The Journal of Asia TEFL

Transcription

1 THE JOURNAL OF ASIA TEFL Vol. 14, No. 3, Fall 2017, The Journal of Asia TEFL e-issn AsiaTEFL.org. All rights reserved. Hedging Strategies in Health Discourse: A Study of Pharmaceutical Leaflets Esther Serwaah Afreh Kwame Nkrumah University of Science and Technology, Ghana Osei Yaw Akoto Kwame Nkrumah University of Science and Technology, Ghana Kodwo Adam-Moses Kwame Nkrumah University of Science and Technology, Ghana Hedging, as a rhetorical tool, has received considerable attention from scholars from diverse fields of study such as advertisement, politics, and religion. Despite the extensive exploration of the topic, it has not received much attention in health communication. Thus, this study, which employs Quirk, Greeenbaum, Leech, and Svartvik s (1985) functional principle for establishing word class and Hyland s (1998) model as frameworks, investigated the various linguistic realizations and functions of hedges in pharmaceutical leaflets (PL), which are also called Patient Information Leaflets. The study, which is based on 50 PLs obtained from some leading pharmaceutical shops within the Cape Coast metropolis and other areas, adopted both qualitative and quantitative content analysis. It was found that pharmaceutical companies employed certain lexico-grammatical hedging devices to mitigate their claims. The study has implications for both patient-pharmacist communication and hedging theories. Keywords: patient pharmaceutical leaflets, hedges, health, rhetoric Introduction Pharmaceutical communication, a sub-branch of health communication, focuses mainly on how information on drugs/healthcare products (medicinal and non-medicinal) can be effectively communicated to drug users or patients in a more rhetorically appropriate, epistemologically acceptable, and ethicallytolerable manner. Like any form of communication, particularly medical or health discourses, English pharmaceutical discourse (Mẫu, 2012) has its conventionalised lexicon, register and norms that govern communication, be it direct-to-consumer or direct-to-prescriber communications (Gilbody, Wilson, & Watt, 2005). Drug manufacturers, therefore, adhere to these norms regarding both the discourse and the metadiscourse in advertising their drugs. At the metadiscoursal level, pharmaceutical leaflet (PL) writers present truths be it contextualised, evidential, and interpreted (Skelton, 1997) in a language that shows deference to readers (Hyland, 2006a, 2006b). Hence, hedging devices are commonplace linguistic phenomena in this genre. Hedges are therefore employed in general medical writings: 515

2 to present a proposition with greater accuracy, or to gain some protection from the possible criticism, or as a device that offers due recognition and deference to readers' views. hedges can be adopted to prevent or temper conflicts or negative reactions of a reader, to protect face of the reader by minimizing certain face-threatening acts in order to pave the way for a smooth, cooperative and successful course of communication. (Zhen, 2007, p. 18) Hedging is therefore perceived as part of the interpersonal communication strategies in pharmaceutical advertisement and communication. Hence, Gross, and Chesley (2012) added that hedges are a rhetorical component of the voice of science, a voice designed to persuade an audience It is a voice that persuades so well because it is generated by a set of social norms (p. 11). del Olmo (2006) also clearly stated that hedging could be described as a tri-dimensional concept which implies: a) vagueness and intentional fuzziness, b) an author s modesty in terms of own achievements and personal implication, and c) impossibility or unwillingness to reach an absolute precision or quantify all the observed phenomena. Similarly, Hyland (2006) has made an insightful observation concerning general medical writing: In a context where the accreditation of knowledge depends on the consensus of the research community and the need to evaluate evidence, to comment on its reliability, and to avoid potentially hostile responses, expressions such as might, perhaps, and possible can contribute to gaining the acceptance of research claims. Medical papers provide interesting and useful examples of the use of hedging in scientific discourse because they relate to matters impinging on significant issues of our lives. (p. 694) There have been a number of studies on PLs which include Vigneshwaran, Padmanabha, and Devanna (2013), Cardeñosa, Gallardo, and Toni (2008), and Paiva (2000). However, few of such have looked at PL from a linguistic perspective. Those on the discourse aspect of PLs have also not given serious attention to hedging strategies employed by the PL writers, even though hedging has been argued to be a commonplace phenomenon in medical discourses (Antic, 2009; Hyland, 2006b). And we have mentioned already that even though hedging has received attention from multidisciplinary scholars, PLs have not been explored in terms of hedging. Hence, the literature focuses on studies within the medical discourse/writing genre in general. As Antic (2009) opined, medical papers provide interesting and useful examples of the use of hedging in scientific discourse because they are related to matters which influence significant issues of our lives (p. 4). A few studies such as Zhen (2007), Hyland (2006b), and Skelton (1997) have explored tentativeness in medical writing in general. However, there has not been an extensive research on hedging strategies in PLs to ascertain the nature of the linguistic realizations of hedging. Thus, the present study seeks to explore hedging strategies employed in PLs to ascertain the linguistic realizations of hedging in this key genre of pharmaceutical literature. Specifically, the study seeks to explore the lexico-grammatical resources of hedging devices employed by PL writers to communicate with their intended audience. The Patient Information Leaflet There are a number of genres through which experts in pharmaceutical sciences communicate to and with their audience. Some of these are research articles, textbooks, and conference papers. However, one key genre is the PL, which is an invaluable component of direct-to-consumer communication. PLs play both informative and educative roles. This then helps demystify the general perception that a pharmacist s job is only to dispense drugs (Waterfield, 2010). Waterfield therefore remarked that the contribution of the pharmacist to patient care is beyond the mechanical and technical supply function associated with dispensing (p. 1). 516

