); 非禮 [feili] (incivility);and lettered-words like CEO, CFO, CGO, CIO, COO, CTO and so on. As for improving the neologism dictionary, and in order to standardize the Chinese neologism including imported words and lettered words, we should investigate into the constituents, phonetic rules and family trend of neologism. It proves that the research on the mechanism and the motivation of word-formation turns out to be an effective way to such an investigation. 5. Motivation of lexical composition of neologism. The motivation of word-formation refers to the ways in which new things, events and phenomena are designated a certain linguistic sign. It illustrates the relationship between the sign and its denotatum, so that the meaning of the word can be ascertained from the sum of the meanings of its constituents. The motivation can be based either on linguistic form and its concept or widely social and cultural background. The former is micro-motivation while the latter macro-motivation. The recognition the motivation of neologism is a key point for the selection of dictionary entry-words and lexicographical definitions, especially for bilingual dictionary translation. According to theoretical modules of cognitive linguistics such as mental space (Fauconnier1985, 1998) and conceptual integration (Turner and Fauconnier 1995, 1998;Turner 1996; Coulson 1997), the representations of language can be illustrated by conceptual schema. It is a visualized structure model entailing certain concepts. The recognition of the composition of linguistic expression and its structure model is the main source of meaning comprehension and representation. The following discussion will probe into the mechanism of word-formation from the perspective of linguistic and conceptual schema, as well as cross-space mapping and conceptual blending, so as to bring to light the micro-motivation of words, including composition, abbreviation, metaphorization, sound imitation, importation as well as calque. It must be indicated that, since new technology and new concepts generally originates from English speaking countries, the motivation of Chinese neologism usually bears influence of English word-formation Metaphorization and its motivation of word-formation Metaphorization is very influential to the development of language; it affects directly the linguistic activities, such as the way of thinking, the creative power of mind 384

398 WORDS IN ASIAN CULTURAL CONTEXTS and the formation of concept. Lakoff has carried out a broader analysis of metaphor in the years 1980 s; two fundamental conclusions can be made from his analysis: (a) all language is metaphorical and (b) all metaphors are ultimately based on our bodily experience. Thus the metaphorization is a conventional cognitive module of a language community. Making use of certain relations between the source and target domains, the metaphor can represent semantic information in a vivid and figurative way. Ability to remember things a: memory in computer b: 存储 Memory 内存 a : memory in brain b : 记忆 Fig.2 Blending of metaphoric concept Blend Target Generic space Source Association From the point of view of morphological structure of specialized words, metaphor is a covert implication without a comparative word. It is only the partial projection and mapping of structure schema between tenor and vehicle. For example, in evolution of informatic terms, inventors usually name a new concept after a well-known term and create the new one by means of conceptual mapping. There are two cognitive bases for metaphor: similarity-based metaphoric mapping and relevant experience based metaphoric mapping. The neologism in the following table is coined by similarity. Sign Tenor Vehicle Similarity mouse/ 滑鼠 hand-operated device for the control of the cursor a small furry animal Form and action on PC screen window/ 視窗 operating environments a space of glass in Structrure for PC the wall memory/ 記憶體 part of a computer that retain information a property of the human mind Function: process information explorer/ 搜索器 package/ 套裝軟體 virus/ 病毒 Microsoft Popular Web browser stored collection of written programs of computer a self-replicating program that spreads and destroys data someone traveling through unknown area to find about it things wrapped in paper or box a very small living thing that causes infectious illnesses Character: seek knowledge of unfamiliar regions Function: collect things Nature: infectious and harmful Metaphor not only applies to noun-formation but also to verb-formation. For example, 炒魷魚 (chao youyu)originally refers to cooking a dish : the squid curls up when being cooked. And now it is used figuratively to describe the scene of a man, who is fired, rolls up his quilt and prepare for departure. 跌眼鏡 (die yanjing), which comes from dialect, literally refers to a man is so surprised that his glasses falls, and it metaphorically means that an event goes beyond one s expectation, extremely astonishing. This overstatement expresses vividly the manner of surprise, very true to life. And the expression 令人大跌眼鏡 (lingren da dieyanjing) is often found in many communication media Composition and its motivation Composition means that two or more than two simple conceptual schemas (input space) are projected into blending space through cross space mapping, completion and 385

399 ASIALEX 2005 elaboration process so as to constitute new terms. The denotatum of the composed concepts is simplex, and readers can easily catch the whole meaning at the first view in light of its compositional motivation. In the process of conceptual integration, both conceptual metaphorization and language schema take effect. For example, the compound spaceship is precisely the product of the conceptual mapping between underwater vehicle and space vehicle and also of the structural projection from the space and ship into the blend space. The compound words listed in the following table all fall into this category: spaceship space+ship 航太 [hangtian]+ 飛船 [feichuan] hypertext hyper + text 超 [chao] + 文本 [wenben] internet inter + net 互聯 [hulian] + 網 [wang] Internet Explorer Internet + Explorer 英特網 [yingte wang]+ 流覽器 [liulan qi] download (down+load)+ ( 下 [xia]+ 載 [zai])+ 加速器 accelerator accelerator [jiasu qi] electronic electronic + 電子 [dianzi] + 商務 [shangwu] commerce commerce The composition shows a strong tendency to form words as a family, for instance, webaholics are initially called 網蟲 [wangchong] (net worm) in Chinese, and then other words derived from the original one to form a large family, e.g. 准網蟲 [zhun wangchong](quasi-worm: green hand on Internet), 爬蟲 [pachong] (reptile worm: webaholic at primary level), 小蟲 [xiaochong] (small worm: webaholic at intermediate level), 大蟲 [dachong] (big worm: webaholic at advanced level), 飛蟲 [feichong] (flying worm: webaholic at high level) and so on. So do the lettered words. Let s see the compound words based on lettered words: 1) Compound words based on RNA (ribonucleic acid/ 核糖核酸 [hetang hesuan]): RNA 干涉 [ganshe] (RNA interference), RNA 雜交 [zajiao] (RNA hybridization), RNA 複製酶 [fuzhi mei] (RNA replicas), etc. 2) Compound words based on DNA(deoxiribonucleic acid/ 去氧核糖核酸 [tuoyang hetang hesuan]): DNA 排序 [paixu] (DNA sequence), DNA 基因圖 [jiyin tu] (DNA images), DNA 基因圖譜測定 [jiyin tupu ceding] (DNA profiling), DNA 親子鑒定 [qinzi jianding] (DNA paternity testing), DNA 電腦 [jisuan ji](dna computers), etc. In the above examples, lettered words RNA and DNA function as an affix, for many new words and terms are derived from them and used widely in Chinese language communication Abbreviation and its motivation Abbreviation incarnates the cognitive economy; it can reduce a group of long string terms to a short expression for the sake of brevity. English abbreviated words 386

400 WORDS IN ASIAN CULTURAL CONTEXTS constitute the main source of Chinese lettered words. Abbreviations are regularly formed in the flowing ways: a) Initialism:BBS:bulletin board system/ 電子公告牌 [dianzi gonggao pai], GPRS : General Packet Radio Service/ 通用分組無線業務 [tongyong fenzu wuxian yewu], IE : Internet Explorer 英特網流覽器 [yingte wang liulanqi], etc. b) Acronyms:CEPA:Mainland and Hong Kong Closer Economic Partnership Arrangement/ 內地與香港更緊密經貿安排 [neidi yu xianggang geng jinmi jingmao anpai], APEC : Asia Pacific Economic Cooperation / 亞太經濟合作會議 [yatai jingji hezuo huiyi], OPEC:Organization of Petroleum Exporting Countries/ 石油輸出國組織 [shiyou shuchu guo zuzhi], etc. c) Clipping:It means coin a shorter word by clipping the head or the end, or both of a long word, e.g. phone = telephone ( 電話 [dian hua]), demo = demonstration ( 示範 [shi fan]), bus=autobus ( 公汽 [gong qi]), flu =influenza ( 流感 [liu gan]), etc. d) Blending:It aims to blend two or more clipped words so as to coin a new term, e.g. telex/ 電傳 [dian chuan]=teleprinter + exchange; taikonaut/ 宇航員 [yuhang yuan] = taikong + astronaut; netzen/ 線民 [wang min] = net + citizen, etc. The above shortened words are introduced from English, while Chinese people themselves coin a lot of abbreviations in Chinese as well. For example: a) Abbreviations of Chinese character: 春晚 [ChunWan]( 春節聯歡晚會 : Chinese New Year s Day Party), 春運 [ChunYun] ( 春節期間的交通運輸 : Passenger Transport around Spring Festival), 邊貿 [Bian Mao] ( 邊境貿易 : Border Trade), 節育 [JieYu] ( 節制生育 : Birth Control), 消協 [XiaoXie] ( 消費者權益保護協會 : Customer s Right Protection Association), etc. b) Abbreviations of Chinese character Pinyin: GB Guo Biao ( 國標 : National Standard), RMB RenMin Bi ( 人民幣 : Chinese People s Currency), HSK HanYu ShuiPing KaoShi ( 漢語水準考試 :Chinese Language Evaluation Test), 3F 現象 Fuzao, Fuqian, Fukua Phenomena. This kind of abbreviation occurs in flood in cyber language. For example: MM (younger sister); JJ (elder sister ); GG (elder brother ); DD (younger brother), PLMM (pretty girl); WAN (I love you), etc. c) Common English abbreviations for Chinese Netizens: IC I see; IDK I don t know; TKS Thanks; TYVM Thank you very much; CU See you; CUL See you later; BRB Be right back; ROTFL rolling on the floor laughing; GTG Got to go Sound imitation and its motivation Sound imitation consists in using certain signs or words to supersede other words with similar pronunciation so as to attract attention by novelty. Sound imitation, mainly used in cyber-language, includes the following two aspects: 387

