1. Introduction
The issue of how to teach or learn a new language has generated an immense literature in English, based upon varying mixtures of assertion, theory, observation and experiment, and written from a variety of perspectives: psycholinguistic, sociolinguistic, pedagogic, educational and political. Since the late nineteenth century, the usual assumption in this literature has been that a new language is best taught and learned monolingually, without use of the students’ own language for explanation, translation, testing, classroom management or general communication between teacher and student. The belief – sometimes explicit, but more often implicit – has been that everything that happens during a language class should be in the language being taught, and that students should be discouraged or even banned from making any use of the language(s) they already know. Since this notion became accepted wisdom in the late nineteenth century, it has been largely taken for granted in the language teaching literature throughout the twentieth century, with only isolated voices of dissent. More recently, however, this monolingual assumption has been increasingly challenged alongside a reassessment of the merits of relating the language being taught to students’ own languages. This article surveys and assesses this new and growing literature.
Bearing in mind, however, that language learning is an international activity, it is important to note that what is in vogue in the literature does not necessarily reflect what is actually happening in all parts of the world. In some places, the latest fashion simply does not reach teachers, syllabus designers or text-book writers; in others, there is a significant time lag before a new approach arrives; and elsewhere, new theories may be actively resisted. Thus, although the mainstream literature has supported monolingual teaching for the last hundred years or so, there are many educational contexts where use of the students’ own language has remained the norm. Similarly, if bilingual teaching returns, as the literature surveyed in this article suggests it will, then it is likely that monolingual teaching will nevertheless continue in many places.
The greater part of the language-teaching literature has concerned the teaching of English, although a great deal of it, acting on the assumption underpinning much second language acquisition (SLA) research that acquiring a new language is a substantially uniform universal process, implicitly claims relevance to the teaching of any language. In our view, this failure to differentiate is a considerable weakness; while the psycholinguistic aspects of language learning may indeed have some universal features, the sociolinguistic factors vary considerably. As English continues to grow and consolidate its position as the dominant international language, its use raises very specific issues of power and identity, which, though shared to a certain extent by other dominant and widely distributed languages (such as French, Spanish, Portuguese and Russian), are very different from those involved in the teaching of smaller and more geographically confined languages (such as Icelandic, Chechen, Navaho and literally thousands of others). The extent and speed of the spread of English make it, in the view of some analysts (following Phillipson Reference Phillipson1992), a threat to other languages, including those which have formerly had international currency in one field or another (e.g. diplomatic French or scientific Russian). In addition, the global distribution of English (Crystal Reference Crystal2003), its fragmenting into new Englishes (Seargeant Reference Seargeant2012) and its use as a lingua franca (Seidlhofer Reference Seidlhofer2011) mean that, for many learners, native-speaker models of English and the goal of cultural integration into English speaking countries are no longer needed, or even desirable.
This confounding of different languages is also sometimes present in the literature advocating a return to bilingual teaching. That is to say, many arguments for bilingual teaching are vague as to whether they are relevant to any language, to a particular language, or specifically to English. It may be, however, that a reintroduction of bilingual teaching will also hasten an end to over-generalised statements about language learning. The learning of Chinese by a French speaker, for example, which necessitates the learning of a new writing system and significantly different structures, is not the same as the learning of Italian by a Spanish speaker. Thus, specificity is needed, and we should make our position and the scope of this survey clear from the outset. This article is concerned with the teaching of English to speakers of other languages, though where relevant, we also review literature referring specifically to languages other than English (almost exclusively other ‘big’ languages such as French and Spanish). We do not, however, claim relevance to the teaching of any language.
Moreover, in the teaching of English, the dominance of the monolingual assumption is particularly important for a number of reasons. Firstly, it has had a devastating effect on the status of non-native English speaker teachers. Additionally, by assuming that the goal of learners should be to emulate native-speaker proficiency, it has presented learners with an unattainable but not necessarily desirable ideal (not all native-speaker English is widely comprehensible, stylistically diverse, literate or eloquent). Furthermore, the monolingual teaching of English has inhibited the development of bilingual and bicultural identities and skills that are actively needed by most learners, both within the English-speaking countries and in the world at large.