3 Patients have rights (Casas & Parada, 2008; Zülfikar & Ulusoy, 2001). One of their foremost rights is the right to information. This right of patients therefore becomes a responsibility of healthcare providers, particularly doctors and pharmacists. Consequently, healthcare providers are obliged to respect this right of patients. Hence, doctors communicate to and with patients through multimedia avenues such as spoken, written, internet communication, radio, and television. One of the key written modes through which direct-to-consumer communication occurs is the patient information leaflet, which is the primary or only source of information about their medicine for many people. In the UK, PLs has been a statutory requirement since The PL is an ethico-legal requirement in the pharmaceutical industry (Medicines & Healthcare Products Regulatory Agency [MHRA], 2008). PLs contain authoritative information on drugs and they are one of the vital means of doctor-patient communication - part of the direct-to-consumer communication. It is very useful because it helps readers to make informed choices even in a doctor s absence, thereby helping patients to ensure self-care. The importance of this is that, unlike other unregulated sources of information such as online adverts and information centre adverts, the patient information leaflet is highly regulated because it has to undergo some approval processes. The Corpus for the Present Study Methodology The data comprises patient information leaflets (i.e., PLs) written in English. The present study realised that language is indispensable in pharmaceutical communication. Hence consumers are communicated to through a number of international languages. The leaflets gathered revealed that PLs can be grouped into three major categories: monolingual, bilingual and multilingual, depending on the number of languages adopted in communicating the pharmaceutical information to the potential users of the drugs. Out of the number gathered, the predominant language was English. It can be said that English, as a global language, is employed as a lingual franca in health communication. The rationale behind could be to reach a wider consumership. For the purpose of this study, only PLs written in English were used. The study was limited to 100 PLs, gathered from three main sources. First, we used what we had in our possession. Second, we collected some from friends. Third, we contacted some pharmacy stores in and around the university community and procured copies from them. The pharmacists and the researchers were aware of the ethico-legal role of the leaflets and its relevance to drug users. Three main steps were taken by the researchers in selecting the hedging devices. In the first step, the researchers read through the leaflets and tagged all the possible hedging devices. The next was that the researcher re-read the text to identify the true hedging devices as means of both intra- and inter-rater reliability tests. This step helped in authenticating the metadiscursivity (specifically hedgeness ) of such linguistic resources. Finally, all the identified hedging devices were manually copied and typed for easy analysis. Analytical Framework The study employed Quirk, Greeenbaum, Leech, and Svartvik s (1985) functional principle for establishing word class as well as Hyland s (1998) model as frameworks to respectively account for the lexical and nonlexical hedging devices used in the data. According to Quirk et al. (1985), lexical hedges, like all words in English, may easily be identified using any criteria for establishing word classes namely, morphological, notional, and functional principles. However, in this study, we used the functional principle because of the inherent challenges the other principles seem to have. The morphological principle, for instance, underscores that the word class of a word may be established by merely looking at the morphemes that constitute the word. For example, this principle indicates that words with derivative endings such as -ism, -ize, -al, and -ly, 517

4 as in the case of metabolism, ironize, fully, national, and fully may be thought to be nouns, verbs, adjectives, and adverbs respectively. This analysis where the word class of the word is determined outside context is believed to be flawed from the onset since the principle cannot account for words which are made up of only one morpheme such as argument, come, nice, and well. Secondly, even though the morphological principle is believed to be able to determine word classes of words with more than one morpheme, it cannot be reliable in all cases. The reason is that some words of different word classes have similar inflectional and derivative endings. It will therefore be misleading to solely rely on this principle to determine word classes of all such words. Like the morphological principle, the notional criterion has its own difficulties and cannot be relied upon in all instances. This criterion identifies the word class of words based on people s perception or notion of the words. It intimates that words are categorized in their respective word classes based on the name the word suggests as perceived by users of the language. As such, nouns may be said to be words that name everything whereas verbs refer to names of words that denote an action or a state of being. Adjectives, on the other hand, refer to names of words that qualify nouns while adverbs are words that modify verbs, adjectives, and other adverbs. The following sentence may be used as an illustration: The quick brown fox skillfully jumped over the lazy dog. Considering the sentence above, notionally, brown, fox, and dog can be said to be nouns since they are names of colour and animals respectively. In the same vein, whereas jump is a verb because it depicts the name of a word that denotes an action, skillfully is an adverb because it is the name of a word that modifies the verb jump. The word lazy can also be thought to be an adjective since it is the name of a word that modifies the noun dog. The seeming difficulty of this analysis stems from the fact that, if all words in English depict names of something, what then is its description or categorization independent of what it is thought to be. More importantly, each of these words above may belong to a different word class all together used in a different context. Thus, quick can be said to be a noun when used in a sentence like: The quick always wins. These fundamental setbacks with the morphological and notional principles undoubtedly make the functional principle the obvious choice for making any thorough analysis of the word class of words in English. The functional principle posits that the word class of words has to be determined based on the context in which they are used. The principle lends credence to the grammatical environment in which the word is used. Thus, round, as an example, can be said to belong to different word classes depending on context of its use. The word can be said to function as a noun, a verb, an adjective, and an adverb in the following sentences: 1. The round is my preference. 2. Muslims round the Kabba when they go to Mecca. 3. I prefer the round pie. 4. Kwame went round the bend. Arguing within the confines of syntactic frames as espoused in traditional grammar, round in 1. above can be said to be a noun because it is only words that belong to the noun category which can fill the slot which comes after the determiner the. In a similar analysis, round in 2. can be thought to be a verb as it functions as the verbal element in the sentence. The same word in 3. functions as an adjective by virtue of the fact that it modifies the noun bend. In 4., round is an adverb, modifying the verb went. It is on the basis of its reliability, as exemplified in the above instances, that the functional principle is chosen over the rest. However, as it has been hinted above, this framework can only account for hedges at the lexical level. It therefore became necessary for us to use another framework which could be used to analyse hedges at the syntactic level. We thus resorted to Hyland s (1998) model. In Hyland s (1998) model, the terms grammatical and strategic (p. 103) hedges are used to distinguish lexical signals from regularly used grammatical patterns and other means of expressing 518