401 ASIALEX 2005 a) Expression of numeral sound imitation It refers to those expressions composed of numerals, which can be classified into different groups by the initial number. Among them, the expressions initiated by 5 are in the majority: 514 無意思 [wu yisi] 5376 我生氣了 [wo shengqi le] 584 我發誓 [wo fashi] 嗚嗚嗚 [wuwuwu ] Nonsense 我依舊想你 [wo yijiu xiangni] I am angry 5366 我想聊聊 [wo xiang liaoliao] I swear [wo yaoni quwo] sound of 56 無聊 crying [wuliao] I miss you as before I want to have a chat I want you to marry me That s boring b) Expression of sound imitation in Pinyin These expressions are adopted by netizens to show their personalities. For example: 共眠 [gong mian] (Sleep Together) refers to 共勉 (Encourage Each Other); 幽香 [you xiang] (subtle aroma) refers to 郵箱 (mail box); 竹葉 [zhu ye] (bamboo leaves) refers to 主頁 (homepage); 酷雞 [ku ji] (cool chicken) refers to 酷機 (cool computer); 溫酒屋 [wen jiuwu] (warm pub) refers to WIN95(Windows 95); 溫酒吧 [wen jiuba] (warm bar) refers to WIN98 (Windows 98); etc. There are also some offbeat expressions, e.g. 瘟都死 [wen dousi](all die of the plague)stands for WINDOWS; 瘟酒吧 [wen jiuba] (a bar spreading plagues) stands for WIN98. These two expressions show that the young men of our time try to titillate the senses by uttering a malediction against themselves Loan translation and its motivation Loan translation consists in projecting the source language spaces directly into those of the target language. Generally speaking, Chinese neologisms or technical terms formed by loan translation preserve wholly (or mostly) the signifiant and signifié of the source words. Due to lexical domestication, some words may differ somewhat in their signifiant and signifié. However, they manage to keep the basic features of the source words. 1) Transliteration It is a product of language schema blending, in which the sound of a source language is projected directly into blending space to form a new Chinese word, that is, the pronunciation of the word is directly represented in the form of Chinese Pinyin with its concept being transplanted wholly or mostly to Chinese. Therefore, sound is the only motivation of this kind of word formation. For examples: 卡通 [ka tong] > cartoon, 特氟隆 [Te fulong] > Teflon, 厄爾尼諾 (E Er Ninuo) > EI Nino, 拉尼娜 [La Nina] > La Nina, 克隆 [ke long] > clone, etc. 2)Transplantation It aims to project the source lexical form directly into target language Chinese,i.e. the words as NBA, CEPA, DNA, APEC, CFO, etc. are directly used in Chinese. The motivation for these terms transplantation is cognitive economy, that is, compared with their Chinese counterparts, the transplanted forms are much 388

402 WORDS IN ASIAN CULTURAL CONTEXTS simpler. Lettered words mainly come from English abbreviation (see 4.3), or daily-used words. Note that lettered words usually co-occur with Chinese translations, but the former are relatively far more economical in cognition, easier to pronounce, so that they are more frequently used than their Chinese equivalents. The following is a contrastive analysis of frequency between them. The data are obtained by searching the simplified Chinese web pages with Google Search Engine on April 2, Lettered words Frequency Chinese Equivalents Frequency Ratio NBA 753,000 美國籃球協會 :1 CEPA 239,000 內地與香港更緊密經貿安 773 排 :1 IC 卡 597,000 積體電路卡 :1 3G 2,600,000 第三代移動通信 ( 系統 ) 69, :1 CFO 180,000 首席財務官 58, :1 WTO 1,020,000 世界貿易組織 507,000 2 :1 WTO 1,100,000 首席執行官 681, :1 GDP 895,000 國內生產總值 709, :1 From the above data we can conclude that the ratio of occurrence frequency between the lettered words and their Chinese equivalents are closely related to the cognitive economy and their memorizability. Such simple words with fewer letters as NBA,CEPA,IC 卡,3G and DNA are easy to utter and memorize, while their Chinese counterparts are complicated and hard to remember due to their lengthy components and indistinct lexical features. Thus, the Chinese translations are seldom used. Yet other Chinese translations with comparatively distinct lexical features like 世界貿易組織 (WTO) 首席執行官 (WTO) 國內生產總值 (GDP) are less complicated and easier to understand. So their occurrence frequency is comparatively higher than the others. (See the table) 5.6.Calque and its motivation Intralanguage calque When scientists try to name a new thing or new concept, they usually abstract the semantic features from it, and find out its conceptual prototype through cognitive association and then look for a linguistic sign within the prototypic categories that is closed related to that concept; if such a term exists, they can change one or two morphemes in that term so as to coin a new word representing this new thing or concept. This kind of word-formation is called intralanguage calque. In the development of modern language, calque is one of the major ways to produce new words. For example, 的士 [di shi] is originated from the Cantonese transliteration of the English word taxi, which initially refers to the car for hire but later with the emergence of other motor vehicles for hire, e.g motorbikes, minibuses, trucks and flatbed cart, the words 面的 (minibus for hire), 摩的 (motor-cycle for hire), 貨的 (trucks for hire), 板的 (flatbed cart for hire) are then calqued on the basis of the prototype 的士 (taxi) ; while the original 的士, which is now called 轎的 for distinction, has become a generic word of this category, and 的 that is originally a functional sign in Chinese has got the 389

403 ASIALEX 2005 sense of vehicle for hire. Words formed in such a way have a strong power of family-grouping. For example, after the word 酷 transliterated form cool is widely adopted in Chinese communication, it spreads as fast as a computer virus: we have 酷哥 (cool boys) for males, 酷姐 (cool girls) for females, 酷文 (cool articles) for writing, 酷裝 (cool suit)for clothing; and many others as 酷言 (cool words), 酷站 (cool website), 酷腳 (cool feet), 酷臀 (cool hips), 酷臂 (cool arms), 酷手 (cool hands), 酷腿 (cool legs), 酷胸 (cool bosom), 酷斃 (cool to die/very cool), 酷機 (cool computer), 酷語 (cool expression), etc. This infectivity is naturally the consequence of the family-grouping power of calques Interlanguage calque This is a cross space and cross language mapping of expressional schemas and concepts, which is effective approach to foreign terms translation. The difference between calque and borrowing consists in that calque doesn t indiscriminatingly transplant the expressional form of the source language. On the contrary, it aims to coin new terms for new concept using the word-building components of target language in the light of the structure patterns of the source language. The related studies show that the interlanguage calques are motivated in the following ways: 1) Semantic Calque The conceptual mapping between correspondent expressions, i.e. a new meaning is to be projected from foreign expression onto a corresponding Chinese word. Let s take the information science for example. By imitating the English words notebook and engine, Chinese counterparts 筆記本 and 引擎 have got respectively a new meaning computer as notebook and searching tool ; in the same way, the Chinese word 鼠 and 窗 have acquired separately a new meaning hand-operated device for PC ( 滑鼠 ) and operating environments for PC ( 視窗 ) from their English counterparts. 2) Phono-semantic calque This refers to the cross language mapping between expressional schema and conceptual schema, not only the pronunciation of the source language gets mapped, but also the concept is conveyed. Phono-semantic calque can be divided into the following types according to their formal structure: a) The source language is the name of a product, the target language, based on transliteration, will select the Chinese morphemes that can represent both the quality and function of the product. In this way, not only the sweet-sounding of the source language is retained, but also the meaning is expressed. Let s take the medical expressions for example: 眠爾通 [Mian er tong] > Milton; 帶爾眠 [Daier mian] > Dakmane; 酣樂欣 [Hanle xin] > Halcion; from their Chinese calques, we can easily infer that they are sleeping tablets. More examples are: 息斯敏 [Xisi min] > Hismanal; 鼻可靈 [Bike ling] > Beconase; 利肝靈 [Ligan ling] > ligalon, and 偉哥 [Weige] > Viagra. Their Chinese equivalents show clearly the function of these medical products. This is also true of calques for some pesticides, such as 速滅殺丁 [Sumie shading] > Sumissadin; 禾大壯 [Heda zhuang] > Oedran; 滅草特 [Miecao te] > Machette. b) The source language words are nouns, the target language will artfully rebuild their sounds and meanings. For example, the word cool is transliterated as 酷 390

