Second and Foreign Language Contexts of Use

Recent research has continued to add to the body of work already existing on the pronunciation needs of L2 speakers to prepare them for interaction with L1 speakers of the language (i.e., for native/nonnative communication). The most recent studies have refined and extended the area of inquiry, moving away from an emphasis on nativelike goals to one that gives greater priority to the listener perspective (both the native's and the nonnative's), with a focus on issues such as the factors involved in the intelligibility or comprehensibility of nonnative speech to native listeners, nonnative listeners' preferred speech rates, and the like (cf. Derwing & Munro, 2001; Derwing, Rossiter, & Munro, 2002). Other recent research in this domain has been investigating the effects of different approaches to pronunciation teaching on learners' accents, including learners' own perspectives, rather than taking the beneficial effects of all pronunciation teaching as a given (cf. Derwing & Rossiter, 2002a, 2002b). Both of these general research directions are already having an effect on second/foreign language pronunciation teaching. Listening activities are becoming more prominent in pronunciation materials, particularly those that help learners deal with the problems of connected speech (in which regard Shockey, 2003, is an important contribution to teacher education). Production activities are moving away from mimicking toward the greater prioritizing of specific pronunciation features, with more priority generally being accorded to suprasegmental than to segmental aspects of the language for this type of communication context.

Meanwhile, contrastive analysis-based research continues in part at least because of the current emphasis in L2 pedagogy on individual learner needs. It would, in any case, have been a serious mistake to throw out the modern contrastive analysis baby with the old contrastive analysis bathwater. Teachers have always continued to believe in the important influence of the mother tongue on L2 pronunciation acquisition, even during periods when researchers were emphatically arguing that L1 transfer was trivial (most notably during the 1970s and 1980s in the United States). Furthermore, the interest in contrastive analytical research itself has never disappeared entirely, even though it is nowadays complemented by an equally robust interest in other approaches to interlanguage phonology, as is evidenced by the range of research papers in Major (1998) and Leather (1999).

Within the current contrastive analysis tradition, there is a growing body of research-based publications for teachers of students from L1s that earlier research had tended to overlook and/or treat superficially, by ignoring, where relevant, the role of local L1-L2 contact. Brown, Deterding, and Low (2000), for example, examine a range of differences between Singaporean and British English, including discourse intonation, pitch range, and lexical stress. Hung (2000, 2002a) uses a contrastive methodology in determining his phonology of Hong Kong English. Deterding and Poedjosoedarmo (1998) is a research-based reference work for teachers, providing both details of the segmental and suprasegmental features of a wide range of different Southeast Asian languages and English, along with practice activities for teachers to use in the classroom. Taking further the current move away from nativelike accents as the goal of pronunciation teaching, the authors question whether learners in countries where the L2 is an official second language should be taught an accent other than their own, a point also raised by Hung (2002b). Further evidence of the continuing pedagogic influence of this type of research is the second edition of Swan and Smith (2002), providing contrastive information for teachers of English of students from a large number L1s, Brown (1997), a book of pronunciation teaching materials for the Singaporean classroom, and Weinberger's (2002) Web site (http://classweb.gmu.edu/accent/), which provides an English passage read by L2 speakers from a comprehensive range of L1 backgrounds.

International Contexts of Use

However, some languages are learned primarily for use in international contexts. In this second case, much of the interaction typically takes place between nonnative speakers (NNSs) from different first language backgrounds, often with no native speakers involved at all. Here, then, we are speaking of an international language, of which English in the expanding circle (Kachru, 1992) is currently the example par excellence, although other languages such as Spanish may, too, be learned for this purpose. Because of the acknowledged position of English at present as the world's principal international language, or lingua franca, the discussion that follows will focus exclusively on the role of international context in the teaching of pronunciation for English as an International Language (EIL). The principles and issues discussed nevertheless apply to the acquisition of the pronunciation of any international language and involve, possibly, the most radical changes of all to L2 pronunciation pedagogy (see also Seidlhofer, this volume).

Essentially research into EIL has demonstrated the importance in EIL communication of pronunciation in general, and of certain pronunciation features in particular. Pronunciation had been marginalized by communicative approaches to language teaching in vogue since the 1980s, in the belief that it was peripheral to successful communication. The EIL research found, on the contrary, both that in interaction between L2 speakers from different L1s, pronunciation plays a critical role in preventing communication breakdowns and that—in line with the distinction between foreign and international languages—the phonological and phonetic factors involved are not necessarily the same as those involved in communication between a native and nonnative speaker of the language.