1.1 Survey outline
After a brief and necessary discussion of terminology (section 1.2), the rest of section 1 outlines the late nineteenth-century origins of monolingual language teaching which, despite changes in theories of language and of learning, was supported by the vast majority of language teaching and learning literature throughout the twentieth century. Section 2, in contrast, charts the continuing widespread use of students’ own languages, even in contexts where the monolingual assumption is notionally unchallenged. Subsequently, section 3 examines the contemporary scholarly and societal trends which have led to a revival of interest in own-language use and translation. Key arguments underpinning this revival are brought together in section 4. Section 5 reviews empirical studies which explore both the extent of own-language use (5.1) and its functions (5.2) within the language classroom. Section 6 examines the theoretical frameworks and research findings which support bilingual teaching, including psycholinguistic and cognitive perspectives (6.1), SLA (6.2), general learning theory (6.3) and sociocultural approaches (6.4), from there moving to questions surrounding the ‘optimal amount’ of own-language use in the classroom (6.5) and ways of integrating psycholinguistic into sociolinguistic frameworks (6.6). Research into teachers’ and learners’ attitudes towards own-language use is discussed in section 7. Finally, this review considers the persistence of monolingual approaches, especially in content-based learning (section 8), before examining how a bilingual approach may be implemented in practice in section 9's concluding discussion.
1.2 Terminology
As so often in academic debate, the terminology which surrounds own-language use in language teaching can cause considerable confusion; different terms are often used by different authors to refer to the same concept, while the same terms are also used in different senses. Clarity is therefore needed in referring to the two opposed approaches to teaching. We shall refer to the notion that a language is best taught without reference to another language as monolingual teaching and the opposite notion, that use should be made of a language the student already knows, as bilingual teaching.
All learners of an additional language are by definition speakers of at least one language and aspire to know at least one new one. At least two languages are therefore involved, and there are established terms to refer to each. Most widespread are ‘first language’ (abbreviated to ‘L1’), ‘mother tongue’ and ‘native language’ for the existing language and ‘second language’ (abbreviated to ‘L2’), ‘foreign language’ or ‘target language’ for the new one.
In our view all these terms are unsatisfactory. In many educational settings, the common shared language is not the ‘first’ or ‘native’ language of all students (for example, although German is the language used in German secondary schools, and therefore the language likely to be recruited as an aid to the teaching of English, it is not the first language of all the pupils in those schools, who may for example be recent arrivals from Turkey or Poland). Moreover, the term ‘native language’ has long been effectively challenged as muddled and imprecise (Coulmas 1981; Rampton Reference Rampton1990). ‘Mother tongue’ is not only an emotive term but also inaccurate – for the obvious reason that many people's ‘mother tongue’ is not their mother's ‘mother tongue’!
As for the language being learned, ‘second’ wrongly implies that all learners know only one other language, when many are already bi- or multi-lingual; in addition, it arguably has unfortunate connotations of ‘second class’. Furthermore, as the use of English in the world becomes ever more global (Graddol Reference Graddol1997, Reference Graddol2006; Crystal Reference Crystal2003; Seargeant Reference Seargeant2012), the distinction between English as ‘second’ and ‘foreign’ language becomes ever harder to draw; in some senses English has become the second language of the whole world. Meanwhile, ‘target language’ may have unfortunate military overtones. For these reasons, this article, while acknowledging the currency of the terms above, will adopt the terms own language – ‘the language which the students already know and through which (if allowed), they can approach the new language’ (G. Cook Reference Cook2010: xxii) – and new language, that is, the language being learned (for further discussion, see G. Cook Reference Cook2010: xxi–xxii).
1.3 Origins of monolingual teaching
The twentieth century was a time of successive revolutions in theories of language and language acquisition, and these have had a marked influence on language teaching. For example, philology was overtaken by structuralist linguistics; behaviourist learning theory retreated in the face of a new Chomskyan nativism; a focus on sentence grammar yielded to functionalist and discourse analytic views; corpus linguistics re-wrote ideas about the relation between vocabulary and grammar. The fact that confidence in monolingual teaching has survived such changes suggests that its origins lie less in the theoretical than the practical sphere. This view is also borne out by the history of its emergence in the late nineteenth century.
Monolingual teaching was not unknown before that time, however. In Europe, there had been medieval schools which used Latin for all purposes, immersion teaching by governesses and tutors in upper-class families, and instances of monolingual teaching in general secondary school education (Phillipson Reference Phillipson1992: 186−187). But it was from the late nineteenth century onwards that it was extensively promoted in the language-teaching literature, rapidly gaining the status of an unchallenged assumption, not only in the teaching of English, but other major European languages too. Although the academics of the self-styled reform movement (Howatt & Smith Reference Howatt and Smith2002, Howatt with Widdowson 2004: 187–210) drew upon research in phonetics and psychology to vigorously oppose the grammar translation method which dominated foreign language teaching in secondary schools at that time (instead advocating a greater emphasis on spoken language, fluency and connected texts), they were by no means dogmatically opposed to the use of students’ own languages. Henry Sweet, for example, one of the leaders of the Reform Movement in Britain, explicitly advocated the use of translation in the teaching of vocabulary (1899/1964: 194).