5 reservation in the data analysed. This division draws attention to the preponderance of lexical mitigation in the English and Chemistry masters theses and to conveniently categorize a diversity of expressions that might fail to be included. According to the model, the means by which grammatical hedges are expressed vary and may include syntactic structures such as questions, conditional clauses, and contrast markers as well as some formulaic phrases. Literature Review Some studies on PLs include Vigneshwaran, Padmanabha, and Devanna (2013), Cardeñosa, Gallardo, and Toni (2008), and Paiva (2000). However, few of such have looked at PLs from a linguistic perspective. Those on the discourse of PL have also not given serious attention to hedging strategies employed by the PL writers, even though hedging has been argued to be a commonplace phenomenon in medical discourses (Antic, 2009; Hyland, 2006b). And we have mentioned already that even though hedging has received attention from scholars from multidisciplinary backgrounds, the PL has not been explored in terms of hedging use. Hence, the literature focuses on studies on hedges within the medical discourse/writing in general. As Antic (2009) opined, medical papers provide interesting and useful examples of the use of hedging in scientific discourse because they are related to matters which influence significant issues of our lives. (p. 4). A few studies such as Zhen (2007), Hyland (2006b), and Skelton (1997) have explored tentativeness in medical writing in general. As an attempt to provide a pedagogical information for students in the medical field, Antic (2009) provided some useful information on medical writing, focusing on both the micro and the macro levels. She highlighted the structure, content, and language use in medical writing. In relation to language use, she was quick to add that hedges are one of the commonplace linguistic resources in medical writing. Medicine is said to be an anthropocentric discipline (Antic, 2009; Biglan 1973), so writers must exercise caution in making claims and advancing propositions not with unwarranted overconfidence (Hyland, 2006a, 2006b). This caution is achieved through speculative language. She delineated the functions of hedges, though not supported by any data, as expression of uncertainty, scepticism, evidence of open-mindedness, and a shield against any hostile counter-claim. These non-empirically grounded functions of hedges outlined are supported by the proceeding empirical studies. Zhen (2007) conducted a pragmatic study on hedges in medical writing. The writer argued that hedges perform a number of pragmatic functions in medical writing. These he outlined as making claim more precise, objective, and polite. The study also categorised the frequently used words or expressions for hedging. The author argued, When writing medical discourse, writers may select a hedge to present a proposition with greater accuracy, or to gain some protection from the possible criticism, or as a device that offers due recognition and deference to readers' views. In summary, hedges can be adopted to prevent or temper conflicts or negative reactions of a reader, to protect face of the reader by minimizing certain facethreatening acts in order to pave the way for a smooth, cooperative and successful course of communication. (p. 7) The study focused on certain words that are employed as hedges, but it did not include a detailed analysis of lexico-grammatical resources as hedges. Besides, it focused on academic medical writing, unlike the present study, which focuses on a specialised genre in medical communication: PLs. Skelton (1997) conducted a study which looked at how medical writers in general talk about things which they deem to be true, possible, and untrue, focusing on research papers published in three leading medical journals since He identified three main types of truth: contextualized truth, evidential truth, and interpreted truth. He explained that these deal, respectively, with truth as the research tradition states it to be, truth as the statistical evidence states it to be, and truth as a matter of deriving possible non-statistical 519