The main focus of EIL research to date has been the role of pronunciation in promoting intelligibility in NNS–NNS communication, including the part played by accommodation skills. Jenkins (2000, 2002) builds on earlier research in which listeners rated the intelligibility of the pronunciation of speakers from different L1s, such as Smith and Rafiqzad (1979), Smith and Bisazza (1982), Smith and Nelson (1985), and Smith (1992). Her Lingua Franca Core targets those features found in her research to be crucial in promoting intelligible pronunciation for an interlocutor from a different L1: most consonant sounds, vowel quantity, initial and medial consonant clusters, and tonic stress (see Seidlhofer, this volume, for details). Drawing also on Speech/Communication Accommodation Theory (cf. Beebe & Giles, 1984; Giles, Coupland, & Coupland, 1991), Jenkins's research also demonstrates that intelligible pronunciation between speakers from different L1s is not a monolithic phenomenon, but one that requires negotiation and adjustment in accordance with the specific context of the discourse and, above all, in relation to addressor/addressee factors (see Jenkins, 2003, for further discussion and examples of accommodation in EIL communication).

Also on the research agenda, though less extensively explored hitherto, is the link between accent and identity in EIL speech contexts. The prevailing concept of “accent reduction,” with its tendency to regard learners as subjects for speech pathology and to exhort them to lose all traces of their L1 accent in their L2 has been questioned by those working from an EIL perspective. Instead, the concept of “accent addition” is being promoted in accordance with the goals of additive bilingualism and in tune with the current emphasis on learner choice (see, in particular, Pakir, 1999). Based on research into pronunciation attitudes, both that of other scholars (see below on sociopsychological issues) and her own EIL research, Jenkins (2000, pp. 209–210) proposes five stages of pronunciation learning:

• Addition of core [i.e., Lingua Franca Core] items to the learner's productive and receptive repertoire
• Addition of a range of L2 English accents to the learner's receptive repertoire
• Addition of accommodation skills
• Addition of non-core items to the learner's receptive repertoire
• Addition of a range of L1 English accents to the learner's receptive repertoire

Learners who have elected to acquire an accent that enables them both to preserve their L1 identity in their L2 English and to be (pronunciation-wise) intelligible to other NNSs will probably aim for the first three stages. However, they may also wish to be able to understand the pronunciation of NSs, certain features of whose speech can, without prior familiarizing, present particular difficulties for NNS listeners (Bent & Bradlow, 2003). In this case, they will probably aim for all five stages. The critical point, though, is that there is no suggestion of losing their L1 repertoire and, by definition, their L1 identity. This change in emphasis has already filtered through to pronunciation materials, which are tending to incorporate a greater degree of learner choice of target than hitherto, and to move away from nativelike targets for learners whose goal is international intelligibility.

Pronunciation materials are responding to the EIL research in other ways. The most noticeable phenomenon is the large increase in the number of NNSs used in listening activities, thus providing exposure to a range of L2 English accents. Things are moving rather less slowly in relation to production, though the first courses to offer learners the choice of an NS or a local (but internationally intelligible) NNS model are appearing (e.g., Cunningham & Moor, 2003; Sato, Kanechiku, Matsumoto, & Miyata, 2003). Otherwise, it is still a case of adapting existing published pronunciation materials. Jenkins (2003), for example, provides suggestions for adapting minimal pair activities from Brown (1997) and Hancock (1995) to promote production of respectively core sounds and tonic stress, and from Gilbert (2001) to develop EIL accommodation skills.

