Eğitimde Kuram ve Uygulama 2008 , 4 (1):159-174 Journal of Theory and Practice in Education Makaleler/Articles ISSN: 1304-9496 http://eku.comu.edu.tr/index/4/1/mcelik.pdf A DESCRIPTION OF TURKISH-ENGLISH PHONOLOGY FOR TEACHING ENGLISH IN TURKEY1 TÜRKİYE’DE İNGİLİZCE ÖĞRETİMİ İÇİN TÜRK-İNGİLİZCESİNİN SESLETİMİNİN TANIMLANMASI 2 Mehmet ÇELİK ABSTRACT The use of English as an international language has resulted inevitably in a blending of English and the first languages of the users. One particular consequence of this contact has been the creation of a system involving the phonemic features of the two languages, one such variety being Turkish-English phonology. The objective of this article is to provide Turkish teachers of English with a description of realistic and at the same time intelligible pronunciation for teaching and testing purposes. For this purpose, data from bilinguals and teacher trainers as well as advanced learners of English have been utilized. In view of the strong arguments in the literature that it is unrealistic to expect learners to accomplish native speaker norms in pronunciation, what is needed, methodologically, is an approach or framework that recognizes the learner’s linguistic background as well as the contexts that involve nonnative-nonnative discourse participants. The suggested framework involves a reduced but intelligible phonological system for language learners to successfully communicate in English internationally. The paper concludes the description of Turkish- English phonology by reducing the number of phonemes by 8 from a studied total of 23 phonemes from Received Pronunciation. Teachers of pronunciation can make use of the suggested pronunciation framework for effective, viable and realistic teaching targets. Key Words: international communication, intelligibility, bilingual phonology, Turkish-English, testing ÖZ İngilizcenin uluslararası bir dil olarak kullanımı kaçınılmaz olarak İngilizce ile bu dili kullananların bririnci dilleri arasında bir karışımı ortaya çıkarmıştır. Bu temasın önemli sonuçlarından biri iki dilin fonemik özelliklerini içeren bir sistem oluşmasıdır ki bu Türk-İngilizce sesbilimidir (sesletimidir). Bu makalenin amacı Türk İnglizce öğretmenlerine test ve öğretim amaçlarına yönelik gerçekçi ve aynı zamanda anlaşılır bir telafuz sunmaktır. Bu amaçla iki dilli bireyler, öğretmen eğiticileri ve ileri seviyede ingilizce öğrencilerinden elde edilen verilerden yararlanılmıştır. Alanyazında öğrencilerden ana dil normlarında ikinci dil telafuzu beklemenin gerçekçi olmaması yönünde yer alan kuvvetli gerekçeler göz önünde bulundurularak, metodolojik olarak gereken, İngilizce’yi ana dil olarak kullanmayan söylem katılımcıları içeren bağlamların yanısıra öğrencinin dilsel birikimini tanıyan bir yaklaşımın veya yapının dikkate alınmasıdır. Önerilen model, İngilizce öğrenenlerin uluslararası platformda başarılı bir şekilde iletişim kurmalarına yönelik indirgenmiş fakat anlaşılır bir ses sistemi içermektedir. Bu makale Standart İngiliz İngilizcesi’nin toplam 23 foneminden 8 tanesinin çıkartılması sonucu ortaya çıkan Türk İngilizce sesletiminin tarifini ortaya koymaktadır. Telaffuz öğretmenleri önerilen telaffuz yapısından etkili, uygulanabilir ve gerçekçi öğretim amaçları için faydalanabilirler. Anahtar Kelimeler: uluslararası iletişim, anlaşılabilirlik, ikidilli sesletim, Türk İngilizcesi, ölçme 1 The article is a revised version of the paper titled ‘Bilingual Phonology as a Basis for Teaching Pronunciation’ presented at the 4th International ELT Research Conference “Reflecting on Insights from ELT Research” 26-28 May 2005, Çanakkale Onsekizmart University. 2 Hacettepe University, Department of English Language Teaching. E-mail:[email protected] © Çanakkale Onsekiz Mart University, Faculty of Education. All rights reserved. © Çanakkale Onsekiz Mart Üniversitesi, Eğitim Fakültesi. Bütün hakları saklıdır. A Description of Turkish-English Phonology for Teaching English in Turkey INTRODUCTION Pronunciation has been an area of major concern in the teaching of second/foreign languages since the demise of the grammar-translation approach. The insistence on perfect pronunciation peaked in the decades when the Audio-Lingual Approach was unrivalled. When it was superseded by the now most popular Communicative Language Teaching (CLT), pronunciation became relegated to a less important position since the chief concern was regarded as the communication of meaning, as opposed to the form (pronunciation) that carries it. However, the issue of pronunciation has been one major point of dissatisfaction and complaint by many learners and users of English (Canagarajah, 1999), whose