6 meaning from findings (p. 9). Writers also make frequent explicit reference to the extent to which they are committed to the propositions expressed in statements about truth the manner in which they do so is discussed, with a distinction being drawn between propositions and comments. The increasing demand of professionalism in pharmaceutical industry requires, among other things, politeness in language use, adherence to conventionalised norms, and respect for consumers rights. Results and Discussion The findings of the present study are discussed along the various lexico-grammatical forms used as hedging devices in PLs. In analysing the data, we employed the terms lexical and strategic (i.e., grammatical) hedges (Hyland, 1998) to distinguish lexical signals from regularly used grammatical patterns and other means of expressing reservation. This division is to draw attention to the preponderance of lexical mitigation in the medical leaflets collected within the Cape Coast Metropolis in order to conveniently categorize a diversity of expressions that might fail to be included. Table 1 shows the frequency of forms of hedging in the data based on lexical and strategic (grammatical) categories. TABLE 1 Frequency of Lexical and Strategic Hedges in PLs (per 8, 387 words) Linguistic Types Category of Hedges Frequency Pragmatic Function Lexical Hedges Non-Lexical (Strategic) Hedges TOTAL Auxiliary Verbs 237 (30.3 %) Writer-Oriented Lexical Verbs 110 (14.1 %) Reader-Oriented Adverbs 100 (12.8 %) Accuracy-Oriented Adjectives 94 (12.0 %) Writer-Oriented Nouns 7 (0.9 %) Writer-Oriented Passives 111 (14.2 %) Writer-Oriented Conditional Clauses 64 (8.2 %) Reader-Oriented Admission to Lack of Knowledge 33 (4.2 %) Reader-Oriented Adverb Phrases 11 (1.4 %) Accuracy-Oriented Reference to Authority 10 (1.3%) Writer-Oriented Comparison to General Behaviour of all 4 (0.5 %) Writer-Oriented Medicines 781 Table 1 above suggests that the forms of hedging devices used in the medical leaflets within the Cape Coast Metropolis are both lexical and grammatical in nature. We first proceed to discuss the findings related to the various lexical expressions used as hedging devices. The strategic or grammatical hedges are subsequently discussed. Lexical Hedges As depicted in the table above, hedging in medical leaflets within the Cape Coast Metropolis is principally a lexical phenomenon. Out of the total number of 781 incidents of hedges employed in the data, 548 representing 70.2% are of the lexical category. They are modal auxiliary verbs, lexical verbs, adverbs, adjectives, and nouns in that order. The preponderant use of lexical forms as hedges in the data is in line with Holmes s (1988) view that lexical hedges, rather than phrasal or syntactic structures, represent the most 520

7 common means of realizing epistemic modality in English. In this section, we attempt to identify the various lexical forms of hedges, paying attention to their semantic implications. Modal auxiliary verbs The data indicates that modal auxiliary verbs present the most frequent means of hedging lexically in medical leaflets. Out of the total of 584 lexical hedges in the data, modal verbs were used in 237 instances, representing 30.3%. The distribution of modal verbs in the data shows that writers of medical leaflets have very limited choices. They only employ may, might, can, and could as hedging devices to the neglect of other epistemic forms of the auxiliary verb such as would, should, and must. However, limited in the data, modals cover a wide range of meanings and may overlap each other, depending on the differences in contexts within which they may occur (Huddleston, 1971; Palmer, 1970). We now attempt to discuss the individual modals, considering the meaning each invests in the expression, thereby bringing their hedging role to the fore. May/Might From the data, May and Might are used interchangeably to indicate an assessment of possibility. (1) If the symptoms are very severe, your doctor may recommend you start taking your previous prescribed dose again and reduce it more slowly. ML 96 (2) In some cases, the drug may prolong prothrombin time ML 111 (3) Although Fluoxetine does not increase the effect of alcohol, it might affect your judgment or coordination. ML 67 (4) You are advised not to drink alcohol whilst taking METROLEX-F as this might cause unpleasant side effects such as ML 507 It can be argued that the modal verbs may and might in the above extracts express the same epistemic sense of possibility where the authors show the degree of commitment they invest in the claims made. Expressing their claims as sheer possibility, the authors invariably indicate that they are not certain in the propositions made. In example 2, for instance, the data indicates that it is possible for a drug to prolong the time within which the drug will take effect. Similarly, the notion of possibility is expressed in the use of might in example 3 where it is again indicated that it is possible for the drug to affect one s judgment or coordination. Can/Could The epistemic can and could could be said to be similar to may and might in that they also express tentative possibility (e.g., Perkins, 1983). Their use in the data suggests the speculative and tentative nature of scientific writing. The following examples are typical of this view: (5) Serious or even fatal reactions can occur if you are taking other medicines used to treat depression. ML 35 (6) High doses of caffeine can cause tremor and palpitations. ML 451 (7) Some antimalarials as halofantrine and quinine could influence the ECG pattern. ML 224 (8) Large doses could cause intestinal obstruction. ML 458 From the extracts above, can and could, like may and might, express possibility. The tentativeness of the modal expression echoed by can and could in examples 5, 6, 7, and 8 can be made more pronounced when each of them is paraphrased as It is possible In that case, extract 6, for instance, would read: It is 521

