Type of task, time-on-task and electronic dictionaries in incidental vocabulary acquisition

Type of task, time-on-task and electronic dictionaries in incidental vocabulary acquisition MONICA HILL AND BATIA LAUFER Abstract Acomputer-based study compared the effect of three task types on incidental L2 vocabulary learning, on task-induced amount of dictionary activity and timeon-task. Ninety-six Chinese ESL students read a text containing twelve unfamiliar target items and performed one of the three tasks: a form-oriented production task, a form-oriented comprehension task or a meaning-oriented task. The text was presented on computer and the words could be looked up in electronic glosses. The results of immediate and delayed vocabulary tests showed that the two form-oriented tasks yielded better results than the meaning-oriented task. There was no significant difference in time-on-task, but there was a significant difference in the amount of dictionary activity the tasks generated. We concluded, therefore, that an important factor determining task effectiveness for vocabulary learning is the amount of word-related activity that the task induces. 1. Background 1.1. Vocabulary learning during reading One of the contentious issues in second language vocabulary learning is the extent to which the exposure to written language input is an effective and efficient means of learning vocabulary. Researchers of reading and L1 vocabulary have shown that with an estimated vocabulary of 5,000 words on entry to primary school, native speakers quickly increase their word power by around 1,000 words per year (Nagy 1997; Saragi et al. 1978). Since learners are deluged with written texts, especially from high school onwards, increase in vocabulary has been attributed to extensive reading. By entry to university, Goulden et al. (1990) suggest that the vocabulary size of educated adults is around 20,000 word families. Indeed such gains in vocabulary could not be accounted for by direct vocabulary acquisition (Nagy et al. 1985; Nation and Coady 1988). IRAL 41 (2003), 87 106 0019042X/2003/041-087 c Walter de Gruyter

88 Monica Hill and Batia Laufer In a second language, however, research on vocabulary growth and ultimate attainment presents a totally different picture. In many countries, the average vocabulary size of L2 learners, graduates of high school and university students of English as a Foreign Language, ranges between 1,500 and 4,000 word families (Laufer 2000). Can we, in all certainty, claim that this vocabulary has been acquired mainly through reading? Studies on vocabulary acquisition from reading (without any enhancement tasks) show that pick up rates of unfamiliar words range from 1 5 words in a text of over 1,000 words (Zahar et al. 2001; Luppescu and Day 1993; Hulstijn 1992; Knight 1994; Paribakht and Wesche 1993). Similar gains occur during reading books. In Horst et al. s (1998) experiment, an average of five words were gained from the reading of a simplified version of The Mayor of Casterbridge, atextof 21,000 words. Lahav (1996) conducted a study with students who read four simplified readers, each one of about 20,000 words, and found an average learning rate of 3 4 words per book. At this rate of growth, a second language learner would have to read in excess of eight million words of texts, 1 or about 420 novels to increase their vocabulary by 2,000 words. This would appear to be a daunting and time consuming means of vocabulary development. It is therefore reasonable that L2 learners acquire their vocabulary not only from input, be it reading or listening, but also through word-focused activities. 1.2. Vocabulary learning during reading supplemented with word-focused activities The simplest form of giving prominence to new words is by using graphic devices such as different font or color. Highlighted hyperlinks were shown to incite readers to click on these words in order to search their meanings. However, this increased clicking did not increase the vocabulary learned incidentally (De Ridder 2002). One of the most common word-focused activities used during reading is consulting a gloss or a dictionary. Glosses and dictionaries are used by learners to facilitate comprehension, but their use can contribute to small increments in vocabulary learning as well (Chun and Plass 1996, 1997; Lyman-Hager and Davis 1996; Lyman-Hager et al. 1993; Watanabe 1997). For example, Luppescu and Day (1993) found that students using a dictionary during reading gained higher scores on a post test than those who did not, and Knight (1994) found that learners of Spanish using an electronic dictionary recalled significantly more words in both immediate and delayed recall and in a comprehension activity than learners who had no access to a dictionary. If we compare the two, glosses and dictionary use, the latter is usually more effective (Hulstijn et al. 1996; Laufer 2000).