404 WORDS IN ASIAN CULTURAL CONTEXTS [ku], both the pronunciation and the meaning fashionable and attractive of the source word are vividly portrayed through the conceptual mapping and blending. The same goes with the words like 駭客 [hei ke] > hacker; 紅客 [hong ke] > hongker; 鬅客 [peng ke] > pank; 基因 [jiyin] > gene; and so on. Such a calque is an excellent combination of paraphrase and transliteration. They are the top-grade of the interlanguage calques. c) The source language usually occurs as a single word, the target language has to generalize its conceptual meaning based on transliteration. For example, from the word ballet, we ve got 芭蕾舞 [balei wu]. In this expression, 芭蕾 carries the sound of the source word, and 舞 expresses its concept. Here are some similar examples: 高爾夫球 [gaoerfu qiu] > golf; 桑拿浴 [sangna yu] > sauna; 華爾滋舞 [huaerzi wu] waltz; and so on. d) The source language is a compound word; the target language is to use one expression to convey both its pronunciation and meaning. For example, the word miniskirt is translated as 迷你裙 [mini qun]. The first two characters 迷你 express both the sound and an emergent meaning (cf. Fauconnier 2001), and the third character 裙 just summarizes the referent content of the source language. According to rules of phono-semantic calque, we have coined new the Chinese words: 迷你光碟 [mini guangpan], 網際網路 [Yinte wang], 蹦級跳 [bengji tiao] and 霹靂舞 [pili wu] respectively from the English words: minidisk, Internet, bungee jumping, break dance. 3) Morpho-semantic calque Some languages may make use of certain letters or a word portraying a form to designate a shaped thing. In this case, the mere referential concept is insufficient to convey the whole meaning of such expressions, and it is preferable to adopt an image mapping to calque a word representing the similar form, and that is the morpho-semantic calque. There two kinds of exemplars: a) 丁字尺 [dingzi chi] > T-square; 工字鋼 [gongzi gang] > I-steel; 工字梁 [gongzi liang] > H-beam; 丁字接頭 [dingzi jietou] > T-junction; K 形坡口 [kaixing pokou] > double-v groove. b) 環形圈 [huanxing juan] > O-ing; 三角皮帶 [sanjiao pidai] > V-belt; 交叉支撐 [jiaocha zhicheng] > X-brace; 叉形曲線 [chaxing quxian] > Y-curve. In group (1), it is evident that each pair of the boldface (both in English and Chinese) are similar in shape, while in group (2), the Chinese characters are simply a description of shapes corresponding to the form of the Latin (and hence, English) letters. 4)Phono-morphological calque In Western languages, some terms may contain special codes or hieroglyphs. When those terms are translated into Chinese, this special sign should be retained in order to convey the special information. That is the phono-morphological calque. Note the pairs of equivalents below: X 射線 [aikesi shexian] > X-ray; γ 射線 [gama shexian] > γ-ray; V 形管 [weixing guan] > V-shaped tube; IP 電話 [aipi dianhua] > IP phone; T 恤衫 [tixu shan] > T-shirt; V 形軸封 [weixing zhouxian] > V-packing; E 型彎頭 [yixing wangtou] > E bend; H 型網路 [aiqu xing wangluo] > H-network; X 形支撐 [aikesi xing zhicheng] > X support, etc. 391

405 ASIALEX 2005 Obviously, the phono-morphological calque not only retains partially or wholly the pronunciation and the spelling of the source language, but conveys the distinctive conceptual features as well. (Note: the Chinese character 形 [xing] means the form or shape in English.) 6. Conclusion After the study on the major neologism dictionaries and relevant materials, we have found that the imported words arouse concerns of linguists and lexicographers in China because their exotic features have important effect on Chinese expressions. The neologism occurrence is a process of conventionalization, which is neither created artificially nor let unchecked. The coinage of a new word or expression must conform to the basic requirements such as rational structure, well specified meaning and unsubstitutability. For this purpose, contrived instruction for the use of neologism is a necessity. One of the most efficient tools for the instruction is neologism dictionaries, which play a significant role in formalization, standardization and utilization of new words and expressions. The studies on the motivation or mechanism of word-formation of neologism with a cognitive approach prove to be an important theoretical evidence for the selection of entry words and lexical definition for the dictionary compilation. In the meantime, the motivation provides an effective guidance for the construction of new words and expressions and for the translation of bilingual dictionaries. Notes References Fauconnier, G. (1985) Mental Spaces: Aspect of Meaning Construction in Natural Language. Cambridge, MA: MIT Press. Fauconnier. G. (1998) Mental spaces, language modalities, and conceptual integration. In The New Psychology of Language: Cognitive and Functional Approaches to Language Structure, edited by M. Tomasello. Lawrence Erlbaum. Fauconnier, Gilles and Turner, M. (1998) Conceptual integration networks, Cognitive Science 22: Turner, M, and Fauconnier, G. (1995) Conceptual Integration and Formal Expression, Journal of Metaphor and Symbolic Activity, volume 10, number 3, pages Turner, M. (1996) Conceptual Blending and Counterfactual Argument in the Social and Behavioral Sciences. In Counterfactual Thought Experiments in World Politics, edited by Philip Tetlock and Aaron Belkin. Princeton University Press. Coulson, S. (1997) Semantic Leaps: Frame-Shifting and Conceptual Blending. UCSD Ph.D. dissertation. Oakley, T. (1998) Conceptual Blending, Narrative Discourse, and Rhetoric. Cognitive Linguistics 9:

406 WORDS IN ASIAN CULTURAL CONTEXTS Interlingual lexicography, with special reference to research priorities R. R. K. Hartmann University of Exeter Introduction My aim here is to survey several important current issues in interlingual lexicography, in order to see what has been done and how much more needs to be done until we have a more complete understanding of what is involved, in terms of six main perspectives of metalexicography, and what the implications of all this are for research. On the status of interlingual lexicography I define interlingual lexicography as a complex of activities concerned with the design, compilation, use and evaluation of interlingual dictionaries ; the interlingual dictionary is in turn defined (in the DoL 1998/2001: 75) as (a) type of reference work with information on more than one language. It is a term particularly needed either when the contrast with monolingual dictionary is stressed or when the distinction between bilingual dictionary and multilingual dictionary is considered irrelevant. In the last two decades, we have all experienced, in conferences like this, the rise and rise of our specialization, in terms of a number of topics and dimensions. These include, above all, the status of our field as a professional activity, and what this involves as a set of compilation activities along a range of stages (from recording or data-collection via editing or description to publishing or presentation), but when we think of the prototypical lexicographer working on the prototypical dictionary, it tends to be the famous Johnson, Webster, Murray, Gove, Barnhart and Simpson working on the monolingual general or historical dictionary rather than the anonymous band of dictionary makers slogging away on a bilingual or multilingual project that few have heard of or written about. Closely linked with professional approaches is the academic status of the discipline we recognize lexicography as being. Is it a scholarly activity that forms part of the university syllabus, together perhaps with such subjects as linguistics or education (linked to problems of multilingualism, language planning, language standardization, literacy, etc.), or is it a minority field struggling (like translation studies, terminology studies and the like) to establish itself in higher education? For this interdisciplinary field of lexicography, there are now many more textbooks available than ever before, but interlingual aspects are often ignored in them. We also do not know much about where and how lexicography is taught, how it relates to practical training provided by publishers, and how the development of national and international associations and conferences sponsored by them has helped to consolidate it. For monolingual lexicography, all this has been difficult enough to strive for (and sometimes to achieve), but for interlingual lexicography, which involves two or more languages, there are almost insurmountable hurdles. For a start, dealing with language pairs is bound to be harder than with single languages. Interlingual activities are also more intricate, involving degrees of bilingualism, language acquisition and translation (and consequently require more intercultural and 393

407 ASIALEX 2005 interdisciplinary collaboration). It is not surprising, then, that fewer experts have dedicated themselves to the task, less research has been done and less literature is available on these topics. Although (paradoxically) foreign language learners are often observed to prefer bilingual to monolingual dictionaries, less attention has been given to, and much less is known about, their purpose, their design and their consultation. The result may well be the relatively low status that seems to be attached to interlingual dictionary-making, and even prejudices against it, as demonstrated by Piotrowski s (1989: 72) sarcastic comment, in an anthology published in Singapore, on the opinions of two eminent British linguists implying that students should be weaned away from bilingual dictionaries, as they tend to perpetuate translation and prevent free creative expression in the target language. An additional problem is lack of familiarity with the literature in languages other than the scholars own, with the regrettable result that, as Zgusta (1984: 275) has pointed out, lexicography still is largely compartmentalized by languages and that the flow of information from one center of work to another is not yet what would be desirable. Research perspectives Wherever lexicography has managed to become established, and whenever it has developed elements of self-reflection, sets of principles and codes of practice, we now speak of metalexicography or dictionary research, for which about half a dozen perspectives have been distinguished in the literature, notably dictionary history, dictionary criticism, dictionary typology, dictionary structure, dictionary use, and dictionary IT (Hartmann 2001, 2003). However, it must be admitted that most of the time these tend to focus on the traditions and practices of unilingual rather than interlingual lexicography. And, unfortunately, we do not know enough about whether and how these developments differ by language, country, and culture, as there are practically no comparative studies (one exception is the paper by Huang 1994, a contrastive account of what distinguishes metalexicography in China from that in the West). I cannot devote equal amounts of attention to all these six perspectives (but more examples will be given in the oral presentation). For the first, dictionary history, there is no general pattern for the treatment of dictionary traditions across language pairs. What several of the relevant authors share is their emphasis on plagiarism (Hayakawa 2001). In some of these historical studies, the claim is made that the bilingual dictionary preceded the monolingual one. Certainly there has been a point of view, although it has been challenged, that this was the case for many languages of Europe at the time when they broke free from the domination of Latin. What points can we generalize from some of the tendencies that have been diagnosed? Most studies have concentrated on the general bilingual dictionary (usually they come in two separate parts, although I know of one or two dictionaries that have combined the two parts into a single alphabetic sequence). Other interlingual genres are occasionally mentioned, such as the special-purpose or technical bilingual dictionary, the bilingual dictionary of idiomatic phrases, and the bilingual pronouncing dictionary. Directionality can also be an important consideration, depending on the status of the languages in question and the needs of the potential users. Thus, the beginning of a tradition is typically dominated by the major language (such as Latin in medieval Europe, and English today in most parts of the world), so bilingual dictionaries are designed initially and predominantly for reception and interpretation 394