Thus, a more credible source for dogmatically monolingual teaching is the work of Maximilian Berlitz. The story is that Berlitz, a German Jewish immigrant to the USA who himself spoke several languages, employed a Frenchman, Nicholas Joly, to cover his French class for English-speaking students (Berlitz International 2008: 4). Having always communicated with Joly in French, however, Berlitz did not realise that his new employee did not speak any English. Joly was therefore not able to follow the usual practice of explaining the French language in English, and had to teach French in French, making his meaning as clear as he could through gestures, facial expressions, pictures and so forth. To Berlitz's surprise, on his return (so the story goes) not only had the pupils enjoyed these lessons, they also seemed to have made progress. This serendipitous event apparently inspired Berlitz to develop the berlitz method, which still survives nearly a century and a half later, proudly described on current Berlitz websites (e.g. Berlitz London: 2011). One of the main tenets of the method remains as follows:
The Berlitz Method excludes any use of the student's native language in either the classroom or in the student's review materials. By totally immersing the student in the new language, we can most closely simulate the real-life situations in which he or she will be using the language, and eliminate the cumbersome process of introducing a concept first in the student's language and then in the target language.
With this method, Berlitz went on to found the Berlitz schools, which spread rapidly throughout the USA and internationally, teaching both English and other languages. There were 200 such schools by 1914, and the organisation is still thriving today, promoting the fact that it ‘is one of the few international organisations to survive two World Wars and the Great Depression’ (Berlitz International 2008: 8).
The monolingual principle of the Berlitz schools became a model for other institutions, and its widespread acceptance can be attributed to a number of factors. Classes in which students are speakers of a variety of languages, and the employment of native speaker teachers who do not necessarily know the language(s) of their students, created situations in which bilingual teaching seemed to be impossible. It was also in the interests of both publishers and language schools based in English-speaking nations to promote monolingual products which could be implemented by native-speaker experts, marketed worldwide without variation, and did not need input from speakers of other languages.
Additionally, the perceived goals of language learning changed. In so-called ‘traditional’ language teaching, and in particular in Grammar Translation, the aim was to develop the ability to translate written language accurately, and ultimately to build upon this skill, thereby enabling literary texts to be read in the original. Since the move towards monolingual teaching in the late nineteenth century, however, a frequent if often unstated assumption has been that, with the exception of courses specifically aimed at training translators and interpreters, the goal of language teaching is to prepare students to communicate in monolingual environments and to emulate as far as possible the use of the new language by its native speakers – a goal which for many learners is neither useful, desirable or attainable (Davies Reference Davies, Cook and Seidlhofer1995, Reference Davies2003; see also section 1). There was little or no acknowledgement of the need of many learners to operate in bilingual or multilingual environments, where translation and appropriate code-switching are needed and valued skills (Sridhar & Sridhar Reference Sridhar and Sridhar1986).
Similarly, the recent reassessment of the merits of bilingual teaching seems also to be driven more by practical considerations and a reconsideration of the goals of language learning rather than by any new theories of language and language learning. Indeed, the proponents of a return to bilingual teaching seem to share many theoretical assumptions with those advocating monolingual teaching in the late twentieth century, including their focus on communication and learner needs. Major factors in the advocacy of bilingual teaching include a recognition that many learners will need to operate bilingually, that they will wish to preserve their own cultural and linguistic identity while speaking English, and that they will not necessarily be using English in a native-speaker environment, but as a lingua franca with other non-native speakers of English (Jenkins Reference Jenkins2000, Reference Jenkins2007; Seidlhofer Reference Seidlhofer2002, Reference Seidlhofer2011). In short, the increasingly accepted view, following a paradigm-breaking argument by Widdowson (Reference Widdowson1994), is that the predominantly English-speaking countries, given the language's international currency, can no longer claim ‘the ownership of English’.
Although belief in the superiority of monolingual teaching was endemic in the most influential English language teaching literature from the late nineteenth century onwards, there were nevertheless important changes within this monolingual paradigm. While actual use of students’ own languages by teachers or students was discouraged or even banned, this did not necessarily initially entail a lack of interest in that own language by teachers and course designers. In the 1940s and 1950s, it was widely assumed that contrastive analysis (a comparison of the learners’ own language and the new language) would be useful, as it would identify points of difficulty for speakers of particular languages (Fries Reference Fries1945; Lado Reference Lado1957). Knowledge of learners’ own language and its relation to the new language was therefore present in the background of monolingual teaching, even when not allowed in the foreground.