number is estimated to have reached one billion. On the right grounds or not, these users wish to get rid of their accent for various reasons, at a cost of reducing the focus on, perhaps, more relevant domains and skills. As second language research has revealed in the past decades, pronunciation is the only component of second/foreign language learning process in which adult learners are highly unlikely to achieve native-like accuracy no matter how much effort is placed on it (Marinova-Todd, Marshall, and Snow, 2000). This testimonial has warranted a more realistic approach by re-examining the pronunciation targets in syllabus designs. In this sense, this study answers McKay’s (2002: 41) plea: “… teaching English as an international language requires that researchers and educators thoroughly examine individual learners’ specific uses of English within their particular speech community as a basis for determining learning goals.” Similar demands are made by others (Alptekin, 2002; Jenkins, 2000). In view of these observations, a framework can be developed for both the teaching and testing of pronunciation, based on recognition of the individual learners’ linguistic background, i.e. first language. The suggested framework is based on the practices and communication strategies of Turkish- English bilinguals as well as learners of English when they communicate with native speakers (NSs) as well as non-native speakers (NNSs) of English. Although the contextual focus of the study is on Turkish-English, its broader appeal lies in its setting an example for other foreign language teaching/learning milieu in their efforts towards a realistic and workable pronunciation syllabus and assessment. The study first examines the stance of CLT in regards to providing viable, achievable and learnable pronunciation objectives. Secondly, it considers recent studies that stress the international nature of the task in hand and, therefore, argue against the imposition of native speaker forms as achievable targets. Thirdly, it examines the communication strategies of Turkish-English bilinguals in struggling with the task of getting the pronunciation right. The study finally suggests a slightly reduced and thus a more feasible and teachable English phonological system by considering the Eğitimde Kuram ve Uygulama / Journal of Theory and Practice in Education 160 http://eku.comu.edu.tr/index/4/1/mcelik.pdf Mehmet Çelik Egitimde Kuram ve Uygulama Journal of Theory and Practice in Education 2008 , 4 (1):159-174 two main and influential varieties of English: Received Pronunciation (RP) and General American (GA). PRONUNCIATION AND COMMUNICATIVE LANGUAGE TEACHING The notion of communicative competence was first used by the sociolinguist Dell Hymes as a reaction to the linguist Noam Chomsky’s notion of language, which takes the competence/knowledge of the ‘ideal speaker’ as the core of language phenomenon. Breaking away with the influential Chomskyan tradition of thinking, Hymes (1972) conceptualized ‘communicative competence’ as the ability not only to apply the linguistic (i.e. phonological, morphological, and syntactic) rules of a language to form correct sentences/utterances, but also the ability to utilize and comprehend these utterances ‘appropriately’ in communicative contexts. The term appropriately assumes significance here: it can mean the use of cultural, sociolinguistic as well as pragmatic knowledge since the act of communication does not take place in vacuum, devoid of these precepts. Hymes’ understanding of language as an instant communicative device has been more appealing to language teaching pedagogy because the needs and demands of second/foreign language learners required a working command of language rather than substantial knowledge ‘about’ the language. Therefore, this conception of language has been extensively applied to second language teaching and learning by methodologists since the prevailing purpose of language learning throughout the world has been to communicate. Elaboration of communicative competence by Canale and Swain (1980) and Canale (1983), to note only a few, has produced an influential approach: the Communicative Approach, also known as Communicative Language Teaching (CLT). In the teaching of English both as a second and foreign language, CLT has been exercising great influence on curriculum design and classroom practice. Originated and elaborated upon in the English speaking countries for application in second language classrooms, CLT has also been adopted for distressed foreign language classrooms all over the world. CLT rests on the recognition of native speaker norms and ‘considers target- language