408 WORDS IN ASIAN CULTURAL CONTEXTS (reading, or decoding, or version in French), while later on the productive uses become more important (writing, or encoding, or thème in French). The overriding impression is that there are enormous divergences between the lexicographic traditions of various language pairs. To turn to the second perspective, dictionary criticism, we note that although there is a relatively long tradition, and several different approaches can be distinguished, the problem is that no-one knows for sure exactly how good (i.e. positive as well as negative) dictionary criticism should be carried out, least of all for interlingual dictionaries. Several criteria or sets of standards for evaluating and assessing dictionaries have been proposed, but only rarely are they applied to interlingual dictionaries. The comparative approach pioneered by Heuberger (2000) for the critical analysis of monolingual ELDs [5 printed and 4 electronic, such as the OALD, with many editions since 1948] has not yet been adapted to the needs of interlingual lexicography. The closest anyone has ever got to this was Iannucci s (1962) critical account of the less than satisfactory treatment of meaning discrimination in bilingual dictionaries, but it is limited to just that: 75 specimen entries in 32 dictionaries with English. So what we can generalize from all this is that we have very few guidelines for reviewing bilingual dictionaries, and those that are available, such as Steiner s (1984) checklist under three headings, inclusiveness, content (including directionality and equivalence), and organization, are not widely known among dictionary critics, and comparative critical accounts of different interlingual traditions are practically non-existent. The third perspective, dictionary typology, deals with the problem of classifying the ever-growing range of reference works. All monolingual dictionaries are potentially useful for interlingual purposes (such as the pioneering dictionary of German regionalisms, the VWBD 2004). Often they contain inherently interlingual information, as in thesauruses and dictionaries of synonyms, where there is a close parallel between explaining the meanings for words in terms of their (intralingual) near-synonyms or finding corresponding words in terms of their (interlingual) translation equivalents in another language. This is why they are often subjected to translation, as in the sub-genres for which bilingualized versions of English dictionaries exist in languages such as Chinese and/or Japanese (e.g. the NODE 1998/2005 or the DoL 1998/2003). The so-called general bilingual dictionary, such as the NACED (1978/2000), is the most well-known interlingual type, but it can be contrasted with more specialized formats (note that the directionality can go either way, or both ways, as in many dictionaries that have two parts addressing users from two languages, cultures, countries and markets). Sometimes more than two languages can be involved, as in polyglot or multilingual dictionaries which have a long-established tradition in fields like law, music, medicine, science and technology. Contrary to general belief, it is not always easier to correlate technical terms across language barriers than it is to find translation equivalents for non-technical general and conversational words and phrases. Finally, elements of all of these sub-genres can be presented in electronic form, or even combined, in a number of different ways. Two recent survey papers are relevant in this context, one written from the Japanese point of view in relation to foreign-language learning, and the other from the Chinese standpoint in relation to translation. Tono (2004) shows us that the typology of various reference materials is 395

409 ASIALEX 2005 confusingly unsettled (from the chip-operated pocket calculator and the CD-ROM-fed computer terminal to various online web-links and parallel-database translation software), while Zhang (2004) demonstrates that the quality of electronic dictionaries is surprisingly deficient, especially in terms of the often inadequate translation equivalents offered in these products. What can we generalize about dictionary typology in relation to interlingual lexicography? It seems to me that we would be justified in saying that almost anything is possible, and therefore we need to keep an open mind on diverse new products, provided that we are confident they appeal and make sense to the average user. The fourth perspective, dictionary structure, is a notoriously difficult and under-researched problem area. Since the classic paper by Haas (1962), compilers have struggled to supply the kinds of information categories that the ideal bilingual dictionary should include, from just the right translation in the target language and all the lexical and grammatical details, levels of usage and personal names to specialised terminology, spelling, pronunciation, etc. The most important point is of course No. 1, translation equivalence. How are lexical equivalents found and codified, how are all their occurrences in real contexts actually illustrated, and how are their various senses properly explained and discriminated? Most of us are fully aware of the fact that we can distinguish several types or levels of equivalence, from complete (or full or exact or absolute) equivalence via partial equivalence and false equivalence to nil or non-equivalence. Some of us may even know how some of these are covered in the dictionaries for a particular language pair, but the considerable variation between different language pairs has not been systematically catalogued, part of the mutual neglect and distrust between the two fields of lexicography and translation studies (Hartmann 2004). I have space here just for one example from a bilingual dictionary compiled at Exeter, the CEDIP (1988) in which the translation equivalents are ingeniously arranged in a 3-step sequence; thus for the Chinese idiom bu ju li jie, the 3 levels are: literal translation: not stick to usual social rules free translation: pay no attention to convention full English equivalent: do not stand on ceremony, the latter marked with the register labels literary and colloquial. So, to generalize on dictionary structure today, more than 40 years after Haas s demands, we no longer regard them as unreasonably optimistic, especially with the means IT has put at our disposal. The fifth perspective is dictionary use, a specialization to which I myself have contributed. First, a few comments on some of the variables within the so-called user perspective. Most publications on the topic of observing and surveying dictionary users start with a reference to the famous American lexicographer Barnhart (1962), whose influential questionnaire survey had established that meaning and spelling outrank grammar and etymology, and may well have contributed substantially to the removal of historical facts from general and learners dictionaries, in favour of stressing semantic and orthographic information. The trend since the 1980s and 90s has been for more direct (rather than indirect) observation, from the relatively large-scale questionnaire survey to the relatively limited direct observation by test or protocol (although large numbers by themselves are no guarantee for reliability, and small numbers can still be useful if they reveal typical behaviour patterns of typical dictionary users). From Barnhart s survey of native- 396

410 WORDS IN ASIAN CULTURAL CONTEXTS speaker students at American colleges we have moved on to the observation of learners of English as a Foreign Language and English as a Second Language, even English for Specific Purposes, and on to British learners of German (Hartmann 1982) and many other language learners around the world. What can we generalize from these studies? The earlier research efforts were concerned with more general notions of dictionary reference needs, but gradually they have tended to focus on more specific instances of dictionary reference skills associated with particular activity contexts. But practically no studies exist of the multifarious uses and users of technical dictionary genres, such as LSP experts. Even less is known about translators as dictionary users, as most studies devoted to dictionary use while translating are based on students of translation rather than professional translators. Let us hope that the trend towards more empirical research methods such as thinking-aloud protocols and experiments, as documented by Tono (2001), will continue. The sixth branch or perspective is dictionary IT, or computational lexicography. I have already hinted that IT has made it possible not only to mechanize many lexicographic processes (e.g. word-processing and corpus technology), but also that many new types of reference works have been developed, under such names as electronic dictionary and terminological database. However, interlingual lexicography has not yet fully explored all the infinite possibilities, it also lags behind unilingual lexicography, and the literature often does not refer to bilingual dictionaries and translation at all, including the problem of how to treat (Haas s) multiple information categories. Research priorities We have seen that lexicography has matured, both in practice and in theory, in terms of more professional approaches to the discipline, more textbooks, more training facilities and more international associations. Some generalizations have resulted from the six perspectives of dictionary research: dictionary history, dictionary criticism, dictionary typology, dictionary structure, dictionary use, and dictionary IT. Now we need to ask about what sorts of desiderata remain. Among the factors of interlingual dictionary use and dictionary provision that need to be investigated more fully are the following 10 problem areas: (1) What do we know about varying levels of mother-tongue and foreign-language proficiency of the dictionary user, from beginner s to advanced level? We need more surveys of dictionary users, e.g. of the comparative sort that arose from the European-Union funded Thematic Network Project, which looked at problems of language learning in higher education over a period of 3 years, and was published as a European Language Council report (Hartmann 1999). Specifically, its Sub-Project 9 concerned with dictionaries found that dictionary awareness among university students, at least in most European Union countries, is still too limited, and also made a number of recommendations to counteract this deficiency. (2) What are the types of activity engaged in by the user, from reading to writing or translating from one language to another? I have already referred to foreign-language learners and translators, but there are other user groups, such as scientists, writers and journalists about whose reference needs and reference skills 397

411 ASIALEX 2005 we know hardly anything at all; some are mentioned in the collection of papers edited by Atkins (1998). (3) What is the degree of knowledge sought, from general information to technical expertise? On linguistic as well as encyclopedic reference needs, there is only a limited literature available. The terminology of law could serve as an example here, especially legal translation, the topic of a recent book by Chromá (2004) who, from the vantage point of the Czech Republic, looked at the possibilities and requirements of a new bilingual dictionary of the language of law, to assist in this area of special-purpose communication. (4) What genres of reference works are on offer, from general-purpose to special-terminological, or from monolingual to interlingual, or from print to electronic? IT has begun to widen the options, producing new types, and even hybrids of existing types, such as the so-called bilingualized dictionary discussed by Thumb (2004) from the point of view of look-up strategies of Hong Kong learners of English. (5) Which information categories are to be made available in the dictionary, from spelling and pronunciation to lexical meaning, usage and encyclopedic details? The only (comparative) treatment of the structure of bilingual dictionaries, Marello s textbook (1989), gives examples of various macrostructural and microstructural profiles in the bilingual dictionaries of the 4 language pairs covered, but this does not actually show all the ways in which all the information categories mentioned by Haas (1962) have been or should be presented. Among the neglected topics are pictorial illustrations (Hupka 1989) and typology/art/design (Luna 2004). (6) Are we familiar with all the types and directionality of the translation process, from mother tongue to foreign language, or vice versa, or from literal to free and idiomatic, including problems of sense discrimination and equivalence? I have mentioned some examples, and could add that Svensén s book (1993, especially its brand-new Swedish 2 nd edition 2004), is particularly helpful in giving many examples of partial equivalence, related to the problem of whether you are going from L1 to L2, or from L2 to L1. And more could be said about the subject of how lexical equivalents can be established by means of contrastive textology or comparable parallel-text analysis (Hartmann 2005). (7) What is the sequence of operations in a typical consultation process, from choice of dictionary to search within the entry and integrating the result of the operation with the requirements of the activity? On the reference skills necessary to look up information in the dictionary, I can add the paper by Nesi (1999), although it still neglects interlingual aspects. (8) Which intradisciplinary methods are appropriate for carrying out research on the topics listed above? For all these we need to elaborate appropriate research techniques within lexicography and dictionary research (Hartmann 2001), although for each of its perspectives (dictionary history, dictionary criticism, dictionary use etc.) specific procedures may still need to be developed. (9) What interdisciplinary methods can we draw on for such studies? One defining feature of all lexicography around the world is its so-called reference base. In our search for knowledge out of information, a basic human need, we all turn to reference books such as dictionaries, handbooks, directories and encyclopedias, monolingual and bilingual, general and technical, print and electronic. In the last few years, I.T. has indeed helped to strengthen these developments, so much so that 398