8 possible that high doses of caffeine will cause tremor and palpitations. The similarity between the epistemic uses of could and may is further illustrated by their combination in the data together with the co-occurrence of possibility as exemplified below: (9) Possibly, Gastric emptying can achieve a worthwhile recovery. ML 473 (10) Metrolex-F could possibly be affected by or have an effect on other medicines. ML 508 Epistemic lexical verbs Following modal verbs, epistemic lexical verbs emerge in the data as the next most preferred form of lexical hedges. Out of the 548 lexical hedges in the data, lexical verbs occur 110 times, representing 14.1%. The lexical verbs in the data express subjectivity and are generally used to hedge either commitment or assertiveness. They express both the mode of knowing and the source of that knowledge, thereby conveying the implications about the reliability of the knowledge itself. The epistemic lexical verbs carry some import of both judgment and evidence to suggest a non-committal means of making an assertion. They can thus be categorised into epistemic judgment verbs and epistemic evidential verbs. Epistemic judgment verbs According to Perkins (1983), epistemic judgment verbs reflect appraisals by the speaker of the factive status of events and include speculation and deduction. In the data, these speculative verbs indicate that there is some conjecture about the truth of the proposition. They comprise the use of performative verbs which perform rather than describe the acts they label: (11) If pains persist after three days use of Fluoxetine 20, it is advised that you consult your doctor. ML 99 (12) It is recommended that you do not drink alcohol with this medicine. ML 66 The tentativeness of the emboldened verbs above clearly demonstrates the speculative nature of the performatives. The writer s use of the verb advised to suggest tentativeness shows that he admits some degree of uncertainty he invests in the proposition. The speculation conveyed by the verb in example 11 is made clearer in that of 12 where it becomes more obvious that the writer is making a conjecture based on some deduction from experience. The claim is thus captured as a suggestion rather than an expression of concrete fact. This speculative category also includes verbs which involve unobservable cognitive states or processes, and which do not obviously perform tangible acts in the way the core examples of speech act verbs as exemplified above. Unlike the performatives, these verbs appear to give a more conjectural than assertive meaning to the propositions which follow, hypothesizing a world in which something might be true. (13) It is hoped that maximum plasma concentrations are generally seen 2 to 4 hours after administration. ML 20 (14) It is believed that in vivo tests revealed no structural chromosome damage. ML 31 As evident in extracts 13 and 14, the propositions made are similar to those in 11 and 12 in the sense that both sets of examples mark the authors degree of commitment to their claims. They both reiterate the notion of uncertainty on the part of writers of medical leaflets. However, the verbs hoped (13) and believed (14) can be said to be different from advised (11) and recommended (12) on the basis that the former category of verbs is unobservable. For example, the import of speculation made with the use of the verb hoped is cognitive in nature. On the other hand, the speculation made possible by recommended is more physical in form and much observable. 522

9 Epistemic evidential verbs As already indicated, the second category of lexical verbs that is present in the data is the evidential verbs. These types of lexical verbs provide evidentiary justification either based on the evidence of the writer s senses or the feasibility of matching evidence to goals (Hyland, 1998). (15) Repeated dose, oral, chronic toxicity in rats showed altered liver weights with some slight centrilobular swelling and hepatocellular vacuolation. ML 28 (16) Other in vitro test suggests that both cause a marked diminution of nucleic acid synthesis. ML 212 Contrary to examples of evidential verbs which are usually based on opinions from the literature (Atkinson, 1999), the evidential verbs in the above extracts suggest that the claim made is based on evidence from the writer s own observation, particularly with the help of their senses. Epistemic adverbs Epistemic adverbs are the third most frequent means of hedging in the data. Out of the total number of 548 lexical hedges used in the data, adverbs comprised 100 instances, representing 12.8%. From a semantic point of view, the epistemic adverbs in the data are integrated as an element of the clause, functioning principally as adjuncts or disjuncts, which may intensify or tone down the proposition expressed in the verb. Quirk, Greeenbaum, Leech, & Svartvik (1972) refer to this category of intensifying adjuncts as downtoners (p. 452) due to the lowering effect they have on the modified verb. A number of frequency adverbs in the data can be identified as downtoners. (17) This effect quite greater when quinine was infused after atemether/lumefantrine combination. ML 292 (18) Peak serum concentrations are achieved approximately 1 hour after ingestion. ML 102 (19) Dizziness, sedation and hypotension are more likely to occur in elderly patients. ML 482 In each of the extracts above, the adverb has a lowering effect on the main verb or the adjective, and it is in this that the modality expressed is made possible. For example, in extract 17, the adverb quite reduces the force to the claim made by the adjective greater, thereby toning down the force of the adjective. In like manner, the adverbs approximately and more likely in 18 and 19 respectively tone down the effects of the verbs achieved and occur. Disjuncts, or what Halliday (1994) calls probability adjuncts, are another type of adverbs which are used to hedge in the data. They detach themselves from the other parts of the sentence, commenting on the style and form of what is being said or defining in some way the conditions under which authority is being assumed for the statement (Quirk & Greenbaum, 1990). Some items in this class of adverbs therefore indicate that a generalization is being made (Quirk et al., 1972) and therefore hedge the accompanying statement: (20) Side effects are generally mild and transient. ML 117 (21) As a weak acid, acetylsalicylic acid is largely unionized in the acid environment of the stomach. ML 471 The element of uncertainty expressed by the hedges above is inherent in the writer s attempt to generalize claims without being specific or emphatic. It tends to show an idealized situation on the ground. 523