Incidental vocabulary acquisition 89 In addition to glosses and dictionaries (paper and electronic), a variety of reading-related vocabulary tasks has been investigated in connection with vocabulary learning. Thus, Hulstijn (1992) added a multiple choice exercise requiring the student to select the appropriate meaning of the new word encountered in a text. This task proved more efficient for retention of the practiced word than a traditional gloss for the same word or no gloss at all. Paribakht and Wesche (1997) compared learners in the reading only condition with learners in the reading plus condition. In the latter, the task consisted of reading a text and doing a range of vocabulary exercises based on the text. The reading plus group acquired significantly more words than the reading only group. Hulstijn and Laufer (2001) produced a gapped text, deleting ten words that were unfamiliar to the subjects. The explanation of these words was provided on a separate sheet. The subjects had to read the text and to insert the new words into it. When tested unexpectedly on the recall of these words, the learners attained higher scores than another group that read a non-gapped version of the text and had marginal glosses of the new words. Laufer (2001) compared three groups: a group which read a text and looked up new words in a dictionary, a group which read the same text and wrote sentences with the target words, and a group that was given a list of the target words and was asked to look them up and write sentences with them (without reading the text). When tested on the retention of the target words, the reading group was significantly worse than the other two. Interestingly, the reading + sentence writing group was the best on the immediate test, but it was not different from the sentence only group on the delayed recall test, suggesting that the additional value of reading a text did not make much difference to retention when sentences had been written with the new words. The implication of this result is that a word-focused task, such as writing a sentence, is more beneficial than a receptive awareness of these words induced by reading. 1.3. Task effectiveness and dictionary activity All the studies cited in Section 1.2 show that words are retained better if, in addition to reading, a word-focused task is performed. The reason for this is the attention to the word which is required of the learner during a word-based activity. Psychologists and L2 vocabulary researchers believe that the more individuals pay careful attention to various aspects of words, and the more they form rich and numerous associations with existing knowledge (e.g., in the form of establishing similarities and contrasts between old and new information), the higher the chances are that the new information will be retained (Anderson 1995; Baddeley 1997). In the case of vocabulary, this means that if learners pay careful attention to the word s pronunciation, orthography, grammatical

90 Monica Hill and Batia Laufer category, meaning and semantic relations to other words, they are more likely to retain the word than if they pay attention to only one or two of the above word properties. Such attention may not be necessary during a reading only task whose aim is global understanding of the text, but it is essential when the focus of a task is vocabulary. It is not difficult to conclude that reading together with vocabulary tasks is better for word learning than reading without vocabulary tasks. The question is whether some tasks are better than other tasks when they accompany reading. Laufer and Hulstijn (2001) point out that task superiority is justified by researchers in terms which defy a simple definition and operationalization: deep processing, elaboration, cognitive effort. They propose a motivationalcognitive construct of involvement which can be operationalized into specific tasks. The involvement load of a task depends on the presence or absence of three components: need (whether the word is needed for the task), search (whether the meaning is given or searched by the learner), and evaluation (which entails a comparison of a given word with other words, or combining the word with other words in order to assess whether a word does or does not fit its context). If we want to predict task effectiveness, the higher the involvement load, the more effective the task. We would like to follow a similar line of thought and claim that the more operations learners perform on a word, the better their retention will be. We suggest that, during reading, such operations may take the form of dictionary activity. Some evidence that the amount of dictionary activity is related to retention is found in Laufer and Hill (2000). They found that learners who used L1 translations for new words together with L2 definitions during a reading task, achieved the best scores on an unexpected vocabulary retention test. We define dictionary activity as the number of all selections at all stages of reading. Forexample, a learner looks up a new word in her native language, then decides to click on its definition in English. Two minutes later, she decides to return to this word and clicks on its L1 translation again. The dictionary activity index for this word, as expressed in the number of clicks, is three. Another learner looks up the word in L1, then clicks on the listen button to hear its sound, then looks up its L2 definition and never returns to the word. The student s dictionary activity index for the word is three as well. Dictionary activity can be observed and measured if specially designed computer programs are used. (We will describe such a program in Section 2.3.4.) If the amount of dictionary activity is related to learning, then the more activity a task induces, the more effective it should be for word learning. Whether this is so is an empirical question which we will address in our study in Section 2. Suffice it to say, at this stage, that the tasks we will describe in Section 2.3.3 were intended to provoke a different amount of dictionary use.

Incidental vocabulary acquisition 91 1.4. Time-on-task or type of task? An important question related to task effectiveness is whether it is the nature of the task that induces effective learning, or the time learners spend on it. A more demanding task is usually more time consuming. For example, in Hulstijn and Laufer (2001), writing a composition using the target words took longer than filling in the target words in sentences. The composition tasks yielded better retention scores, but the question is whether the better results were due to the type of task, or time-on-task. The researchers claimed that the fact that the composition took longer than the fill-in task does not necessarily mean that more time was devoted to the target words. They distinguish between time-on-task and time-on-target items. For example, a learner may have spent fifty minutes on the task of writing a composition, but only fifteen minutes on checking the use of the target words and incorporating them in the composition. Hence, time-on-task and time-on-target items are different. Similarly, in Ellis (1995), interactionally modified input, i.e., input that included clarifications, resulted in more words that were learnt than pre-modified input (input prepared on the basis of prior student interaction before the experimental task). In the latter case, however, more words were acquired per minute. The basic question of what leads to better learning type of task, or time-ontask has not been answered yet. 2. The study 2.1. Aim The aim of the study was twofold. Our first aim was to explore the different effects that message-oriented and form-oriented tasks might have on the retention of unfamiliar words which were encountered during reading. Our second aim was to attempt to characterize each task in terms of the dictionary activity it provoked, as well as the time required for its completion. Following the discussion in section 1, we assumed that retention of new target words would depend on what was done with them, i.e., the type of dictionary activity. However, we did not dismiss the possibility that different retention results could be contingent upon the different amount of time spent on each task. Our objectives were therefore to find out whether task type affected incidental vocabulary acquisition and, if it did, to explain why such effect occurred. 2.2. Subjects Initially, 128 subjects took part in the experiment. All the subjects were enrolled on English enhancement courses in the English Centre at the University of Hong Kong. They were all young adults about twenty years old, most of whom had been educated locally. All were second language learners of English