However, in the 1970s and 80s, with the emergence of SLA as the dominant body of academic theory informing language teaching, even this background consideration of learners’ own languages was removed. Early SLA claims that a natural order of acquisition would be followed by all learners of English whatever their own language (Krashen Reference Krashen1982), and that first language interference was only a minor cause of error (Selinker Reference Selinker1972), dealt a formidable blow to the apparently commonsense view that syllabuses should vary with the first language of the students (whereby, for example, the teaching of English to Chinese speakers would demand a different approach from the teaching of English to French speakers). Additionally, the emphasis on a natural and universal order of second language acquisition through exposure to meaning did not favour the conscious attention to differences between new language and own language which is inherent in bilingual teaching. There are, however, some notable exceptions to the dismissal of contrastive analysis as a guide to learner problems, such as Swan & Smith's edited collection (1987/2001), which brings together insights into likely learner difficulties across a range of own languages.
2. Acknowledging own-language use
However, despite the monolingual imperatives which dominated theories of language teaching and SLA research until the late twentieth century, learners’ own languages and, indeed, translation, have continued to be used in many language classrooms around the world (Benson Reference Benson2000; V. Cook Reference Cook and Beaven2008), especially in contexts where learners share a language which is also known by the teacher. For example, Adamson (Reference Adamson, Davies and Elder2004) notes that Grammar Translation in English language teaching prospered until the 1960s in China, whilst Nasrin (Reference Nasrin2005) outlines the continuing use of the same method in Bangladesh.
Yet, as Pennycook (Reference Pennycook and Byram2004) points out, there is more to own-language use than the ‘static’ and ‘traditional’ impression that the term ‘Grammar Translation’ portrays, an image which has served to stereotype and marginalise non-monolingual teaching practices around the world (see also G. Cook Reference Cook2010). Numerous studies have documented bilingual teaching, code-switching and code choice (emphasising learner choice during classroom interaction and teacher choices in curriculum design and teaching practice (Levine Reference Levine2011)) in a range of English language classrooms around the world, for example in Botswana (Arthur Reference Arthur1996); Brazil (Fabrício & Santos Reference Fabrício, Santos and Edge2006); China (Qian, Tian & Wang Reference Qian, Tian and Wang2009; van der Meij & Zhao Reference van der Meij and Zhao2010; Littlewood & Yu Reference Littlewood and Yu2011); Cyprus (Copland & Neokleous Reference Copland and Neokleous2011); Finland (Nikula Reference Nikula2007); Germany (Butzkamm Reference Butzkamm1998); Hong Kong (Pennington Reference Pennington1995; Lin Reference Lin1996; Carless Reference Carless2002, Reference Carless2004, Reference Carless2008; Littlewood & Yu Reference Littlewood and Yu2011); Hungary (Harbord Reference Harbord1992; Nagy & Robertson Reference Nagy, Robertson, Turnbull and Dailey-O'Cain2009); Italy (Moore Reference Moore2002); Japan (Hobbs, Matsuo & Payne Reference Hobbs, Matsuo and Payne2010); Malta (Camilleri Reference Camilleri1996); South Africa (Adendorff Reference Adendorff, Bailey and Nunan1996); South Korea (Liu et al. Reference Liu, Ahn, Baek and Han2004; Kang Reference Kang2008); Spain (Unamuno Reference Unamuno2008); Sri Lanka (Canagarajah Reference Canagarajah1999); Sweden (Cromdal Reference Cromdal2005); Thailand (Forman Reference Forman2007, Reference Forman2008); and Turkey (Eldridge Reference Eldridge1996; Ustunel & Seedhouse Reference Ustunel and Seedhouse2005). Studies from Canada, with its particular history of bilingual education, also document own-language use and code choices (e.g. Behan & Turnbull with Spek Reference Behan, Turnbull and Spek1997; Swain & Lapkin Reference Swain and Lapkin2000, Reference Swain and Lapkin2005; Liebscher & Dailey-O'Cain Reference Liebscher and Dailey-O'Cain2005; Cummins Reference Cummins2007; Dailey-O'Cain & Liebscher Reference Dailey-O'Cain, Liebscher, Turnbull and Dailey-O'Cain2009).