412 WORDS IN ASIAN CULTURAL CONTEXTS there is now a strong case for a so-called reference science, defined by McArthur (1998: 218) as the study of all aspects of organizing data, recognizing at least 3 branches (traditional lexicography, encyclopedics, and a third that does not really have a name yet), each with many connections to other disciplines. (10) What are the implications of all of this not only for lexicography, but also a number of other fields such as translation studies, technical terminology and foreign-language learning? The problem gets ever more complicated, as there we take on the role of responsible representatives of our discipline, which (almost like missionaries) we have to explain to others, in education (e.g. teacher training), in media studies, even in politics. And this brings me back to the example I gave in point No. 1: what is the relevance of the observations and the recommendations we made under the banner of the European Thematic Network Project for the rest of the world, or what might be the relevance of what we do in English lexicography and dictionary research for other languages, and is what we know about interlingual dictionaries for one language pair such as English and German relevant to the lexicography of another language pair such as English and Chinese? Conclusion I have surveyed the scene of interlingual lexicography and metalexicography, and I hope I have managed to diagnose a few trends, such as professionalization, theory formation and research methods. I trust you will share with me the realization that we need to do more in all these areas. I recently acquired the LBCP (2001), a monolingual dictionary of Chinese proverbs presented through the medium of English, from which I finally quote the following by HAN Fei Zi (Warring States Period) as a motto for the research programme advocated here: Claiming certainty without corroborating evidence is stupid. Bibliography (a) Cited dictionaries and other reference works: CEDIP = (A) Chinese-English Dictionary of Idioms and Proverbs comp. by HENG X.J. & ZHANG X.Zh. (Lexicographica Series Maior 24). Tübingen: M. Niemeyer 1988 DoL = Dictionary of Lexicography comp. by R.R.K. Hartmann & G.C. James. London: Routledge 1998/2001 [Japanese translation Jishugaku jiten ed. by Takebayashi S. Tokyo: Kenkyusha 2003; Chinese translation ed. by YONG H.M. et al. in preparation] LBCP = The Little Book of Chinese Proverbs comp. by J. Clements. Bath: Parragon 2001/2004 NACED = New Age Chinese-English Dictionary/Xinshidai Hanying Dacidian ed. by WU J.R. & CHENG Z.Q. Beijing: The Commercial Press 1978/2000 NODE = (The) New Oxford Dictionary of English ed. by J. Pearsall. Oxford: Oxford University Press 1998 [Chinese bilingualization Xin Niujin Yingyu Cidian (Yinghan Shuangjieben). Shanghai: Foreign Language Education Press due 2005] OALD = Oxford Advanced Learner s Dictionary of Current English comp. by A.S. Hornby. Oxford: Oxford University Press [1 st edn 1948, 6 th edn 2000 ed. by S. Wehmeier. CD-ROM edn Chinese bilingualization ed. by WU X.Z., CHEUNG F.K. (et al.). Taipei/Hong Kong 1966/1970/1984/1986/1994/1997] VWBD = Variantenwörterbuch des Deutschen. Die Standardsprache in Österreich, der Schweiz und Deutschland sowie in Liechtenstein, Luxemburg, Ostbelgien und Südtirol ed. by U. Ammon et al. Berlin: W. de Gruyter

413 ASIALEX 2005 (b) Other literature: Atkins, B.T.S. ed. (1998) Using Dictionaries. Studies of Dictionary Use by Language Learners and Translators (Lexicographica Series Maior 88). Tübingen: M. Niemeyer Barnhart, C. (1962) Problems in editing commercial monolingual dictionaries, in Householder & Saporta (1962), [reprinted as Chapter 14 in Hartmann 2003 Vol. I: ] Chan, S.-W. ed. (2004) Translation and Bilingual Dictionaries (Lexicographica Series Maior 119). Tübingen: M. Niemeyer Chromá, M. (2004) Legal Translation and the Dictionary (Lexicographica Series Maior 122). Tübingen: M. Niemeyer Haas, M. (1962) What belongs in the bilingual dictionary?, in Householder & Saporta (1962), Hartmann, R.R.K. (1982) Das zweisprachige Wörterbuch im Fremdsprachenerwerb, in Studien zur neuhochdeutschen Lexikographie Vol. II ed. by H.E. Wiegand (Germanistische Linguistik 3-6/80). Hildesheim: G. Olms, ed. (1999) Dictionaries in Language Learning. Recommendations, National Reports and Thematic Reports from the TNP Sub-Project 9: Dictionaries. Berlin: Freie Universität (2001) Teaching and Researching Lexicography (Applied Linguistics in Action Series). Harlow: Longman/Pearson Education ed. (2003) Lexicography: Critical Concepts [3 volumes]. London: Routledge/Taylor & Francis (2004) Lexicography and translation, in Chan (2004) 7-21 (2005) Interlingual references: On the mutual relations between lexicography and translation, The Hong Kong Linguist 25 (2005 i.p.) Hayakawa, I. (2001) Methods of Plagiarism. A History of English-Japanese Lexicography. Tokyo: Jiyusha Heuberger, R. (2000) Monolingual Dictionaries for Foreign Learners of English: A Constructive Evaluation of the State-of-the-art Reference Works in Book Form and on CD-ROM (Austrian Studies in English 87). Vienna: W. Braumüller Householder, F.W. & Saporta, S. eds. (1962) Problems in Lexicography (Publications of Indiana University Research Centre in Anthropology, Folklore, and Linguistics 21). Bloomington IN: Indiana University Press Huang, J.H. (1994) Chinese and Western metalexicography, in James (1994), [reprinted as Chapter 67 in Hartmann 2003, Vol. III: ] Hupka, W. (1989) Wort und Bild. Die Illustrationen in Wörterbüchern und Enzyklopädien (Lexicographica Series Maior 22). Tübingen: M. Niemeyer [Chapter 8, Die Funktion und Leistung der Illustration neben den anderen Bestandteilen des Wörterbuchartikels eine Fallstudie, reprinted in English translation as Chapter 66, How pictorial illustrations interact with verbal information in the dictionary entry: a case study, in Hartmann 2003, Vol. III: ] Iannucci, J.E. (1962) Meaning discrimination in bilingual dictionaries, in Householder & Saporta (1962, [reprinted as Chapter 10 in Hartmann 2003, Vol. II: ] James, G.C. ed. (1994) Meeting Points in Language Studies. A Festschrift for Ma Tailai. Working Papers. Hong Kong: Hong Kong University of Science and Technology, Language Centre Luna, P. (2004) Not just a pretty face: The contribution of typography to lexicography, in Proceedings of the Eleventh EURALEX International Congress, Lorient, July 6-10, 2004 ed. by G. Williams & Sandra Vessier. Lorient: Université de Bretagne Sud, Faculté des Lettres et des Sciences Humaines, Vol. III: Marello, C. (1989) Dizionari bilingui con schede sui dizionari italiani per francese, inglese, spagnolo, tedesco (Fenomeni Linguistici 6). Bologna: Zanichelli [Chapter 1, Definizione, storia, bidirezionalità, reprinted in English translation as Chapter 44, The bilingual dictionary: definition, history, bidirectionality, in Hartmann 2003, Vol. II: ] McArthur, T. (1998) What then IS reference science?, in Living Words. Language, Lexicography, and the Knowledge Revolution. Exeter: University of Exeter Press, [reprinted as Chapter 69 in Hartmann 2003, Vol. III: ] 400

414 WORDS IN ASIAN CULTURAL CONTEXTS Nesi, H. (1999) The specification of dictionary reference skills in higher education, in Hartmann (1999), [Thematic Report 3, reprinted as Chapter 20 in Hartmann 2003, Vol. I: ] Piotrowski, T. (1989) Monolingual and bilingual dictionaries: Fundamental differences, in Learners Dictionaries: State of the Art ed. by M.L. Tickoo (Anthology Series 23). Singapore: SEAMEO Regional Language Centre, Steiner, R. (1984) Guidelines for reviewers of bilingual dictionaries, Dictionaries 6: Svensén. B. (1993/2004) Practical Lexicography: Principles and Methods of Dictionary-making. Oxford: Oxford University Press [2 nd Swedish edn Stockholm: Norstedts Akademiska Förlag] Thumb, J. (2004) Dictionary Look-up Strategies and the Bilingualised Learner s Dictionary (Lexicographica Series Maior 117). Tübingen: M. Niemeyer Tono, Y. (2001) Research on Dictionary Use in the Context of Foreign Language Learning. Focus on Reading Comprehension (Lexicographica Series Maior 106). Tübingen: M. Niemeyer [Chapter 4, Research on dictionary use: Methodological considerations, reprinted as Chapter 21 in Hartmann 2003, Vol. I: ] (2004) Research on the use of electronic dictionaries for language learning: Methodological considerations, in Computer-mediated Lexicography in the Foreign Language Learning Context ed. by M.C. Campoy Cubillo and M.P. Safont Jordà (Col lecció Estudis Filològics 18). Castelló de la Plana: Publicacions de la Universitat Jaume I, Zgusta, L. (1984) Review of Studien zur neuhochdeutschen Lexikographie Vols. I - IV ed. by H.E. Wiegand (in Germanistische Linguistik ), Dictionaries 6: Zhang, Y. (2004) An empirical study of electronic dictionaries and translation software, in Chan (2004), Socio-psychological motivation for code-switching/mixing: the Maithili-Hindi situation Girish Nath Jha Jawaharlal Nehru University girishj@mail.jnu.ac.in Abstract Code-switching/mixing (CS/CM) is a very common phenomenon among bi/multilinguals under different socio-psychological motivations. In a country like India, where multi-lingualism is a norm, switching from one language to another or mixing two codes is part of people s everyday use of language mostly spoken. The Mithila region in north India presents a curious socio-linguistic case for linguists because of its sociolinguistic diversity. The use of Hindi in this region follows different courses as the state and central official government communication medium, as primary medium of instruction in most schools and colleges, as the vehicle of print and electronic media including films and music, as a tool for bilingual communication, and also by way of CS/CM. The goal of the paper is to understand the phenomena of code-switching and mixing by the Maithili speakers of this region and try to ascertain the complex socio-psychological or contextual reasons which characterize its prevalence and persistence. Review of literature CS/CM can be understood as a sudden or pre-meditated switch on the part of a bilingual to another code either inter-sententially or intra-sententially. While CS is inter-sentential, CM is intra-sentential. Some examples follow- 401