92 Monica Hill and Batia Laufer whose mother tongue is Cantonese or Mandarin. While the medium of instruction at their secondary schools is officially English, few students speak English outside of the classroom and so their use of L2 is slight and their vocabulary is relatively low. The minimum level of English for entry to the University of Hong Kong is grade D in the Use of English A level exam, roughly equivalent to a TOEFL score of about 550. The subjects were all of a similar proficiency, between grades C and D. Of the original cohort, the data from thirty-two subjects were excluded for a variety of reasons, including knowledge of more than one of the target words and being absent for the post-test. The final sample size was ninety-six. All the subjects were familiar with the use of a computer and a mouse as IT proficiency is considered part of their academic study. Subjects were randomly assigned one of the investigated tasks. 2.3. Materials 2.3.1. Reading passage. In looking for the reading passage, it was considered necessary to ensure that the text was of an appropriate level of difficulty with a lexical density that would allow general comprehension but encourage the subjects to exploit the computer dictionary resources provided to help them fully understand the text. Finding a suitable text proved extremely difficult and so a short passage of 165 words was written for the purpose of the experiment. Since the text was presented on a computer screen, its text length was restricted by the size of the computer screen and font size so that it would fit the screen without requiring the reader to scroll up and down. As the purpose was to stimulate vocabulary retention following the use of computer dictionary resources, careful consideration was given to the length and density of unknown vocabulary in the text. Laufer (1992) suggests that a minimum of 95 % of the words in a text need to be known for any reasonable chance of guessing the meaning of unknown words from context to occur. Hu and Nation (2000) suggest that the learner may reach an adequate comprehension level of a text at a density of 98 % familiar vocabulary, but that in intensive reading of short passages, less than 95 % coverage may be suitable for developing language and the use of reading strategies. The text used in the study (see Appendix), therefore included 93 % of words which we assumed (and later verified) were familiar to the learners. This density was regarded as being an appropriate level to enable the learners to be sufficiently challenged so as to want to use the computer dictionary resources. 2.3.2. Target words. The text contained twelve words, of which four were nouns, four adjectives and four verbs, that were unlikely to be familiar to the subjects. This was verified in a pilot test in which twenty students of a similar

Incidental vocabulary acquisition 93 English proficiency who did not participate in the experiment were asked to underline any words in the passage whose meaning they did not know. The twelve target words were the only words that were underlined. They were: indigenous (adj), arduous (adj), affability (n), itinerary (n), saunter (v), boisterous (adj), squander (v), weave (v), stunning (adj), remuneration (n), dusk (n) and toil (v). They were highlighted in the text on the screen. 2.3.3. Tasks. We investigatedthree tasks. Each task required the subjects to attend to each of the twelve target words in a different manner. The tasks themselves were presented on paper. The information necessary for the completion of the task was on the computer screen. In Task 1, the learner was required to answer twelve yes/no comprehension questions, each question on a different portion of the text containing one target word. Example (1) is an example of a Task 1 question and the sentence in which the answer is found. (1) Have the villagers of Sofa moved there from nearby towns? (yes/no) The Nanu region is famous for its wonderful scenery, ancient temples and, above all, the indigenous people of Sofa. The target word indigenous had to be understood in order to answer the question correctly. But the task does not require the learner to do anything with the word. It draws the learner s attention to the entire sentence. Hence, we refer to this task as message-oriented. In Task 2, the target word appeared on the screen and the learners had to select its meaning from four options. The options were high frequency words likely to be familiar to the learners. Example (2) illustrates this task: (2) Underline the correct answer: indigenous means in line 2 a) quiet b) dependent on c) originating in d) shy In this task, attention is drawn to the word itself, rather than to the context surrounding it. The word is provided by the researchers and the learner has to demonstrate comprehension of it. Hence, we refer to this task as a formoriented comprehension task. In Task 3, a synonym or paraphrase of the practiced word was provided and the learners had to select its corresponding word form from four options which consisted of the correct target word and three distractors selected from the twelve target words. Example (3) illustrates the task: (3) Which word means friendly nature? a) indigenous b) toil c) affability d) remuneration