Similarly, own-language use in US language classrooms (both English and other languages) has been documented by, for example, Brooks & Donato (Reference Brooks and Donato1994); Polio & Duff (Reference Polio and Duff1994); Blyth (Reference Blyth and Kramsch1995); Kramsch (Reference Kramsch, Byram and Fleming1998); Antón & DiCamilla (Reference Antón and DiCamilla1999); Alley (Reference Alley2005); Belz (Reference Belz2002); Levine (Reference Levine2003, Reference Levine, Turnbull and Dailey-O'Cain2009, Reference Levine2011); Edstrom (Reference Edstrom2006); Scott & de la Fuente (Reference Scott and de la Fuente2008); and Brooks-Lewis (Reference Brooks-Lewis2009). Meanwhile, Rolin-Ianziti & Brownlie (Reference Rolin-Ianziti and Brownlie2002), Crawford (Reference Crawford2004), Kim & Elder (Reference Kim and Elder2008) and Ellwood (Reference Ellwood2008) document own-language use in Australian and New Zealand language classrooms, whilst Mitchell (Reference Mitchell1988), Macaro (Reference Macaro1997, Reference Macaro2001), Allford (Reference Allford and Pachler1999) and Meiring & Norman (Reference Meiring and Norman2002) detail the use of learners’ own languages in British secondary school foreign language classes. Moreover, the continued use of translation activities in language classrooms around the world has also been increasingly acknowledged in recent years. For example, Kern (Reference Kern1994) investigates the role of translation in L2 reading; Lally (Reference Lally2000) and Kim (Reference Kim2011) outline the use and effects of translation tasks within writing activities; Tudor identifies a use for translation within ESP (1987a; 1987b); and Malmkjær (1998) points out that translation and own-language use remain the norm in university-level language teaching.
According to V. Cook (Reference Cook2001), language teachers who are able to do so use the learners’ own language ‘every day’. Similarly, Lucas & Katz's (Reference Lucas and Katz1994) study of English-only programmes in the US reports that ‘the use of native language is so compelling that it emerges even when policies and assumptions mitigate against it’ (p. 558). As Levine (Reference Levine2011) observes, therefore, the language classroom has remained a multilingual environment in spite of the monolingual principles and norms which have been assumed within the language teaching literature over the last hundred years. Despite its disappearance from the public discourses of language teaching and learning, in many contexts, own-language use and translation has never entirely ceased – or been ‘stamped out’ (Butzkamm Reference Butzkamm2003: 29).
What has developed in language teaching, therefore, is a divide between those who have regarded monolingual classrooms as a given and those who teach using the learners’ own language. Yet twelve years into the twenty-first century, there is evidence that this division, which, in many contexts, might arguably be characterised as one between theory and practice, may be coming to an end, and that the existence and advantages of using the learners’ own language in class are increasingly recognised. Next, therefore, we shall briefly examine how societal and scholarly trends have led to an environment in which own-language use might be viewed more positively.
3. The changing context
The recent interest in own-language use and, to a lesser extent, translation has been made possible by changes in the academic and political climate surrounding language teaching and learning. Early SLA ideas about natural acquisition through attention to meaning have now been effectively discredited (Gregg Reference Gregg1984, McLaughlin Reference McLaughlin1987, Widdowson Reference Widdowson and Widdowson1990inter alia), and early SLA research has been widely criticised for its asocial and apolitical approach to language learning (Block Reference Block2003). The development of a social turn (ibid.) within applied linguistics more easily acknowledges complexity, diversity, difference and uncertainty within language and language learning, and has complemented a growing interest in the relevance of complexity theory to new language development (e.g. Larsen-Freeman Reference Larsen-Freeman1997; Larsen-Freeman & Cameron Reference Larsen-Freeman and Cameron2008: 115–160); of sociocultural theory (e.g. Lantolf Reference Lantolf2000, Reference Lantolf and Atkinson2011); and of ecological approaches (e.g. van Lier Reference van Lier and Lantolf2000, Reference van Lier2004; Kramsch Reference Kramsch2002) as ways of understanding and explaining language and language-related phenomena.
At the same time, the effects of contemporary migration and globalisation alongside the increasing recognition that non-native speakers of English have long outnumbered native speakers globally (Crystal Reference Crystal2003) have led to a re-evaluation of the importance of bilingualism and multilingualism for individual and societal language use. As central and creative elements of multilingual discourse which create and maintain speaker identity (Norton Reference Norton2000; Rampton Reference Rampton2005; Block Reference Block2007; De Fina Reference De Fina2007), code choice and code-switching have become increasingly de-stigmatised beyond the classroom and, consequently, are also starting to be seen as a ‘normal behaviour’ (Levine Reference Levine, Turnbull and Dailey-O'Cain2009) within language classrooms. From this perspective, therefore, language learners are increasingly seen as multiple language users (Belz Reference Belz2002), with the language classroom conceptualised as a multilingual speech community (Blyth Reference Blyth and Kramsch1995; Edstrom Reference Edstrom2006). This contrasts with what V. Cook (Reference Cook2001) characterises as a ‘traditional’ view of learners as deficient ‘imitation natives’ learning in monolingual classrooms.