based communicative competence to be essential in order for the foreign language learner to participate fully in the target language culture’ (Alptekin 2002:58). Thus, the workload of the learner included learning the target culture as well as a mastery of the target language similar to that of a native speaker, an objective far fetched in view of the amount of time and effort expended by average learner. As has been the case with its predecessors, CLT has developed into various versions as it ran into difficulties arising from its application both in second and foreign language classrooms. The more recent (strong) versions of CLT have had to somewhat de-emphasize the role of native speaker norms in © Çanakkale Onsekiz Mart University, Faculty of Education. All rights reserved. 161 © Çanakkale Onsekiz Mart Üniversitesi, Eğitim Fakültesi. Bütün hakları saklıdır. A Description of Turkish-English Phonology for Teaching English in Turkey setting goals for all competencies, in particular, pronunciation (phonological), as accent can never be rectified to satisfactory levels after the critical period, i.e. puberty. Therefore, the objective of ‘native-like’ pronunciation had to be abandoned as two major observations have appeared to invalidate the position taken by the proponents of the original (weak) versions of communicative competence: 1) impossibility of the native speaker’s phonological system for the (typical) learner to accomplish, and 2) the notion of ideal native speaker is both void and dialectal (Alptekin 2002; McKay 2002; Jenkins 2000). Therefore, this conception of an ego-centric prescription of authentic language use, or real English, primarily designed for use by immigrants in English speaking countries, has recently been challenged extensively for various reasons, ranging from inadequacies of target language culture application in the classroom to the utopian and unrealistic nature of expectations from the learners, and to the impossibility of the acquisition of native speaker phonological systems in non-English speaking countries (Alptekin 2002; McKay 2002; Jenkins 2000; Crystal 1998). Furthermore, the aim of the learner may not be to perform transactions with monolingual native speakers. The aforementioned, and similar, considerations have shaped the recent (strong) version of communicative language teaching, where broader emphasis is placed on the communication of meaning, reducing expectations in respect of phonological accuracy. The issue of whether pronunciation should be taught with as much vigor as other competences (i.e. grammar, vocabulary) and if so, what model should be adopted for pronunciation teaching, has been extensively debated in recent times (Yasukata, 2001; Van der Walt, 2000; Shibles, 1995). Among those that agree on the teaching of pronunciation, Shibles (1995), for instance, pointed out the complicated problems in defining a standard (i.e. uniform, compact) pronunciation and further noted that even Received Pronunciation (RP, which is standard British English) is marred with many dilemmas relating to its extensive variations in written-spoken, institutional-social, formal- colloquial choices. A viable solution, however, comes from studies such as Yasukata (2001) and Van der Walt (2000), who suggest that the notion of comprehensibility and intelligibility could be adopted since the primary purpose in communication is the apprehension of meaning in the broadest sense, and not necessarily the form of the utterance itself. Recent methodological tendency recognizes the plural nature of English; English does not have a monolithic structure but a varied one in numerous respects, as evident in the existence of continental varieties (e.g. General American, Received Pronunciation, Australian and New Zealand English) and their dialects (e.g. Scottish English, Irish English, and Birmingham English in the United Kingdom). In addition to recognition of linguistic varieties, which necessarily forces us to choose one variety as the native speaker form, Alptekin (2002), alongside many others (i.e. McKay, Eğitimde Kuram ve Uygulama / Journal of Theory and Practice in Education 162 http://eku.comu.edu.tr/index/4/1/mcelik.pdf Mehmet Çelik Egitimde Kuram ve Uygulama Journal of Theory and Practice in Education 2008 , 4 (1):159-174 2002; Nihalani, 1997), suggests that both the local and international needs and contexts should be prioritized, which inevitably requires syllabus designers to consider international needs and thus perspectives in the teaching of English. TEACHING ENGLISH AS AN INTERNATIONAL LANGUAGE English is no longer in the hegemony of monolingual native speakers as they are outnumbered