94 Monica Hill and Batia Laufer Figure 1. Selection of affability and its English meaning In Task 3, attention is drawn to the word itself. But, unlike in Task 2, the search performed by the learner is not for the meaning of a given word, but for the word itself. Hence, we refer to this task as a form-oriented production task. It may be claimed that since nothing is produced (said or written) by the learner, the term production may be inadequate. Yet productive knowledge is defined by Nation (2001: 359) as that used in speaking and writing, and (which) involves going from the meaning to the word form. Nation argues that productive knowledge can be checked by a recall task where the target word has to be recalled, as in translating a word from L1 into L2, or by a recognition task where the target word has to be recognized and selected from several options. Our Task 3 is, therefore, a productive recognition task. 2.3.4. Computer program. The study was made possible by a computer program designed specially for it. The text was reproduced in a program created in Macromedia s Authorware. Each subject is logged in by name. At the side of the text are four buttons, each of which provides a source of information about the target words. The line along the bottom of the screen is the timer. If the subject selects affability and checks its meaning in English, the information in Figure 1 will appear.

Incidental vocabulary acquisition 95 Figure 2. Selection of affability and its Chinese meaning The translation is given when Chinese Meaning is selected (Figure 2). Extra dictionary information includes, as shown in Figure 3, the part of speech, the preposition that follows a verb and an example of the use of the word in a new context. Hear Word can be clicked on to hear the pronunciation of the word. When the subject has completed the task, he or she clicks on Done and the timer notes the time taken to complete the task. The last, and perhaps the most important, data appear in the results screen shown in Table 1. Each student s look-up behavior is recorded by the computer. In the example in Table 1, the subject has selected mainly the English meanings, but has also listened to the pronunciation of four items and checked extra information on squander. The column showing times selected showsthe numberof times the subjectselected information on the word. Chappelle (2001) notes that this use of IT in research can provide valuable data on language learning. 2.4. Research questions The specific research questions were as follows: 1. Do learners acquire different numbers of words following the completion of different tasks:

96 Monica Hill and Batia Laufer Figure 3. Extra information on squander a. as reflected in the results of an immediate test, and b. as reflected in the results of a delayed test? 2. Is there a difference between the tasks in the time learners spend on task completion? 3. Is there a difference between the tasks in the dictionary activity they provoke: a. as reflected in the number of clicks on the target words, and b. as reflected in the type of dictionary information? 2.5. Procedure The subjects were asked to come for individual tutorials of about fifteen minutes. The study was conducted in five stages. 1. Pre-test: A pre-test was administered in which the twelve words were listed in alphabetical order and subjects were asked to give the meanings of any words they knew in English or Chinese. The sheets were collected for subsequent analysis. 2. Reading for general meaning: Each subject was then asked to read the text (on paper) and try to understand the general meaning. All the target words

Table 1. Look-ups recorded in the computer log Incidental vocabulary acquisition 97 Student Name: Chan, Mary Faculty: Arts Word English Hear Word Chinese Extra Info Times Meaning Meaning selected indigenous 1 1 0 0 2 arduous 0 1 0 0 1 affability 1 0 1 0 2 itinerary 0 1 1 0 2 sauntered 1 0 0 0 1 boisterous 1 1 0 0 1 squander 1 0 0 1 1 weave 1 0 0 0 1 stunning 1 0 1 0 2 remuneration 1 0 1 0 2 dusk 1 0 0 0 1 toil 1 0 0 0 1 appeared in text context and were in bold font. This stage was introduced into the study in order to avoid too many look-ups for general comprehension, during the task, as opposed to look-ups which were provoked by the task. 3. Task performance: After a couple of minutes the subject was invited to sit at the computer where the same text was shown on the screen, together with information that would help the student complete their task. Subjects were randomly assigned one of the three tasks and given the following instructions: Information is available on the screen to help you complete the task. If you would like to know more about any of the words, just select the word with the mouse, then click on whichever information you want: English meaning, Chinese translation, pronunciation or extra information that you would find in a dictionary. You have ten minutes to complete the task. 4. Immediate post test: Once the task was completed, the sheet was collected and the subject was asked unexpectedly to give the meaning of each word either in English or Chinese. Hence incidental acquisition of vocabulary is operationalized in the study as the ability to recall the word s meaning in L1 or L2. 5. Delayed post-test: The test above was repeated a week later in class without any warning.