10 Epistemic adjectives The last predominantly used lexical hedges in the data are epistemic adjectives. With a total of 94 instances in the data, epistemic adjectives represent 12.0% of the total number of lexical hedges. The most predominantly used epistemic adjective in the data is possible. According to Coates (1987), possible has both root and epistemic meanings, roughly parallel to the modal verbs can and may. It may also be more accurately related to the notions of dispositions or abilities and to writer s beliefs which may be paraphrased as possible for or possible that, as in the following usage: (22) It is possible that iso-enzyme induction could alter the therapeutic effects of drugs that are predominantly metabolized by these enzymes. ML 298 In the above excerpt, the epistemic adjective expresses possibility, which involves the writer s confidence in the truth rather than an assessment of enabling conditions, even if the writer seeks to hide this judgment. The epistemic meaning in the extract suggests an interpretation that the outcome of the claim is currently only imaginable rather than likely. In addition to taking complementation with be, possible, as a stance marker, is used attributively with some range of nouns: (23) The possible risks associated with prescribing VERMOX 500 mg drug during pregnancy should be weighed against the expected therapeutic benefits. ML 8 (24) The apparent elimination half-life after an oral dose ranges from 3 to 6 hours in most patients. ML 25 This use restricts the reference of the noun while still qualifying the writer s position. For example, possible and apparent in 23 and 24 restrict the nouns risks and elimination. The epistemic use of possible in these contexts, like other contexts, does not refer to physical objects in the real world, but to what Lyons (1977) calls second or third order entities or abstract propositions which may or may not exist. Thus, concepts and experiential content can be organized into impersonal constructions, the contribution of modal adjectives allowing the writer to hedge the process described. Other adjectives, which only occasionally occur in the data as hedging devices, include significant, apparent, and potential. (25) There was a significant (around 30-40%) reduction in plasma levels of lumefantrine, possibly due to lower absorption ML 167 (26) Amodiaquine should be used during pregnancy only if the potential benefit justifies the potential risk to the foetuses. ML 250 Epistemic nouns From the data, epistemic nouns occur as the least preferred lexical means of hedging. The data provides 7 modal nouns representing (0.9 %). They include probability, possibility, and tendency. (27) The probability of elimination of half-life after an oral dose ranges from 3 to 6 hours in most patients. ML 25 (28) To avoid the possibility of VITAFORCE having any effects on other prescribed medicines, always take VITAFORCE at least 2 hours before or 2 hours after taking any other medicines. ML 314 (29) It (Asmadrin) has the tendency oxidize enzyme system found in hepatic endoplasmic reticulum. ML 401 In each of the instances above, the epistemic nouns hedge the proposition of the writer either by indicating 524

11 that his/her position on the matter is just an opinion rather than a statement of fact or likelihood. Non-Lexical Hedges Passive constructions Apart from the use lexical structures as hedging devices, writers in the data explore other grammatical means to hedge responsibility. The most pervasive of all these non-lexical forms of hedges is the use of passive constructions, occurring 111 times, representing 14.2 % of the hedges. With the use of passive voice as hedging devices, the writers are able to suppress human agency by way of distancing themselves from the claims made. (30) Corticosteroids are thought to act by the induction of phospholipase A2 inhibitory proteins, collectively called lipocortins. ML 146 (31) It is recommended that you do not drink alcohol with this medicine. ML 66 (32) No carcinogenic effects were observed in the mouse or rat. ML 29 In addition to the use of passives, the writers distance themselves from their propositions through the use of abstract rhetors (Halloran, 1984), which imply that rhetorical acts can be accomplished without human volition: (33) Repeated dose, oral, chronic toxicity in rats showed altered liver weights with some slight centrilobular swelling and hepatocellular vacuolation. ML 28 (34) Other in vitro test suggests that both cause a marked diminution of nucleic acid synthesis. ML 212 (35) Studies have demonstrated a pharmacokinetic interaction between diclofenac and salichlates when both are co-administered in anti-inflamatory doses. ML 200 In the above extracts, the authors seem to suggest that whatever claims being made in the various medical leaflets emanate from the data themselves. The absence of human agency in examples 33, 34, and 35 indicate that the outcome of the respective research is made possible by the data under discussion rather than a discovery made by the writer. It is worth noting that this convention of scientific discourse allows the writer to employ verbs which contribute to the impersonalization of the discourse, thereby making the text or the data the source of epistemic judgment. The implication is that any reasonable and informed reader would draw the same conclusion. Such forms obscure the verbal and mental processes by nominalising a personal projection, thereby encouraging an interpretation close to make us think that X or lead us to the conclusion that X, rather than my interpretation is that X (Halliday, 1988, p. 174). Conditional clauses Another non-lexical means of hedging that the data presents is the use of conditional clauses. In this kind of strategy, the writers distance themselves from the proposition made by drawing their audience s attention to the fact that the claim that is being made, usually in the main clause, holds true only on condition that what is expressed in the subordinate clause (i.e., the conditional clause) is true. More often than not in the data, the epistemic import of the claim rendered in the conditional clause is deepened with the use of epistemic modal verbs in the main clause, as in the case of 36 and 38 below: (36) If you have liver problems, a lower dose may be required. ML 36 (37) If you have any of these symptoms (i.e. serotonin or neuroleptic), contact your doctor immediately as you might need to stop taking Fluoxetine. ML