by bilingual speakers of English (Crystal 1998). It is in this context that the term “English as an International Language” (EIL), which encapsulates the practices and contributions of users of English with non- English backgrounds, finds currency among scholars such as Alptekin (2002), McKay (2002), and Jenkins (2000, 2002). Adoption of this notion brought with it what would become key terms such as ‘international intelligibility’, ‘international pronunciation’, ‘mutual intelligibility’, and ‘mutual understanding’ in order to emphasize the optimum, perhaps the least number of, requirements for communication. In the same vein, McKay (2002) demands the replacement of native speaker forms with a model that recognizes the international character of English as spoken all over the world. She suggests that language teaching norms not be confined to the Inner Circle countries (such as England and the United states). They should be extended to include not only the Outer Circle countries such as Singapore and India where English has undergone a nativization process (Nihalani, 1997), but also to the Expanding Circle countries such as Turkey and Italy where English is taught as a foreign language. In quests calling for an international perspective, the pronunciation aspect of International English has been one area of research. For instance, Nihalani (1997: 99) adopts a pragmatic approach in the planning of teaching pronunciation when he suggests that “the foreign learner adopts certain core features of English in his pronunciation if he is to use English effectively as an ‘international language’.” He argues convincingly that it is somewhat odd to emphasize the need to teach learners to conform to Received Pronunciation (RP) at a time when no one thinks or even dares ask Scottish speakers or North Americans to do the same. Nihalani’s basic suggestion is that pronunciation should be universally intelligible. In the same vein, Jenkins (2000) argues, through an examination of discourse between non-native speakers, for the recognition of what she terms ‘lingua franca core’, that is, those phonological features that have to be present if non-native speakers wish to remain intelligible among each other. She observes that /T/ and /t/ are permissible as they do not hinder communication: “there is really no justification for doggedly persisting in referring to an item as ‘an error’ if the vast majority the world’s English speakers produce and understand it” (p.160). © Çanakkale Onsekiz Mart University, Faculty of Education. All rights reserved. 163 © Çanakkale Onsekiz Mart Üniversitesi, Eğitim Fakültesi. Bütün hakları saklıdır. A Description of Turkish-English Phonology for Teaching English in Turkey BILINGUAL PHONOLOGY AS A MODEL FOR TEACHING Nihalani’s arguments for the need to form a model of pronunciation for International Language are almost identical to those of Jenkins (2002). What Jenkins calls a ‘lingua franca core’ for pronunciation can actually be worked out to enable non-native speakers of English to converse among themselves, as opposed to discourse between native and non-native speakers (which was essentially the point taken up in Nihalani). Jenkins’ study can be considered a major attempt in that it investigates conversations between non-native speakers of English with Taiwanese and Korean backgrounds. Noting the break-down boundaries for intelligibility based on examples from the recorded conversations, she concludes with an elaborate system of rules for the phonology of English as an International Language. The present study is primarily inspired by both Niahalani (1997) and Jenkins (2002), and therefore takes them as a model for creating a Turkish- English phonology that can be used in communication between bilingual Turkish speakers and NSs on one hand, and NNSs of English on the other. The suggested phonology resembles many other models in that it shares common core features with those of Standard Singaporean English (SSE) and Educated Indian English (EIE) (see below). The recognition of the international status of English and what follows from this into syllabus design and classroom practices is bound to bring about inherent problems associated with the attempt to suggest a framework. Questions such as these are not easy to answer: What is intelligibility? What is the optimum number of pronunciation features for relatively successful communication between NSs of English and NNSs? Smith and Rafiqzad (1983:61) define intelligibility as the “capacity for understanding a word or words when spoken/read in the context of a sentence being spoken/read at natural speed.” It needs to be emphasized here that contextual