98 Monica Hill and Batia Laufer 2.6. Data analysis The sources of our data were the pre-test, the two post-tests and the students personal log files. Each word, whether translated, explained, or left unanswered on the pre- and post-tests, was scored dichotomously. A score of 0 was given for an incorrect or not attempted answer and a score of 1 was given for a correct answer. Spelling and minor grammatical errors were disregarded as the primary focus was on the ability to recall the meaning of the word. For example for affability, the answer friendly was accepted. However, for weave, the response making clothes was not accepted as it was not sufficiently close to the explanation provided in the program: to make threads into material, to twist or bind. The function of the pre-test was twofold: to exclude students who knew more than one target word and to adjust some scores for pre-knowledge of one word. The test scores of subjects who knew one word (most frequently remuneration in true Hong Kong style) were adjusted accordingly by converting their score out of 11 (instead of 12) to a percentage grade 2. The immediate and the delayed post-tests provided data on incidental learning of the target words. The maximum raw score on each of the tests could be twelve, if all twelve target words had been retained. The individual log files provided us with the following information: time spent on the task and dictionary activity. Dictionary activity included all the dictionary selections for each word. All the data were arranged in two databases. One database contained the students ID, task type code, pre-test score, immediate post-test score, delayed post-test score. The other database was arranged by student and showed information about every single word looked up: dictionary activity that the word induced and whether the word was retained or not on the two post tests. The database also included the student s ID and task code he or she was assigned to. Example (4) shows a database for the target word affability. The other eleven target words are listed with their various look-ups in a similar way. (4) Subject Task IR DR Affability Affability Affability Affability Time ID type score score English Hear Chinese Extra on Meaning Word Meaning Info task The data collected allowed us to compare tasks on students word recall, on the amount and type of dictionary activity, and on time-on-task.

Incidental vocabulary acquisition 99 3. Results 3.1. The first research question Our first research question addressed the effect of each task on word learning, as reflected in the results of an immediate and delayed test. Tables 2a and 2b present immediate and delayed recall scores in percentage. The difference between the tasks was measured by one way ANOVA. Tables 2a and 2b show that Task 3 yielded the highest sores in both the immediate and delayed recall tests, followed by Task 2, while Task 1 yielded the lowest scores. The differences between the three tasks were significant in both tests. A post hoc analysis (Tukey Kramer Multiple comparisons test) showed that, in immediate recall, Task 1 was significantly less effective than Task 2 (p < 0.01)and Task 3 (p < 0.001). There was no significant difference between the means of Tasks 2 and 3. In delayed recall, Task 1 yielded significantly lower scores than Task 3 (p < 0.05). There was no significant difference between the other tasks. 3.2. The second research question The second research question focused on whether there was a difference between the time taken by the learners to complete the tasks. The log noted how long each subject spent on completion of the task and the mean was 5.5 minutes out of a maximum of ten minutes permitted. Table 3 presents time-on-task results in minutes. Table 2a. Task effect on immediate recall Task Test score Mean (%) Test score S.D. Task 1 (n=32) 42 22.7 Task 2 (n=33) 62.2 24.3 Task 3 (n=31) 71.9 22.3 Difference F (2,93) = 13.4 p < 0.001 Table 2b. Task effect on delayed recall Task Test score Mean (%) Test score S.D. Task 1 (n= 32) 29.4 20.4 Task 2 (n=33) 37.9 21.1 Task 3 (n=31) 45.3 24.3 Difference F (2,93) = 4.12 p < 0.05

100 Monica Hill and Batia Laufer Table 3. Time spent oneachtask Task Minutes Mean S.D. Task 1 (n=32) 5.23 1.71 Task 2 (n=33) 5.62 1.89 Task 3 (n=31) 5.62 1.92 Difference F (2,93) = 0.47, not significant Table 4. Dictionary activity by task Dictionary information Task 1 Task 2 Task 3 n=32 n=33 n=31 English Meaning 7.1 11.9 6.9 Hear Word 2 6.6 5.2 Chinese Meaning 5.6 3.6 10.3 Extra Info 0.9 0.4 0.6 Mean total clicks 15.5 22.5 23 One way ANOVA comparing mean times (in minutes) spent on the tasks showed that there was no significant difference between the time required for completing the tasks. Hence, the difference in the nature of the task did not result in difference in time-on-task 3. 3.3. The third research question The third research question examined the amount of dictionary activity provoked byeach task, as reflected in the number of clicks on the target words, and in the type of dictionary information sought. Table 4 presents dictionary look-up activity by task. Each figure represents the mean number of clicks per student on particular dictionary information, over all twelve words. The bottom row sums up the mean number of clicks per student. All the figures are higher than twelve, which is the number of target words. This means that, for some words, the learners looked up more than one item of dictionary information. Interestingly, out of the thirty-two learners who knew one word on the pre-test, eleven learners looked up the known word anyway during the task. Table 4 shows that Task 1 generated a lower number of clicks than Tasks 2 and 3. We compared the three tasks on dictionary activity by ANOVA and found that the difference between them was significant (F [2,93] = 3.45 p < 0.05). A post hoc analysis (Dunn s multiple comparisons test) showed that