Such considerations highlight the need to consider the links between code-switching, speaker identity and the symbolic value of languages; a number of studies have addressed these issues in postcolonial settings. Focusing on English-medium primary classrooms in Botswana, Arthur (Reference Arthur1996) suggests that code-switching from English into the learners' own language (Setswana) creates a ‘safe-space’ in which learners can contribute more to a lesson and engage more critically with the curriculum. With only limited access to English beyond the classroom, learners and teachers employ Setswana as a ‘language of complicity’ to overcome problems of English-medium classroom interaction. Lin (Reference Lin1996) takes an equally critical approach to the symbolic domination of English in Hong Kong schools, also suggesting that own-language use is a pragmatic response in English-medium classrooms and calling for a ‘balanced academic bilingualism’ (p. 79), both to reflect the reality of classroom life and to challenge the subordination of ‘all cultural and educational goals to the single dominant goal of learning English’ (ibid.). Similarly, Katunich (Reference Katunich and Edge2006) argues that English-only or English-mainly teaching in post-apartheid South Africa is a form of colonialism which denies black (Katunich's terminology) language learners the possibility of additive multilingualism and limits their identities and educational trajectories – see also Chick & McKay (Reference Chick and McKay2001). For further discussion of the functions of classroom code-switching in other postcolonial contexts in Africa, see, for example, Ferguson (Reference Ferguson2003) and Opoku-Amankwa (Reference Opoku-Amankwa2009). Overall, therefore, these perspectives highlight the ways in which debates surrounding own-language use and code-switching in the English language teaching (ELT) classroom are not ‘just’ technical issues surrounding how languages are learned, but can underpin learners’ sense of who they are and who they want to be in a complex multilingual world.
Although particularly relevant within postcolonial environments, concerns surrounding identity and cultural integrity are by no means limited to these contexts. Numerous studies have focused upon the links between identity, own-language use and the development of intercultural communicative competence, that is, the ability of speakers to communicate in culturally appropriate ways as they move between languages and cultural groups (Stiefel Reference Stiefel, Witte, Harden and Ramos de Oliveira Harden2009; Levine Reference Levine2011). Crawford (Reference Crawford2004) finds that teachers see the learners’ own language as the most appropriate medium for cross-cultural comparisons, whilst Elorza (Reference Elorza2008), Stiefel (Reference Stiefel, Witte, Harden and Ramos de Oliveira Harden2009) and House (Reference House2009), amongst many others, suggest that, in addition to a focus on linguistic accuracy, in-class translation activities can also bring to learners’ attention cross-cultural differences in the ways speakers communicate. Meanwhile, from a more critical perspective, Fabrício & Santos (Reference Fabrício, Santos and Edge2006) outline the ways in which a group of Brazilian schoolchildren use their own language to reflect on the relationship between English and Portuguese both in their classroom and beyond. Noting that English, the language of globalisation and the market, ‘is everywhere in Brazil’ (p. 68), their study illustrates how own-language use facilitates learners’ awareness of the symbolic competence of languages (Kramsch Reference Kramsch2006), whereby the decision to use one language over another at a particular time is in part influenced by the history and power relationships between those languages, that is, by their ‘symbolic value’ (see also Kramsch Reference Kramsch2009).
Thus, changing academic and contemporary political, societal and personal perspectives on bi- and multilingualism provide a context in which interest in own-language use in language classrooms has developed and the reality of in-class code-switching is more readily acknowledged. However, research into, and arguments for, own-language use draw upon a range of differing perspectives, from theories of cognition and learning to concepts of power and classroom management, and from the search for ‘optimal’ own-language use to the role of teacher and learner beliefs in supporting (or rejecting) bilingual teaching. It is to these arguments that we now turn in more detail, looking first at broad theoretical reconceptualisations of language learning and teaching which acknowledge and favour own-language use in the classroom, before moving on to examine specific evidence and insights provided by empirical studies.