12 (38) If the symptoms are very severe, your doctor may recommend you start taking your previous prescribed dose again and reduce it more slowly. ML 96 The notion of uncertainty in the claims above is expressed with use of conditional clauses, making one circumstance dependent on another and thus hedging the certainty of the outcome. More precisely, the conditional clauses in the examples above hedge the propositions, creating a possible alternative world where the expressed fact is not epistemically accessible, but relate to a speculative state of affairs. The import of this type of non-lexical hedges is that the writers suggest to have some doubts regarding the authenticity of the claim made. In the case of example 38, for instance, the truth of the proposition (i.e., recommendation from a doctor) is premised on the fact that the symptoms be severe. And if it turns out that there are no severe symptoms, then there is the likelihood that there would be no recommendation to be made by one s doctor. The same is the case with example 36 where the claim is believed by the writer to be true if and only if the premise (i.e., having lever problems) is true. Adverb phrase Adverb phrases are yet another means of hedging non-lexically in the data. In this strategy too, like before, the scope of hedging covers the use of other strategies such as conditional clauses, epistemic modal verbs, and epistemic adjectives. These strategies, which accompany the adverb phrase, seem to reinforce the import of hedging expressed by the writer. (39) Talk to your doctor as soon as possible if you might be pregnant. ML 69 (40) In some cases, the drug may prolong prothrombin time ML 111 (41) In rare cases significant poisoning, acute renal failure and liver damage are possible. ML 429 In addition to the grammatical forms of hedges discussed above, it was discovered that the data employs some discourse-based strategies whose grammatical patterns are almost difficult to determine. These strategies are distinguished by their reference to a particular aspect of scientific procedures of reporting which exhibit formal realizations. They are thus labelled in our analysis according to the function they seem to perform. They include strategies which suggest Lack of Knowledge on the part of the writer, Reference to Authority, and Comparison to General Behaviour of all Medicine. Admission to lack of knowledge According to Hyland (1998), admission to lack of knowledge is a non-lexical or grammatical strategy of hedging which comments on the existing knowledge of the author. The essence of this strategy stems from the view people hold that when it comes to judging claims, readers often would want to know how much confidence the writer invests in the propositions made. And one of the clearest means of distinguishing between conditionally true statements and speculative possibilities is making known the depth of your knowledge with respect to the claim. It is important to note that this use of hedging is different from Swales (1990) discussion on establishing a research niche, which is an important means by which writers are able to fix their work in an evidential context of uncertainty. (42) It is not known whether mebendazole is excreted in human breast milk. ML 9 (43) To date, there are no known interactions with other drugs. ML 131 (44) There is no evidence of teratogenic effect to the fetus. ML 106 If we consider the excerpts above, we can say that the structures suggest some non-commitment on the part of the authors. The examples indicate that the writers proclaim boldly that they are not knowledgeable 526

13 on the realities on the ground. They therefore admit their lack of full knowledge, which is an indication of humility, an important feature of academic writing (Irvin, 2001). In another instance, the authors admit that they lack full knowledge on a phenomenon due to inadequate information. (45) Only limited information concerning the long-term safety of Fluoxetine on growth, puberty, mental, emotional and behavioural development in this age group is available. ML 51 (46) At present, there are insufficient clinical data to recommend the use of cetirizine in children under the age of 2. ML 129 Reference to authority The data again suggests that writers of medical leaflets hedge their claims non-lexically by making reference to an authority. By this means, the writers usually make their claims based on the views of an institution, which is an umbrella body that presides over the practices of the medical field. This non-lexical form of hedging usually begins with the phrase according to : (47) According to the new WHO malaria treatment guidelines, uncomplicated falciparum malaria must be treated with artemisinin when used correctly ML 274 (48) In accordance with currently accepted views on the employment of drugs during pregnancy, it should not be prescribed in the first three months of pregnancy unless its use is strictly indicated. ML 187 (49) Both amoxicillin and clavulanic acid belong to category labeled B according to Food and Drug Administration (FDA) definitions. ML 107 As evident in extracts 47, and 49, for instance, the researchers present their claims from the view point of WHO and FDA, which are institutions in the medical field manned by experts in the field. With the use of this strategy, the writers indicate that they are not speaking from a secure source of knowledge and that their claims are only but opinions expressed by the institutions. Thus, the claims are merely what pertains in the literature and perhaps have been accepted to be true. In some instances, too, the reference is not made to an institution, but to studies in the literature. (50) Inhibition of protein synthesis as the basic mechanism as the basic mechanism of action is suggested in studies which showed more morphological changes ML 213 (51) Experts recommend we eat five portions of fruits and vegetables a day and keep to a diet low in saturated fats. ML 317 It is important to note that this non-lexical means of hedging identified in the data is different from the use of reportive verbs (see Hyland, 1995) since in this, the hedging element is not limited to any verb but the entire structure which is emboldened. Comparison to general behaviour of medicines The final non-lexical means of hedging claims in the data hinges on drawing a comparison between the medicine or phenomenon under discussion and the seeming behaviour of all medicines in general. (52) Like all medicines, Fluoxetine can cause side effects, although not everyone gets them. ML 97 (53) As with all medicines, you may occasionally experience some side effects. ML 498 (54) Like all other quinolone derivatives, it is thought to inhibit heme polymerase activity. ML