Incidental vocabulary acquisition 101 there were significant differences between Task 1 and Task 2 (p <.05) and between Task 1 and Task 3 (p <.05). There was no significant difference between Tasks 2 and 3. 4 With regard to the type of information selected, we can see that in Task 3, the majority of look-ups were translation of the target word. In the two other tasks, students preferred to look up the English explanation of the words. 4. Discussion In this article we examined incidental vocabulary acquisition from reading a text and performing a task which required some focus on unfamiliar words in the text. (Acquisition was defined as recall of the word s meaning.) In each task, learners had to look up unfamiliar target words in an electronic dictionary. Our results show that all three tasks led to some vocabulary learning. This finding is not new. It is similar to the findings of other studies, which showed that learners who read a text and perform a task which requires some operations on the new words, including looking up the words in a dictionary, will consequently remember some of these words. To use Laufer and Hulstijn s (2001) involvement model as an explanation of this finding, all three tasks were effective since they had the component of search in them, one of the three components of involvement which are presumed to aid learning. However, the aim of the paper was to go beyond mere demonstration of acquisition. The question we raised was whether message-oriented and formoriented tasks differed in their effectiveness on vocabulary acquisition. To this effect, we compared three tasks: a message-oriented task (Task 1), a formoriented comprehension task (Task 2), and a form-oriented production task (Task 3). The two form-oriented tasks resulted in a higher number of acquired words. Task 1 (the meaning-oriented task) proved to be the least effective. It yielded significantly lower vocabulary scores than Tasks 2 and 3 on the immediate recall test. On the delayed recall test, even though Task 1 was significantly different from Task 3 but not Task 2, the higher scores on Task 2 suggest that with larger samples the difference between the two might have reached significance. The two form-oriented tasks were more effective, but why weren t they different from each other? A possible reason for the lack of significant difference between Task 2 and 3 may lie in the nature of our post-tests. In these tests we required the learner to recall the meaning of the target word. Similarly, Task 2consisted of finding the meaning for each target word. Task 3, on the other hand, was more demanding in that it required selecting the target word form for a given meaning. A more demanding test which would ask the learner to recall the target word for a given translation or paraphrase, might have produced larger differences between the two tasks, showing Task 3 to be even better than

102 Monica Hill and Batia Laufer in our results. Even though we did not demonstrate this claim in our study, we showed the superiority of form-oriented tasks. How can we explain the difference in task effectiveness? Our initial assumption was that these differences could be accounted for by the different amounts of dictionary activity that each task induced. The amount of dictionary activity was defined as the number of look-ups, or clicks, which a student performed during the task and which were registered in the individual log files. Comparison of the three tasks on the number of look-ups showed that there were indeed differences in dictionary activity. Task 1, which yielded the lowest acquisition scores, induced the lowest dictionary activity. Tasks 2 and 3, the form-oriented tasks which proved more effective than Task 1, induced significantly more dictionary activity than Task 1, the message-oriented task. This supports Laufer and Hill s (2000) study which found that students who looked up English target words in English and in their L1 did better than those who looked them up in one language only. In other words, subjects who engaged in more dictionary activity did better. Furthermore, the increased dictionary activity included a more frequent use of the Hear button, thereby providing auditory reinforcement. It is possible that this additional dimension of activity contributed to the superiority of Tasks 2 and 3. 5 The weakness of Laufer and Hill (2000) was in the lack of control for the precise time-on-task. Even though there was a time limit of ten minutes, many students finished earlier. It could be argued that what led to better learning was not the type of information looked up, or the amount of dictionary activity, but simply the larger amount of time which the students with better vocabulary learning results spent on the target words. One aim of the present study was to check whether a different amount of time was required to complete the tasks under investigation. We found that there was no significant difference in the time spent on the three tasks. Therefore, the different acquisition scores which we found could not possibly be attributed to time-on-task. Hence, we conclude that the type of task, form-oriented rather than message-oriented, accounted for better acquisition results. The reason for this effectiveness probably lies in the increased dictionary activity that they induce. It may seem strange that learners with more dictionary look-ups completed the task in the same amount of time as learners with fewer look-ups. We cannot know for sure why this was so, but we can suggest a possible explanation. Task 1, the meaning-oriented task, consisted of comprehension questions and required the learner to focus not only on the target word, but on the sentence, or sentences in which the answer to the question could be found. Thus, part of the time-on-task was devoted to decoding the message of the sentences. The time spent on the target items was lower, as shown by the significantly lower dictionary activity. Tasks 2 and 3 required more time than Task 1 for the actual target words. Therefore the three tasks were similar in the overall time required