4. Own-language use: (re-)emerging arguments
4.1 The ‘crosslingual’ language classroom and ‘bilingualisation’
In his balanced consideration of crosslingual and intralingual language teaching (i.e. teaching which utilises or makes reference to the learners’ own language versus teaching that uses only the new language), Stern (Reference Stern1992: 279–299) suggests that these two approaches are not ‘opposites’ but, rather, form a continuum whereby learners’ own languages will be used in different ways and to differing extents at various stages during instruction. Arguing from both a practical perspective (for example, purely intralingual teaching is unrealistic; searching for an interpretation or translation when we do not understand the L2 – to use Stern's terminology – is ‘natural language behaviour’) and drawing upon more theoretical perspectives (for example, the own/new language connection in a learner's mind is an ‘indisputable fact of life’ which generally leads to the use of the learner's own language as a reference system for the new language), Stern suggests that crosslingual and intralingual techniques and practices can complement each other. For Stern, intralingual teaching strategies will be more effective if crosslingual activities such as translation, the use of bilingual dictionaries, ‘consecutive and simultaneous interpreting’ and the ‘interpretive treatment of texts’ are more clearly recognised as strategies ‘in their own right’ (p. 295). Meanwhile, the most appropriate balance of intra- and crosslingual approaches will depend on the specific purpose and context of learning: if communicative proficiency in the new language is the principal goal of teaching and learning, intralingual strategies will dominate; however, if mediating skills such as translation and interpreting are the goal of learning, crosslingual strategies should be an important part of classroom life (p. 301).
Like most advocates of own-language use (as we shall see), Stern is careful to emphasise that his support for crosslingual teaching is not a call to abandon intralingual activities or to ignore ‘the crucial role of monolingual communicative activities and tasks’ in the language classroom (Butzkamm & Caldwell Reference Butzkamm and Caldwell2009: 14); he also points out that further research is necessary into the most appropriate combination of intra- and crosslingual techniques for different language learning contexts, age groups and abilities (echoed by Macaro Reference Macaro1997, Reference Macaro and Llurda2006). But his case for the abandonment of anti-own-language dogma is clear: ‘crosslingual techniques have a theoretical justification and can also be efficient, helpful to the learner and interesting’ (Stern Reference Stern1992: 289).
Stern's discussion cites Widdowson's consistent concern with the unthinking abandonment of translation and own-language use in language teaching and learning. At the height of the ‘communicative revolution’, Widdowson, in his Teaching language as communication, the seminal text for communicative language teaching (CLT), observed that:
What we are aiming to do is make the learner conceive of the foreign language in the same way as he [sic] conceives of his own language and to use it in the same way as a communicative activity. This being so, it would seem reasonable to draw upon the learner's knowledge of how his own language is used to communicate. That is to say, it would seem reasonable to make use of translation. (1978: 159)
Widdowson develops this argument in much greater detail in his more recent discussion of ‘bilingualisation’, that is, the process of acquiring a new language, in ELT (2003: 149–164). Widdowson identifies the essentially bilingual nature of the foreign language classroom whereby ‘our students come to class with one language (at least) and our task is to get them to acquire another one’ (p. 149), and argues that conventional (i.e. monolingual) language teaching procedures fail to recognise the ways in which all bilingual language users fuse their knowledge of two languages into a single system of compound bilingualism (Weinreich Reference Weinreich1953). Compound bilingualism implies that two (or more) languages ‘are interwoven in the L2 user's mind’ (V. Cook Reference Cook2001) in terms of, for example, vocabulary, syntax, phonology and pragmatics. Consequently, learning a second language cannot be the same as first language acquisition, for the learners’ own language plays a central role in the development and use of their new language. For both V. Cook and Widdowson, monolingual teaching not only overlooks the ways in which learners will always draw upon their own language in order to learn a new language, but is paradoxically designed to stifle natural second language learning processes (Widdowson Reference Widdowson2003: 150).
Thus, V. Cook (Reference Cook2001) criticises the notion that successful language acquisition depends on the separation of languages in the learner's mind (i.e. coordinate bilingualism) which, he suggests, is a central tenet of monolingual teaching. Both V. Cook (ibid.) and Widdowson (Reference Widdowson2003) trace the attempt to compartmentalise and separate the learner's own language and the new language back to theories of transfer such as Contrastive Analysis (see section 1.3) in which the learners’ own language was seen as the major source of difficulty and interference in new language learning. However, they both also ask why, when cognitivist perspectives on SLA now accept that the learners’ own language is a resource which learners actively draw upon in interlanguage development (Ellis Reference Ellis1994, cited in Widdowson Reference Widdowson2003: 152) – although exactly how remains unclear – the reconceptualisation of own-language use in language pedagogy remains ‘conspicuous by its absence’ (ibid.).