415 ASIALEX 2005 (1) I told him that Ram bahut bimar hai Ram is very sick (Di Sciullo, et al 1986) (2) a. to ghari ala ani enjoyed the icecream he home came and He came home and he enjoyed the ice-cream b. to ghari ala ani (tyane) ice-cream avdine khalle He home came and he ice-crem with pleasure He came home and he enjoyed the ice-cream (Pandharipande, 1987:6-7) In the data above, all except (2b) are examples of sentential CS. (2b) has a lexical item embedded in the Marathi sentence. So while (2b) can be called an example of CM, the rest are examples of CS. More often than not CS/CM turns out to be a communicational necessity in bilingual societies. Let us look at the following conversation from Kachru (1982a:33)- (3) a. [On the telephone] When will you come? b. [to his wife in Kashmiri] me ditav kagaz pensali. Please give me paper and a pencil c. [On the telephone] What is your address? d. [to his children in Hindustani] shor mat karo bhai zara chup karo Don t make noise, keep quiet (Kachru, 1982a:33) The person on the phone is someone not known to the family, and since English is the default language in a multilingual country like India, the communication medium in this case is English. The husband and wife appear to belong to the Kashmiri linguistic community, so the husband uses his native language to communicate with his wife. The kids are more fluent in Hindi. Therefore (3d) is in Hindustani and neither in English or Kashmiri. Let us look at this Maithili-Hindi CS example. (4) a. [to father in Maithili] kakaji, ee keen na liya father why don t you buy this b. [to the shopkeeper in Hindi] iska kitna keemat hai bhai? what is the price of this (object), brother? (4) is an example of necessary code-switching. Since the shopkeeper may not understand Maithili, he is being addressed in Hindi in (4b). Swigart (1992), Goyvaerts & Zembele (1992) believe that CS/CM can be both a norm or exception depending on what kind of linguistic community it is. According to Valdes-Fallis (1977), code-switching is the use of two languages simultaneously or interchangeably. It implies some degree of competence in the two languages even if bilingual fluency is not yet stable. Gumperz (1982) describes code-switching as discourse exchanges which form a single unitary interactional whole. Corder, (1981) has found it to be a rule-governed behaviour and a communication strategy. Gysels (1992) states two purposes for CS/CM for filling a linguistic / conceptual gap, or for other multiple communicative purposes. CS/CM can be conversational and metaphorical. Studies by Labov (1971) and Gumperz and Hernandez-Chavez (1975) are remarkable in this context. Some case-studies Grosjean and Soares (1986) observed that a bilingual has the choice of activating both CS/CM or of deactivating one and activating the other in a monolingual context. They propose with Obler and Albert (1978) a general language monitoring device that is flexible, rapid, and automatic, and they agree with Paradis (1980) that the bilingual has two language lexicons, each of which is connected to one conceptual store (Paradis does 402

416 WORDS IN ASIAN CULTURAL CONTEXTS not posit a bilingual monitoring device). Gibbons (1983) studied language attitudes and code-switching in Hong Kong between Cantonese and English. They indicate that when Chinese speakers use English with one another they give an impression of status and westernization. When they use Cantonese they give an impression of Chinese humility and solidarity. However, a mix was considered ill-mannered, show-off, ignorant, aggressive, and proud from the Cantonese point of view. Stevens (1983) for French-Arabic and Chana (1984) for Punjabi-English studied listeners evaluative reactions to code-switched speech. Sridhar and Sridhar (1980) studied Kannada-English and found that there was a basic language bilingual discourse and proposed the terminology of guest and host languages. Lipski (1982) speaks of a bilingual grammar especially during intra-sentential code-switching in written text. There are some other recent case studies done by Cozens (1996), Bediafi (1997). Constraints/principles of CS/CM Given that CS/CM are well motivated by bilingual speakers, there are certain constraints/principles which underlie such usages. Poplack (1980:586) talks about an Equivalence Constraint (EC) according to which code-switches will tend to occur at points in discourse where the juxtaposition of L1 and L2 elements does not violate a syntactic rule of either language. Poplack (1980: 585-6) also puts forth a Free Morpheme Constraint (FMC) which states that codes may be switched after any constituent in discourse provided that [that] constituent is not a bound morpheme. Di Sciullo, Muysken and Singh (1986) come up with Government Constraint (GC) on code-switching, according to which code-switching within a governed element is possible if and only if the governed element includes a language-carrier whose language index is identical with the language index of the governor. The Adjective [Noun] Phrase Constraint (ANPC) of Pfaff (1979: 306) states that adjective/noun mixes must match the surface word order of both the language of the adjective and the language of the head. Pfaff (1979: 303) also came up with Clitic Pronoun Constraint (CPC) according to which clitic pronoun objects are realised in the same language as the verb to which they are cliticised, and in the same position required by the syntactic rules of that language. The Dual Structure Principle of Sridhar and Sridhar (1980: 412) states that the internal structure of the guest constituent need not conform to the constituent structure rules of the host language, so long as its placement in the host sentence obeys the rules of the host language. Rakesh Bhatt (1995) develops an optimality-theoretic account (Prince and Smolensky, 1993) for the observed patterns of intra-sentential code-switching (ICS), the proposal is premised on the simple assumption that ICS strives for well-formedness. It begins with the assumption that when the embedding constituents are mixed into the matrix language, the syntax operates to optimize well-formedness(wf). Maithili-Hindi Maithili speakers are very fluent bilinguals and most of them would have almost equal fluency in both Maithili and Hindi. Sometimes, Maithili is the language with a restricted domain of use than Hindi for most city bred individuals in the Mithila region. For most Maithili speakers of third generation, Hindi is as natural as Maithili, 403

417 ASIALEX 2005 sometimes even better. In certain domains, they may be more fluent and expressive in Hindi than Maithili. But for those of the first or second generation who did not have much contact with towns or cities, Hindi is not a fluent language, though very well understood. Therefore, the phenomenon of code-switching and mixing takes place under certain socio-psychological contexts which can be outlined as follows- Context of switching After a careful study of the contexts in which most Maithili speakers are likely to take recourse to code-switching of one kind or another, the following have been listed as potential characterization of the phenomenon. For intense agitated expression It has been a very common observation that Maithili speakers (if they know sufficient Hindi) can not quarrel or fight in Maithili. They code-switch to Hindi as can be seen in the following example- (5) ekhan ahaan san keechh nai kahab. Lekin baad men main tumko dekh loonga now you (formal) to nothing say(1-per, fut.) but later I to you see take (fut) I wont say anything to you now, but later I will teach you (familiar) a lesson Such switches have been noticed from Hindi to English also. There can be several reasons for this. First may be the femininity of the Maithili language which makes it hard to express violent and angry expression with the same effect as could be done in Hindi or English. Another reason could be the very alien nature of the code to be switched which gives the impression of more distancing between the people quarrelling. However, the case may be different for those ladies from villages who are not fluent in Hindi. Because when they quarrel they do it with all possible special effects and metaphors. Speaking to a Maithili and Hindi speaker simultaneously A Maithili speaker may often be seen doing a difficult balancing act when engaged in a multi-cornered communication. A very interesting situation is being cited here. My younger brother s two European friends came to visit my village and lived there for sometime in our family. My younger brothers speak Hindi among themselves and Maithili with other members of the family. So these visitors were being communicated to everybody in the family in Maithili, but when the brothers talked among themselves they switched to Hindi. (6) Gautam: (to grandmother in Maithili) Dadi, nai bujhaliye, ee kahait chhathinh je hinka desh men besi jaarh parait chhai (Dadi, you did not understand. These two gentlemen are saying that in their country they have more cold than here). (turning to his younger brother) Dadi ko kaise samjhayen? Inko lagta hai ki duniya men isase jyaada thandh kahi par hi nahi sakti. (How to make her understand? She thinks that nowhere in the world can have more cold winter than this place) (turning to his European friends) Well, she wouldn t believe this. You should have brought some snow in your bag. This kind of linguistic jugglery is purely need based, and involves a lot of interfacing and planning on the part of a bi/multilingual. Teacher-student communication in elementary schools Most elementary schools in the region suffer from this strange linguistic predicament. While the official medium of instruction and communication on campus is to be Hindi, the teachers and students are often seen using CS/CM for their ease of 404