Incidental vocabulary acquisition 103 to complete each task, but the time in Task 1 was distributed differently than in Tasks 2 and 3. We do not contest the position that learning is related to the amount of time spent on what is being learnt. We contend, however, that the time which contributes towards learning is not time spent on the task,buttime spent on the target item. Why did the tasks we assigned induce different amounts and types of dictionary activity? The answers to questions in Task 1 required understanding of the target words, but no direct questions were asked about their meaning. Therefore learners may have remained satisfied with superficial understanding of these words, reflected in one click on translation or explanation in L2, or they may even have tried to guess some words without looking them up as long as they felt they provided a correct answer to the comprehension questions. In Task 2, the learners were specifically required to provide the correct meaning of the target word selecting it from four options which consisted of high frequency familiar words. It was only sensible to check the electronic dictionary for the meaning of the target word. Since the options were given in English, most look-ups revealed a search for the English explanation of the word. In Task 3, the learners had to match a familiar word with a target word which appeared together with three other target words. Since the four options were target words, i.e., new unfamiliar words, learners had to look up each option. Differentiating between four new words is more demanding than selecting the correct meaning for one new word (Task 2). This demanding nature of the task may account for the large number of L1 look-ups. Learners had to be certain about the meaning of each new word and therefore resorted to the mother tongue equivalent. The higher number of look-ups, auditory reinforcement and search for the precise meaning is probably what made Tasks 2 and 3 more effective in learning the new words than Task 1. 5. Conclusion The two objectives of the study were to find out whether task type affected incidental vocabulary acquisition and, if it did, to explain why such effect occurred. Specifically, we compared three tasks: a message-oriented task, a formoriented comprehension task and a form-oriented production task. The comparisons were of the number of words learnt after the completion of each task, of the time that the tasks took to complete, and of the dictionary activity the tasks induced. We found that the two form-oriented tasks were more effective for vocabulary learning than the meaning-oriented task. The production task yielded the best results. There was no significant difference in the time that was necessary to complete the tasks. Hence the different effectiveness of the tasks could not be attributed to time-on-task. The tasks were different in the amount of dictionary activity they provoked. The order in effectiveness(task 1 the worst,

104 Monica Hill and Batia Laufer Task 3 thebest) was the same as the order in the dictionary activity. Task 1 provoked the lowest activity and Task 3 the highest. We conclude, therefore, that an important factor determining task effectiveness for vocabulary learning is the amount of word-related activity that the task induces. University of Hong Kong <mmhill@hku.hk> University of Haifa <batialau@research.haifa.ac.il> Appendix Text The Nanu region is famous for its wonderful scenery, ancient temples and, above all, the indigenous people of Sofa. In their remote village, reached after an arduous and dusty drive, live the Sofans, with their welcoming smiles and warm affability.fewvisitors, however, make this trip as it is not on the normal itinerary of most tour groups. As I sauntered through the village, I stopped for a few moments on a bridge to watch the boisterous children splashing each other with water. School hours areshort hereand manyof theteenagersseemtosquander their daylight hours riding around the square on noisy old motorcycles. Groups of local women spend their days weaving coloured silk into stunning fabrics. Their remuneration is only a few dollars although they may have worked for several days to create each one. As dusk approaches, the men return from the fields where they have been toiling all day, hungry and ready for a quiet evening with their families. Notes 1. A novel of simplified English of 20,000 words is a rather short novel, but foreign language learners can hardly be expected to read long authentic novels, or nonfiction material before they have reached the vocabulary size of at least 5,000 word families, or 8,000 lexical items (Laufer 1992). 2. In Task 1, 8 out of 32 students knew one word. In Task 2, 8 out of 33 knew one word, and in Task 3, 16 out of 31 knew one word. 3. The fact that some students knew one of the twelve target words did not affect the results. We compared the mean time-on-task of all the students who knew one word and those who did not know any of the words. The mean difference was 0.13 and wasfound not to be significant: T (two tailed, df 94) = 0.33, p=0.74. Similarly, there was nosignificant difference between the standard deviations of the two groups (F =1.32, p=0.17). 4. The fact that some students knew one of the twelve target words did not affect the results. We compared the mean number of clicks of all the students who knew one