cues provide vital support for intelligibility. It is important to identify the extent to which intelligibility in NS-NNS or NNS-NNS interaction can break down at the phonological level. If certain phonological features are implicated in the breakdown, they should be part of the pronunciation teaching. Then, the optimum number of phonological features, which can be called ‘bilingual phonology,’ would include those sounds whose less-than- perfect pronunciation (or replacement by features present in the mother tongue) will not break down natural communication. On the issue of what pedagogic model should be adopted for English in a foreign language context, Alptekin (2002:63) suggests that “successful bilinguals with intercultural insights and knowledge serve as pedagogic models in English as an International Language (EIL) rather than the monolingual native speaker.” We may comfortably extend Alptekin’s plea to the area of pronunciation too. Support for this kind of stance comes from McKay (2002), who questions the hegemony of native speaker forms in Eğitimde Kuram ve Uygulama / Journal of Theory and Practice in Education 164 http://eku.comu.edu.tr/index/4/1/mcelik.pdf Mehmet Çelik Egitimde Kuram ve Uygulama Journal of Theory and Practice in Education 2008 , 4 (1):159-174 English language teaching given the de-territorialized nature of international English. Instead, she proposes a model based on the norms of bilingual speakers. It is common knowledge that there are two major varieties (i.e. dialects) that command respect in their countries and elsewhere in the world: RP in the United Kingdom and General American (GA) English (though they suffer from the inherent problems associated with their descriptions). Turkish- English bilinguals usually get their bilingual status through exposure to the varieties spoken in these two countries. More often than not, there are more TE bilinguals with the GA accent because of the intense educational, economic, diplomatic, and military contacts with the United States Besides, what makes a TE bilingual’s accent sound more like GA lies in the relative approximation of the GA and the Turkish phonological systems. For instance, generally speaking, the pronunciation of / r / is audible in the GA variety in all positions in the syllable whereas in RP it is limited to syllable-initial and syllable-medial positions. Given that learners are exposed to English spelling and that Turkish is fundamentally a phonemic language, speakers of Turkish are tempted to pronounce / r / even when they wish to speak with an RP accent. Therefore, TE bilinguals are usually described as having a flavor of the GA accent. METHOD The participants in this study can be classified as Turkish-English bilinguals (5 participants), English-Turkish bilinguals (2), teacher trainers (4), and advanced learners of English (5). The Turkish-English bilinguals were very advanced and balanced users of the two languages with at least a few years of exposure to English in an English speaking country. English-Turkish bilinguals were the children of Turkish immigrants to English speaking countries with their primary, secondary and tertiary education taking place in the English speaking countries. Advanced learners of English were the top five students in an English Language Teaching department where the author worked. The instruments used included interviews, reading tasks, and informed judgments. First, participants were interviewed in a free conversation, recorded for later transcription and analysis. Secondly, they were asked to read some words (see Appendix I), which was also audio recorded. The analyses obtained from the transcriptions clearly described the phonological features. Finally, they were asked to report on the results of transcriptions in regard to whether and how often the identified phonological features occur in the speech of their colleagues, bilinguals and advanced learners of English. Further, they were also asked to listen to the taped conversations and reading tasks to © Çanakkale Onsekiz Mart University, Faculty of Education. All rights reserved. 