418 WORDS IN ASIAN CULTURAL CONTEXTS expression while perhaps trying to maintain their linguistic identity. (7) teacher: (to a naughty student): kee re (Maithili) kya kartaa hai(hindi) (hey! what are you doing? Teaching Hindi to children by parents/teacher Hindi is often the first language for the school-going children in the Mithila region. In school or at home, when they are studying Hindi, the medium generally happens to be Maithili. This leads to necessary CS/CM situation as in the following example, (8) father: (to daughter) : Maitreyi, Hindi me dahi pum-ling hoit ye. Hindi me ekar sahi vaakya hoit dahi khattaa hai. Bujhalaun? (Maitreyi, yoghurt is masculine in Hindi. The correct sentence in Hindi would be (the sentence). Did you understand? Reporting a communication in Hindi This is a typical case of CS in bilingual societies. The switch often occurs in Hindi in the Mithila area. Let us look at the following example (9) kaail ham Patna men dost ohithaam gel chhalaun. jakhan vida bhelaun ta roik lelain. kahlain aaj bhar ruk jaao, kal chale jaanaa (in Maithili yesterday I had gone to my friend s place in Patna. When I departed, he stopped me by saying (Hindi) please stay today. You can leave tomorrow. Conclusion The CS/CM in the bilingual Mithila region follows a certain pattern and can be categorized as optional and necessary under different socio-psychological contexts. These contexts are intense agitated expression, speaking to Maithili and Hindi speakers simultaneously, classroom situation where the medium is Hindi but the local language is Maithili, bilingual instruction and reporting a situation verbatim. When such a switching/ mixing takes off from Maithili to Hindi, the quality of Hindi sentences and grammar seems to be heavily influenced by Maithili as can be seen in the following example, (10) re narena ki karai chhe? (when after repeated calls the servant does not respond) katai mair giya re? (kahaan mar gaya re) : where have you died? References Bhatt, Rakesh M. (1995), Constraints, optimality and code switching in Linguistic Society of America, Annual Meeting, New Orleans Bokamba, Eyamba G. (1988), Code mixing, language variation and linguistic theory, Lingua 76: Bokamba, Eyamba G. (1996), Classnotes, Lin350, Fall 96, UIUC Cozens, P. (1996), Code choice, domain configurations, code-mixing and code-switching: a study of language use in the Brigade of Gurkhas in Hong Kong, dissertation, Applied Linguistics Research Group. Gumperz, J.J. and Hernandez, Edward (1969), Cognitive aspects of bilingual communication (Berkeley: Institute of International Studies, University of California). Gumperz, J.J. (1982), Discourse strategies (Cambridge: Cambridge University Press). Kachru, Braj B. (1982), The bilingual s linguistic repertoire, in B. Harford and A. Valdman (eds), Issues in international bilingual education: the role of the vernacular (New York: Plenum). Kachru, Braj B. (1976), Dimensions of bilingualism, theory and case studies, in Proceedings of the conference, UIUC on May 8-9, 1975, Urbana-Champaign (Unit for Foreign Language Study and Research, University of Illinois). Labov, W. (1972), Sociolinguistic patterns (Philadelphia: University of Pennsylvania Press). Labov, W. (1994) Principles of linguistic change (Oxford: Blackwell) Pandharipande, Rajeshwari, V., (1986), Language and language variation: Nagpuri Marathi, in B. Krishnamurti (ed.), South Asian languages: structure, convergence and diglossia (Motilal Banarasidas, New Delhi). 405

419 ASIALEX 2005 Pandharipande, Rajeshwari V. (1998), Is genetic connection relevant in code-switching? Evidence from South Asian languages, in Rodolfo Jacobson (ed.), Code switching worldwide: trends in linguistics: studies and monographs 106 (Berlin: Mouton De Gruyter). Pfaff, Carol (1979), Constraints on language mixing, Language 55: Sridhar, S. N. and Kamal K. Sridhar (1980), The syntax and psycholinguistics of bilingual code-mixing, Canadian journal of psychology 34: Misinterpretation analyses of data from 2004 pop song translating competition comparing electronic handheld dictionaries and printed dictionaries users Abstract Miwa Nakamura Seiko Instruments Inc.(SII) This study investigates the comparisons of misinterpretations between handheld electronic dictionary (ED) and printed dictionary (PD) occurring, which were analysed from the marketing data of the 2004 Pop Song Translating Competition in Japan. This competition was held by Seiko Instruments Inc (SII)., the pioneer maker of handheld dictionaries in Japan. For the past three years, the use of handheld dictionaries in high school has taken a steep increase, which indicates growing importance and need for handheld dictionaries in schools. As such it is imperative that the dictionary manufacturers seek to understand the needs of this growing market and continually develop user-friendly dictionaries. This comparative study found two plausible results. First, the limited size of ED screens did not pose a significant factor in causing misinterpretations by the user. Second, for PD users, there was high tendency to skip words more frequently than ED users when they met unknown words while translating. Introduction The electronic dictionary market in Japan has been rapidly increasing for the past three years. The total size of the Japanese electronic dictionary market has grown to approximately six billion US dollars according to SII research. As for the dedicated electronic dictionary targeted especially for senior high school learners, the market has dramatically grown more than 200 percent compared with one year ago. One of the reasons for the rapid growth is that senior high school teachers began allowing electronic handheld dictionaries to be used in classrooms. Some schoolteachers require all the students to have EDs. Others provide alternatives so that the students and their parents can choose inexpensive PDs or lower cost EDs. In 2004, SII held the first annual Pop Song Translating Competition for Japanese high school students. The aim of this competition is to motivate students to enjoy studying English and provide stimulating dictionary use whether electronic or printed. SII selected five popular songs that were provided by the Toshiba EMI record company, which is one of the sponsors of the competition. The five songs were; 406

420 WORDS IN ASIAN CULTURAL CONTEXTS Blondie s Call me, Culture Club s Karma chameleon, Paul Hardcastle s 19 (Nineteen), A Taste of Honey s SukiyakiI and KC & The Sunshine Band s Shake your booty. From this selection, applicants were allowed to choose whatever songs they preferred to translate up to five songs. The contest opened in April 2004 and was promoted by direct mail advertising to high schools, posters in stores and an official website to download application forms and lyrics. Remarkably over 18,000 entries were received. About 400 high school teachers let their students participate in the competition either working in their classrooms or assigning it as homework during their summer vacation. The review of student entries found frequent misinterpretations. Surprisingly, they were very similar types of misinterpretations. Purpose of this study The purpose of this study was to compare EDs and PDs user misinterpretations of word definitions caused by user interface and content structure. Research Topics 1. To determine whether or not limited screen sizes in ED led to incorrect translations 2. To determine whether ED users have higher tendency to adopt or reference the first displayed definition of the looked up word regardless of inappropriate context usage. Subjects 173 applications containing translations of the songs 19 were studied. 110 entries students used EDs whereas 63 used PDs. English proficiency levels were not analysed for the both groups, nor the sex and grade level since they were random samples. Materials In the subject song of the 19 (Nineteen), the following part was focused; According to a veterans administration study Half of Vietnam combat veterans Suffer from what psychiatrists call Post traumatic stress disorder (19 by Paul Hardcaslte) Specific attention was paid to the following four words and phrases; (1) according to (2) study (3) suffer (4) call. Data collection methodology Students recorded the kind of dictionaries used while translating the song. The entries without dictionary information were eliminated from this research. The 110 ED users were categorized as one group since the English<>Japanese dictionary was the same and their functions were very similar across all ED makers. Out of 91 PD users, 407

421 ASIALEX used the same Taishukan s English<>Japanese dictionaries. The remaining of 28 PD users were eliminated from the study because they used the different kinds. Data Analysis Table 1 below shows the correct usage and students misinterpretations. The misinterpretations were categorized into two parts; (1) lack of readability, (2) no response. The misinterpretations which referred to the first word of the definition were categorized as lack of readability. Also, the misinterpretations where the word was taken from the etymology descriptions were grouped into the category of lack of readability. On the other hand, in case of skipping the target words, these were categorized as no response. The only exception was students who mistakenly translated the word call as to make a phone call were categorized as relying on their existent knowledge, since this exact definition appears towards the bottom of EDs and PDs headwords. Target w ords C orrect context usage (1) according to ~ によると 一致して, 従って =as stated by ~ =to agree w ith 調和 =harm ony (2) study 研究 勉強する 学習 =research =to learn 努力 =effort Possible source of the m isinterpretation Error categories 空白 =no response no response no response 空白 =no response no response no response (3) suffer 苦しむ 経験する =affected by a disease =experience 支える =to support 空白 =no response no response no response (4) call が~と呼ぶ 言う ~を呼ぶ 呼んでくる =describe Misinterpretations =order to com e 電話 =to telephone the first w ord definition of "according" lack of readability the first w ord definition of "accord" the first w ord definition The etym ology w ritten just under headw ord "study" the first w ord definition The etym ology w ritten just under headw ord "suffer" the first w ord definition existent know ledge? lack of readability lack of readability lack of readability Relying on existent know ledge 空白 =no response no response no response Table 1. Correct usage, misinterpretations and error categorization Results Table 2 shows the number of the correct usage and the incorrect usages which were categorized as lack of readability and no response. The chi-square tests were performed in the case of four target words. The result showed in Table 2 that there is no 408

422 WORDS IN ASIAN CULTURAL CONTEXTS significant difference between EDs and PDs except for word suffer. (χ2(2)=9.404, p<.01). Then a residual analysis was conducted as shown in Table 3. In case of no response, there was a significance between EDs and PDs. (1) according to ED PD Total correct usage lack of readability no response (χ2(2)=0.1814,p.>0.1) Total (2) study E D P D T otal correct usage lack of readability no response (χ2(2)=3.1365,p.>0.1) Total (3) suffer E D P D T otal correct usage lack of readability no response (χ2(2)=9.404,p<.01) Total (4) call ED PD Total correct usage lack of readability no response (χ2(2)=3.4245,p.>0.1) Total Table 2. Correct usage, lack of readability and no response data (3) suffer ED P D residual >1.65 p<.10 * C orrect answ er * residual >1.96 p<.05 ** Lack of readability ** residual>2.58 p<.01 *** N o response *** Table 3. Residual Analysis Discussions With regard to research topic 1, this study possibly found that the limited screen size of the EDs might not lead to misinterpretations. However, if it is a totally new word for the student probably, like suffer, they might pick up the first word on the dictionary screen. In a future study, we should examine which words students looked up. 409