Incidental vocabulary acquisition 105 word and those who did not know any of the words. The mean difference was 3.06 and was found not to be significant: T (two tailed, df 94) = 1.19, p=0.24. Similarly, there was no significant difference between thestandard deviations of the two groups (F = 1.49, p=0.09). 5. Previous research on Chinese L2 learners has shown that hearing the pronunciation of an unfamiliar word on first presentation can help them to learn the word (Hill 1999). References Anderson, John R. (1995). Cognitive Psychology and its Implications (Fourth edition). New York: Freeman. Baddeley, Alan D. (1997). Human memory: Theory and Practice (Revised edition). Hove, UK: Psychology Press. Chappelle, Carol (2001). The attraction of technology. Paper presented at the American Association of Applied Linguistics conference, Applied Linguistics for the 21st Century, 24 February 2001, St Louis, Missouri, USA. Chun, Dorothy M. and Jan L. Plass (1996). Effects of multimedia annotations on vocabulary acquisition. The Modern Language Journal 80: 183 198. (1997). Research on text comprehension in multimedia environments. Language Learning and Technology 1: 60 81. De Ridder, Isabelle (2002). Visible or invisible links: Does the highlighting of hyperlinks affect incidental vocabulary learning, text comprehension, and the reading process? Language Learning and Technology 6: 123 146. Ellis, Rod (1995). Modified input and the acquisition of word meaning. Applied Linguistics 16: 409 441. Goulden, Robin, Paul Nation, and John Read (1990). How large can a receptive vocabulary be? Applied Linguistics 11: 341 363. Hill, Monica (1999). English vocabulary learning by Chinese students: How can phonology help? Paper presented at the AILA Congress, August 1999, Tokyo. Horst, Marlise, Tom Cobb, and Paul Meara (1998). Beyond a clockwork orange: Acquiring second language vocabulary through reading. Reading in a Foreign Language 11: 207 223. Hu, Marcella and Paul Nation (2000). Vocabulary density and reading comprehension. Reading in aforeign Language 13: 403 430. Hulstijn, Jan H. (1992). Retention of inferred and given word meanings: Experiments in incidental vocabulary learning. In Vocabulary and Applied Linguistics, Pierre J. L. Arnaud and Henri Béjoint (eds.), 113 125. London: Macmillan. Hulstijn, Jan H., Merel Hollander, and Tine 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: 327 339. Hulstijn, Jan H. and Batia Laufer (2001). Some empirical evidence for the Involvement Load Hypothesis in vocabulary acquisition. Language Learning 51: 539 558. Knight, Susan M. (1994). Dictionary use while reading: The effects on comprehension and vocabulary acquisition for students of different verbal abilities. Modern Language Journal 78: 285 299. Lahav, Yehudit (1996). Vocabulary acquisition through extensive reading: Incidental and controlled acquisition compared. Unpublished MA thesis, University of Haifa. Laufer, Batia (1992). How much lexis is necessary for reading comprehension? In Vocabulary and Applied Linguistics, Pierre J. L. Arnaud and Henri Béjoint (eds.), 126 132. London: Macmillan.

106 Monica Hill and Batia Laufer (2000). Electronic dictionaries and incidental vocabulary acquisition: Does technology make a difference? In EURALEX, Ulrich Heid, Stefan Evert, Egbert Lehmann, and Christian Rohrer (eds.), 849 854. Stuttgart: Stuttgart University. (2001). Reading, word-focused activities and incidental vocabulary acquisition in a second language. Prospect 16 (3): 44 54. Laufer, Batia and Monica 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): 58 76. Laufer, Batia and Jan H. Hulstijn (2001). Incidental vocabulary acquisition in a second language: The construct of task-induced involvement. Applied Linguistics 22: 1 26. Luppescu, Stuart and Richard R. Day (1993). Reading, dictionaries and vocabulary learning. Language Learning 43: 263 287. Lyman-Hager, Mary-Ann and James Davis (1996). The case for computer-mediated reading: Une vie de boy. The French Review 69: 775 790. Lyman-Hager, Mary-Ann, James Davis, Joanne Burnett, and R. Chennault (1993). Une vie de boy: Interactive reading in French. In Proceedings of the CALICO 1993 annual symposium on assessment, F. L. Borchardt and E. M. T. Johnson (eds.), 93 97. Durham, NC: Duke University. Nagy, William E. (1997). On the role of context in first and second language vocabulary learning. In Vocabulary: Description, Acquisition and Pedagogy, Norbert Schmitt and Michael McCarthy (eds.), 64 83. Cambridge: Cambridge University Press. Nagy, William E., Patricia A. Herman, and Richard C. Anderson (1985). Learning words from context. Reading Research Quarterly 20: 233 253. Nation, Paul (2001). Learning Vocabulary in Another Language. Cambridge: Cambridge University Press. Nation, Paul and James Coady (1988). Vocabulary and reading. In Vocabulary and Language Teaching, Ron Carter and Michael McCarthy (eds.), 97 110. London: Longman. Paribakht, T. Sima and Marjorie Wesche (1993). Reading comprehension and second language development in a comprehension-based ESL programme. TESL Canada Journal 11: 9 27. (1997). Vocabulary enhancement activities and reading for meaning in second language vocabulary acquisition. In Second Language Vocabulary Acquisition. A rationale for Pedagogy, James Coady and Thomas Huckin (eds.), 174 200. Cambridge: Cambridge University Press. Saragi, T., I. S. P. Nation, and G. F. Meister (1978). Vocabulary learning and reading. System 6: 72 78. Watanabe, Yuichi (1997). Input, intake and retention: Effects of increased processing on incidental learning of foreign language vocabulary. Studies in Second Language Acquisition 19: 287 307. Zahar, Rick, Tom Cobb, and Nina Spada (2001). Acquiring vocabulary through reading: Effects of frequency and contextual richness. Canadian Modern Language Review 57: 541 572.