165 © Çanakkale Onsekiz Mart Üniversitesi, Eğitim Fakültesi. Bütün hakları saklıdır. A Description of Turkish-English Phonology for Teaching English in Turkey discover whether the phonological units under investigation were intelligible both in and out of context. The results of the analyses are presented below. Three strategies were observed to be used by the participants in the study: 1) utilizing variation across GA and RP, 2) collapsing similar sounds, and 3) orthographically induced pronunciation. As the number of participants for each category was few, tables do not include percentage and frequency of use, production and endorsement. Instead, the results should be taken to be the general endorsement of the participants. RESULTS Utilizing Variation across Major Varieties One strategy taken by learners was one that could accommodate both types of preferences as well as uses: GA and RP. This strategy can be justified given that a majority of GA speakers can converse with their RP counterparts with relative ease. As such, TE bilinguals deserve to be allowed to use whichever form they are capable of producing providing they stay within the limits of intelligibility. Having in mind the proximity of Turkish to GA in terms of sound characteristics, I present below certain variations in the two major varieties and indicate for each example what typical strategies TE bilinguals are more likely to adopt. Table 1: Free Variation between GA /r/ and the Absence of it in RP.3 Examples RP GA TE car /kA:/ /kA:r/ /kA:r/ before /bI"fO:/ /bI"fO:r/ /bI"fO:r/ Turkish /"t3:kIS/ /"t3:rkIS/ /"t3:rkIS/ Table 2: Free Variation between RP /@U/ and GA /oU/ Examples RP GA TE go /g@U/ /goU/ /goU/ so /s@U/ /soU/ /soU/ gold /g@Uld/ /goUld/ /goUld/ goal /g@Ul/ /goUl/ /goUl/ post /p@Ust/ /poUst/ /poUst/ 3 The phonetic transcriptions given are all taken from Oxford Advanced Learner’s Dictionary, 2000. Eğitimde Kuram ve Uygulama / Journal of Theory and Practice in Education 166 http://eku.comu.edu.tr/index/4/1/mcelik.pdf Mehmet Çelik Egitimde Kuram ve Uygulama Journal of Theory and Practice in Education 2008 , 4 (1):159-174 Table 3: Free Variation between RP /Q/ and GA /A/. Examples RP GA TE got /gQt/ /gA:t/ /gA:t/ pot /pQt/ /pA:t/ /pA:t/ lot /lQt/ /lA:t/ /lA:t/ lottery /"lQt@ri/ /"lA:t@ri/ /"lA:t@ri/ Table 4: Free Variation between RP /U@/ and GA /Ur/. Examples RP GA TE pure /pjU@/ /pjUr/ (both) cure /kjU@/ /kUr/ (both) sure /SU@/ /SUr/ (both) Collapsing Similar Sounds The second strategy bilinguals appeared to use is the replacement of some English sounds by their approximations in English, which also exist in Turkish. Table 5: /e/ and /{/ are Collapsed to Produce /e/ Examples Phonemically Different TE for both man vs. men /m{n/ vs. /men/ /men/ bad vs. bed /b{d/ vs. /bed/ /bed/ marry vs. merry /"m{rI/ vs. /"merI/ /"merI/ Another case of approximation occurs with the two consonant phonemes (/T/ and /D/) that Turkish does not have. /T/ is replaced by /t/ while /D/ is replaced by /d/, both of the latter are phonemes in Turkish. Consider Table 6 for examples. Table 6: Replacement of /T/ and /D/ by /t/ and /d/ Respectively Examples RP / GA TE three /TrI/ /trI/ thank /T{Nk/ /t{Nk/ think /TINk/ /tINk/ this /DIs/ /dIs/ there /DeIr/ /deIr/ them /Dem/ /dem/ © Çanakkale Onsekiz Mart University, Faculty of Education. All rights reserved. 167 © Çanakkale Onsekiz Mart Üniversitesi, Eğitim Fakültesi. Bütün hakları saklıdır. A Description of Turkish-English Phonology for Teaching English in Turkey Although the consonant /v/ is a phoneme in Turkish whose allophone is /w/, it is not used for its counterpart in English. Interestingly, the glide /w/ replaces most occurrences of /v/ in English. Further, words like ‘have’ /h{v/ can also be pronounced /h{f/ because of the devoicing rule in word final position in Turkish. Table 7: Replacement of /v/ by /w/ Examples RP / GA TE have /h{v/ /h{w/ - /h{f/ very /veri/ /weri/ several /"sevr@l/ /"sewr@l/ moving /"mu:vIN/ /"mu:wIN/ prevent /prI"vent/ /prI"went/ A slightly different strategy involved both collapsing and replacement of the collapsed sounds by a distinct Turkish phoneme. One such case involves three vowels: /A:/ /Q/ and /V/. While they can be interchangeably used, one form, /A/, dominates in bilingual pronunciation. /A/, which exists neither in RP nor GA, is described as a “low, back, open, unrounded vowel” by Lewis. It is more like the ‘a’ of French avoir. Consider Table 8. Table 8: /V/, /Q/ and /A:/ are Collapsed to Produce Turkish /A/ Examples RP GA TE but /bVt/ /bVt/ /bAt/ come /kVm/ /kVm/ /kAm/ car /kA:/ /kA:r/ /kAr/ part /pA:t/ /pA:rt/ /pArt/ lot /lQt/ /lA:t/ /lAt/ got /gQt/ /gA:t/ /gAt/ /pAt/ pot /pQt/ /pA:t/ Orthographically Induced Pronunciation Given the phonemic nature of Turkish, when Turkish learners of English do not know for sure the pronunciation of a letter or sequences of letters in an English word, their strategy is usually one of pronouncing that letter as it is pronounced in Turkish. Including the examples given in Table 9, some pronunciations based on orthography are also stated as intelligible by the participants. Eğitimde Kuram ve Uygulama / Journal of Theory and Practice in Education 168 http://eku.comu.edu.tr/index/4/1/mcelik.pdf