423 ASIALEX 2005 We found that more PD users tend to skip words (=no response) than ED users. Furthermore, the ED s easy access might help students to actively look up words. However, once ED users enter a target word, there is a tendency for students to mistakenly pick the first word in the screen. In order to decrease the possibility of misinterpretations, the ED screen and dictionary content structure need to be improved more. There may be plenty of rooms for further improvement to decrease selection errors. Limitations of this study There are two limitations in this study. First of all, the subjects were random samples which were not controlled in terms of their proficiency and pedagogic circumstances. The students language proficiency level might affect the usability of the dictionary information. Secondly, this research could not capture words they looked up. For comparatively easy words like study and call, the students may have remembered the meaning and they might have never looked them up such and just applied their knowledge without questioning. Conclusions The dramatic increase in use of EDs by high school students has had an affect on learning behaviour. We, as a manufacturer, have a responsibility to make this change over more effective. Up until now, we designed EDs the same way as the PD does, however, for the future, we should consider the useful user interface to inspire users learning activity. Acknowledgement We would like thank Professor Tono (Meikai Univ. Japan) for giving his advice and guidance for this study. References 田中敏 山際勇一郎著 (1989), ユーザーのための教育 心理統計と実験計画法 ( 教育出版 ). Hattori Atsushi (2005), The Effectiveness of Electronic Dictionaries on English Reading Comprehension and Vocabulary Learning with Japanese High School Students (3), unpublished Licentiate thesis (Graduate School of Nagoya Gakuin University) Satsuki Osaki, Natsue Ochiai, Tatsuo Iso and Kazumi Aizawa(2003), Electronic Dictionary vs. Paper Dictionary: Accessing the Appropriate Meaning, Reading Comprehension and Retention, ASIALEX 03 Tokyo Proceedings Dictionaries and Language Learning: How can Dictionary Help Human & Machine Learning?: Words & Music by Paul Hardcastle, Mike Oldfield, W Couturie and J McCord Copyright 1985 by Oval Music Ltd. Rights for Japan controlled by JVC Entertainment Networks, Inc. EMI VIRGIN MUSIC LTD. Permission granted by Virgin Music Japan Ltd. Authorized for sale only in Japan Used by Permission of JASRAC Licence No

424 WORDS IN ASIAN CULTURAL CONTEXTS Cultural information in a bilingual dictionary Jan-Olof Svantesson Department of Linguistics, Lund University Jan-Olof.Svantesson@ling.lu.se Abstract This paper discusses how to handle cultural information in a biligual dictionary from our experience with a Kammu English dictionary being prepared at Lund University. Background A group at Lund University (Kristina Lindell, Håkan Lundström, Jan-Olof Svantesson and Damrong Tayanin) has been working for a long time on a bilingual Kammu-English dictionary. Kammu (also known as Khmu, Kmhmu, etc.) is an unwritten Mon-Khmer language spoken by some 500,000 persons, most of them in northern Laos, and some in adjacent areas of Vietnam, China and Thailand. Traditionally the Kammu live on mountain slopes where they grow glutinous rice on dry swidden fields, and also partially depend on hunting and fishing for their living (see Tayanin 1994, Tayanin and Lindell 1991). Since the 1970s, many Kammu have been forced by economical necessity to leave their mountain villages and move closer to roads and rivers, where a traditional lifestyle cannot be upheld. All school education is given in Lao, the majority language of Laos, which is genetically unrelated to Kammu, and most Kammu in Laos are at least bilingual, knowing Kammu and Lao. Although the language is not immediately endangered, it is now changing under the influence of Lao. Under these circumstances, traditional Kammu culture is rapidly disappearing, and during the work on the dictionary we gradually realized that the knowledge that our main language consultant, Damrong Tayanin (whose Kammu name is Kàm Ràw), has about traditional culture is no longer shared by the younger Kammu. Furthermore, most Kammu of his generation (he was born around 1938) are no longer with us, and those who have survived are not in a position to transfer their knowledge to the younger generation or to an international audience, most of them being illiterate or having adopted a Lao lifestyle. Kàm Ràw was taught to be a shaman in his youth and thus has unusual knowledge of Kammu culture, including the spirit world and traditional medicine, in addition to his more general knowledge of Kammu society, agriculture and hunting, shared by all Kammu of his generation. Work on a Kammu dictionary When we started to work on this dictionary, the goal was to include mainly linguistic information of the usual kind. We soon decided not to include modern political and scientific terminology, which is invariably borrowed directly from Lao. Almost all Kammu in Laos can speak Lao, and those who can read and write do so in Lao. As a result, many speakers now include Lao words in their Kammu speech more or less freely, so, at least in theory, any Lao word can be included in the Kammu lexicon. Thus it is a bit arbitrary to decide which Lao loans should be included in a dictionary of this kind and which should not. Some Lao loans are very old and form a natural part 411

425 ASIALEX 2005 of the traditional Kammu lexicon. Partially they can be recognized formally, having preserved phonological features from older stages of Lao (or rather of Proto-Southwestern Tai, Li 1977), and being integrated in the Kammu derivational morphology. Having realized that Kàm Ràw s unique knowledge of his culture had to be preserved in some way, we also realized that our resources were sufficient for making deeper studies of only a fraction of it, including especially music, which was investigated by Håkan Lundström (see Lundström, in press). For this reason we decided to include as much cultural information as possible in the dictionary, which seems to be the best we can do given the available resources. Since the format is an ordinary alphabetically ordered dictionary, we have included cultural information as comments to the individual words. These comments are usually written by Kàm Ràw and have not been analyzed further, but should be taken as his knowledge of the role of the respective phenomenon in Kammu culture and society. Sometimes his descriptions were given as responses to questions of other members of the team, but most of the information was provided for those words for which he himself thought it was suitable. We also included condensed versions of information given by him in connection with previously published work from the Kammu language and folklore project. Cultural information Several kinds of information are contained in these notes, and some main types are outlined below: Plants. About 500 plant names are registered in the dictionary, and for each of them Kàm Ràw has written a short description of the use of this plant in Kammu culture. Some 300 of these plant species have been scientifically identified by botanists from Lund University doing fieldwork with Kàm in Laos and Thailand. The descriptions include information on how the plants are cultivated or how wild plants are gathered and used as food, as medicine, as building material, etc. Animals. Similarly, short descriptions have been added to all dictionary entries registering animal names. The information includes how the animals are hunted and eaten and their place in different religious ceremonies. Artefacts. Different kinds of man-made implements and utensils are described, how they are made and used. They include agricultural implements, cooking utensils, traps and musical instruments. Also buildings of different kinds (family houses, barns, etc.) are described in detail, including how they are built Kinship and marriage system. The Kammu kinship and marriage system is fundamentally important for the functioning of Kammu society. Everybody in a Kammu village has a kinship relationship to everybody else, and these relations are upheld by ceremonies and rules. In particular the rules for proper marriage relations are regarded as the basis for Kammu society. Spirits. Different spirits take part in Kammu life. They include the ancestor spirits who look after their living descendants and malignant spirits who punish people with illness or death, for example if they break the rules for proper marriage. Illness is regarded as caused by spirits, and it is the job of the village shaman to locate and 412

426 WORDS IN ASIAN CULTURAL CONTEXTS identify the spirits who caused the illness and to drive them out. Ceremonies and rituals. Ceremonies and rituals are performed in all phases of life, including when a child is born, when people get married and when they die. There are also ceremonies connected with the agricultural year, for instance when the rice is sown or harvested and other ceremonies are performed when someone falls ill. The dictionary also contains information on the meaning of personal names, and about different traditions and habits. There are some 500 drawings illustrating the words, most of them drawn by Kàm Ràw. Some of these are necessary for example to show the different parts of traps or other complicated artifacts. Altogether some 1,200 notes of this kind are included in the dictionary, and one problem is how to retrieve the information contained in them, so that they are not buried in this bilingual dictionary, ordered alphabetically after the Kammu forms. We plan to use a kind of analytical index ordered by different fields (such as plants, animals, etc., exemplified above) in addition to an alphabetically ordered English index so that a more systematic retrieval of this information can be done. Conclusion In conclusion, we decided to include cultural information provided by a native speaker as notes to words in a linguistic dictionary as a way of preserving the native knowledge of this unusually knowledgeable native speaker of the language. Although this creates some problems of retrievability and transparency of the material, it was taken as a way to preserve this knowledge which we had not the resources to investigate in a more systematic way. Hopefully it can be used as primary material for future research. References Li, Fang-kuei (1977). A handbook of comparative Tai (Honolulu: University Press of Hawaii). Lundström, Håkan (in press). I will send my song: Kammu vocal genres in the singing of Kam Raw (Copenhagen: NIAS Press). Lundström, Håkan (in press). Kammu songs: the songs of Kam Raw (Copenhagen: NIAS Press). Svantesson, Jan-Olof (1983), Kammu phonology and morphology (Lund: Gleerup). Svantesson, Jan-Olof, Damrong Tayanin, Kristina Lindell, Thongpheth Kingsada and Somseng Xayavong (1994), Watcanaanukom khamu-laaw [Kammu-Lao dictionary] (Vientiane: Institute for Research on Lao Culture, Ministry of Information and Culture). Tayanin, Damrong (1994), Being Kammu: my village, my life (Ithaca: Cornell University). Tayanin, Damrong and Kristina Lindell (1991), Hunting and fishing in a Kammu village (London: Curzon Press). 413

427