However, at this stage, the emphasis is more on raising awareness of EIL contextual factors in manuals and materials for teacher education than on providing classroom pronunciation courses. Pennington (1996), though described as “an international approach,” largely restricts the NNS element to description while providing production activities that promote NS norms. McKay's (2002) handbook for teachers, on the other hand, raises awareness of the possibility of teaching productive pronunciation for EIL by focusing on the Lingua Franca Core items, although she concludes, in line with Dalton and Seidlhofer (1994), that it may still be valuable to maintain an NS accent as a point of reference in the classroom, thus “preventing speakers of English from moving too far apart in their pronunciation” (p.72). Dalton and Seidlhofer (1994) is itself a handbook for teachers and an early example of a teacher education book taking EIL into consideration in its concern with identity and intelligibility. More recently, Walker (2001) is the first attempt to provide teachers within a specific L2 context, in this case Spain, with a taxonomy of core features that their learners should focus on for EIL purposes. And still more recently, Hung (2002b) has addressed the issue of dictionary transcripts, arguing that these should reflect local (in his case Hong Kong) pronunciations of English rather than elite British or American accents. Although this shift has begun to take place with respect to the indigenized varieties of English of the outer circle, however, more data will be needed before the same approach can be applied systematically to expanding circle Englishes. Finally, Keys and Walker (2002) address the inevitable concern of teachers that a move away from exonormative British or American accent norms and models will be accompanied by a decline in pronunciation standards.

However, there are also sociopsychological factors to be taken into account, particularly those relating to language attitudes, motivation, and identity. Dörnyei and Csizér argue on the basis of empirical evidence that traditional orientations to motivation are being challenged by current developments: “World English is turning into an increasingly international language and it is therefore rapidly losing its national cultural base while becoming associated with a global culture,” which, they contend, “undermines the traditional definition of integrativeness as it is not clear any more who the ‘L2 speakers’ or the members of the ‘L2 community’ are” (2002, p. 453). In this respect, the sociopsychological situation is not only unclear, but is also sending out contradictory signals. Bamgbose (1998) describes L2 attitudes to English accents as “a love–hate relationship” and goes on to claim that in the outer circle “one does not wish to sound like a native speaker, but still finds the accent fascinating” (p. 7), but his claim would receive a very ambivalent response in the expanding circle. For, as the research of Dalton, Kaltenboeck, and Smit (1997), Grau (in press), Smit and Dalton-Puffer (2000), Timmis (2002) and others demonstrates, despite recent EIL developments, many teachers and learners still prefer to aim for an approximation of a nativelike rather than a local or internationally acceptable accent. This seems, paradoxically, to be the case even when, as Grau finds in her study, they simultaneously believe that the objective should be international intelligibility and that an L2 accent is acceptable.

Meanwhile, Smit (2002) finds that orientation to the target accent and L1 speaker group as well as self-efficacy and anxiety (that is, “how (in)adequate they feel in their pronunciation,” p. 95) play important roles in the acquisition (or not) of a near-native accent. She concludes that her study “supports the so often invoked character of pronunciation as being that aspect of a language which is closest to its speakers' feelings of identity” (p. 102). Her findings resonate in some respects with Lippi-Green's (1997) account of L2 accents in an inner circle context, the United States. Accent is seen both to arouse in L2 speakers feelings of linguistic insecurity, and to relate in critical ways to social identity and the construction of self and other. This takes us back to the problem highlighted by Dörnyei and Csizér: the difficulty of establishing the social identity of the L2 community in an international context and the implications for EIL pronunciation. While teachers and learners are becoming aware to some extent of the complex sociopsychological issues involved, there has as yet been no attempt to address them at the wider level through pronunciation teaching methodologies and materials, although more enlightened teacher education courses are beginning to grapple with them.

This is, nevertheless, still early in the day for EIL phonology, and it is likely to be some time before large numbers of teachers elect to offer their students a selection of context-based pronunciation goals—or for students to wish to take advantage of the offer. The EIL perspective is also beset by misinterpretations and misconceptions, particularly by those from regions where there is a strong tradition of educational investment in and attachment to the RP accent, such as Central Europe. For a typical example, see Sobkowiak (2003), who fails to grasp the essential difference between EIL and EFL and the implications for pronunciation norms and goals. The best, then, that may be said about pronunciation in EIL contexts at present is that those who support an EIL approach to pronunciation teaching alongside an ESL/EFL approach can be cautiously optimistic. Some learners are at least beginning to be offered a small element of choice in their pronunciation goals, and with further researching into and refining of the Lingua Franca Core and greater publicizing of the sociolinguistic, sociopsychological, and intelligibility imperatives, the process is likely to gather momentum in pronunciation as in the other linguistic levels (Seidlhofer, 2001; Seidlhofer, this volume).