14 From the above, the writers hedge by providing a caveat that the claim they are making is based on the general behaviour of all medicines as commonly known by all. What this means is that they can only be wrong when or if the common view held by all about the behaviour of all medicines is wrong. Conclusion Exploring the lexico-grammatical forms used in PLs as realizations of hedging, the present study sought to find out what is peculiar about pharmaceutical discourse in general and PLs in particular. In summary, we have provided an account for the ways hedging is linguistically realized in PLs in the general pharmaceutical discourse community, given that the corpora were gathered from all forms of drugs from different geopolitical settings around the globe. From the data, it was observed that the forms of hedges are both lexical and non-lexical, with the former being the most preferred form. According to the data, the lexical hedges comprised modal verbs, adverbs, lexical verbs, adjectives, and nouns. It was found that whereas the most frequently used lexical forms of hedges in the data were modal verbs, the least featured were nouns. Concerning the non-lexical hedges, however, it was observed that apart from the use of passive constructions, conditional clauses, and adverb phrases, there are other forms of discourse-based grammatical strategies that are used to hedge. They include admission to lack of knowledge, reference to authority, and comparison to general behaviour of medicines. The aforementioned findings have implications for health communication in general and for pharmaceutical discourses in particular. First, the study contributes to the scholarship on direct-to-consumerbased discourse in the pharmaceutical discourse community. The use of hedging, in some sense, shows how key players in this community respect patients rights by showing some deference to them through the use of hedging strategies. Again, the study has confirmed the assertion that hedging devices are part of the medical language (Antic, 2009). Most importantly, it has implication for hedging in PLs as there has been some tentative realizations about certain peculiar hedging strategies in PLs, for example, a lack of knowledge and comparison to the general effects of medicines. The Authors Esther Serwaah Afreh is a Lecturer in the Department of English. She holds a PhD in Linguistics. Her research interest includes metaphors in political discourses and language and identity. Her papers have appeared in the International Journal of English Language, Literature, and Humanities. Department of English College of Humanities and Social Sciences Kwame Nkrumah University of Science and Technology, Ghana esafreh@gmail.com Osei Yaw Akoto is an Assistant Lecturer in the Department of English, Kwame Nkrumah University of Science and Technology. His research interest includes onomastics, academic discourse using corpus linguistics approaches, and language and identity. He has published in the Journal of Language and Literature. Department of English College of Humanities and Social Sciences Kwame Nkrumah University of Science and Technology, Ghana oseiyaw.akoto@yahoo.com 528

15 Kwodwo-Adam Moses is an Assistant Lecturer in the Department of English, Kwame Nkrumah University of Science and Technology. His research interests include English for academic purposes and legal discourse. He has published in English for Specific Purpose World. Department of English College of Humanities and Social Sciences Kwame Nkrumah University of Science and Technology, Ghana References Antic, Z. (2009). Some implications for teaching scientific medical writing. Acta Fac Med Naiss, 26(1), Atkinson, P. (1999). Medical discourse, evidentiality and the construction of professional responsibility. In S. Sarangi & C. Roberts (Eds.), Talk, work, and institutional order (pp ). Mouton de Gruyter. Bazerman, C. (1988). Shaping written knowledge: The genre and activity of the experimental articles in science. Madison, WI: University of Wisconsin Press. Benesch, H. (2008). Buildings, change, futurity: Reading Seattle Public Library and Sendai Mediateque. Proceedings of Design Inquiries, May (pp ). Sweden: Stockholm. Biglan, A. (1973). Relationships between subject matter characteristics and the structure and output of university departments. Journal of Applied Psychology, 57(3), Brown, G. (1995). Speakers, listeners and communication: Exploration in discourse analysis. Cambridge, UK: Cambridge University Press. Brown, P., & Levinson, S. (1987). Politeness: Some universals in language usage. Cambridge, UK: Cambridge University Press. Burns, L. C. (1991). Vagueness: An investigation into natural languages and the Sorites paradox. Dordrecht, NL: Kluwer Academic Publishers. Cardeñosa, J., Gallardo, C., & Toni, A. (2008). Methodology for language analysis and generation in closed domains: Pharmaceutical leaflet. Information Science and Computing, 2, Casas, A. V., & Parada, C. R. (2008). Patients' right to information: a review of the regulatory and ethical framework. Bulletin of Medical Library Association, 66(1), Channel, J. (1994). Vague language. Oxford, UK: Oxford University Press. Coates, J. (1987). Epistemic modality and spoken discourse. Transactions of the Philological Society, 85, Creswell, J. W. (1994). Research design: Qualitative and quantitative approaches. Thousand Oaks, CA: Sage. Dubois, J. W. (1987). The discourse basis of ergativity. Language, 63(4), Gilbert, G. (1976). The transformation of research findings into scientific knowledge. Social Studies of Science, 6, Grabe, W., & Kaplan, R. B. (1997). On the writing of science and the science of writing: Hedging in science texts and elsewhere. In R. Markkanen & H. Schroder (Eds.), Hedging and discourse: Approaches to the analysis of a pragmatic phenomenon in academic texts (pp ). New York, NY: Walter de Gruyter. Gross, A. G., & Chesley, P. (2012). Hedging, stance, and voice in medical research articles: A study of deviance. In K. Hyland & C. S. Guinda (Eds.), Stance and voice in written academic genres (pp ). London, UK: Palgrave Macmillan. Fraenkel, J. R., & Wallen, N. E. (1990). How to design and evaluate research in education. New York, NY: McGraw-Hill. 529