Both V. Cook's and Widdowson's critiques of monolingual teaching are at the forefront of attempts to re-evaluate the use of the first language in the classroom. Indeed, V. Cook (Reference Cook2001) follows up his broadly psycholinguistic arguments by suggesting a range of ways in which learners’ own language might be used positively in class, including conveying meaning and explaining grammar, organising classroom activities, maintaining discipline, building rapport and forming relationships between teacher and learners, and use of the learners’ own language for testing. Cook also suggests the development of learning activities which build up connections between own and new languages in the learners’ minds and the deliberate use of the learners’ own language during classroom tasks and activities. Throughout, he emphasises the importance of seeing such techniques positively rather than as fall-back activities which teachers feel guilty about. Thus, for V. Cook, as for Widdowson and Stern, the need to maximise new language in the classroom does not preclude the use of the learners’ own language, as it may provide efficient shortcuts within the learning process, be more related to the learning processes the learners are using, or be more relevant to their external learning goals (2001, 2002).
4.2 ‘Bilingual Reform’
Whilst Widdowson and V. Cook offer a broad theoretical re-evaluation of the use of learners’ own languages in the classroom, Butzkamm & Caldwell (Reference Butzkamm and Caldwell2009) provide a more detailed overview of how a bilingual approach to language teaching might operate in practice in their book The bilingual reform: A paradigm shift in foreign language teaching. Drawing upon both authors’ longstanding interest in this area (for example, Butzkamm Reference Butzkamm and Weinstock1989/2002; Caldwell Reference Caldwell1990) which, in turn, builds upon Dodson's earlier work and ‘Bilingual Method’ (1967/1972), the book acts both as a call for change in which the ‘mother tongue taboo will be swept away. . .[in an] act of theoretical house-cleaning’ (2009: 13) and as a practical guide for using the learners’ own language in the classroom. Indeed, building on Butzkamm's 2003 paper, they offer ten maxims for using the ‘mother tongue’ (their term), highlighting, for example, issues of learner confidence, a focus on meaning, and links between the learners’ own and the new language.
Thus, Butzkamm & Caldwell (Reference Butzkamm and Caldwell2009: 13) argue that own language is ‘the greatest pedagogical resource’ that a learner brings to foreign language learning, as it ‘lays the foundations for all other languages we might want to learn’. They therefore advocate not just a ‘flexible and less rigid attitude’ towards own-language use, but the systematic exploitation of the potential of learners’ own language(s) ‘where that is appropriate’ (original emphasis). Like Stern, V. Cook and Widdowson, therefore, Butzkamm & Caldwell emphasise the importance of meaning and communicative tasks in teaching and learning, and the need to create a ‘foreign language atmosphere’ by using the foreign language to perform the ‘normal business’ of the classroom, such as spontaneous communication between teachers and learners and classroom management tasks (thereby ‘cocooning’ learners in a ‘language bath’ (p. 31)). But their call is for a ‘new balance’ in which own-language use compensates for the limited time and exposure to new language which learners experience as a ‘natural’ part of instructed language learning. They also highlight the need for learners both to ‘decode’ language in order to understand its ‘message’, and to ‘code-break’, processing language to understand the ‘rule’ and how meanings are encoded so that language patterns can subsequently be used creatively to produce new meanings (V. Cook's terms (e.g. Reference Cook and Beaven2008), and a distinction similarly made by Butzkamm's notion of doppelverstehen/‘dual comprehension’, that is, understanding both what is said and how it is said (1989/2002: 12)). For Butzkamm & Caldwell, therefore, this ‘generative. . .and combinatorial power of language’ (p. 14) is best facilitated by use of the learners’ own language, and they outline a comprehensive repertoire of classroom techniques which, they suggest, will facilitate the paradigm shift towards own-language use in language teaching and learning, including idiomatic translation, bilingual dictionary work, and the development of dialogues and drama. We shall return to these practical issues in section 9.
4.3 A case for translation
Thus far, this review has examined the broad approaches to own-language use that underpin the cautious revival of interest in bilingual teaching. However, whilst pointing out that own-language use is not the same as translation, but also suggesting that the rehabilitation of own-language use may in fact open a gateway for translation, G. Cook (Reference Cook and Baker1998/2009, Reference Cook2007, Reference Cook and Beaven2008, Reference Cook2010) moves the debate towards this more specific (and controversial) possibility, presenting a case for the revival of translation within language teaching and learning.
Arguing that translation cannot be treated separately to, or be compartmentalised from, other forms of own-language use in the classroom, G. Cook's Translation in language teaching (2010) draws upon language-learning research, pedagogical theory and practice, and educational philosophy to make a case for a major reassessment and reintroduction of translation into language teaching and learning. The argument is that in many contexts translation is a natural and effective means of language learning, develops an important skill, answers students’ needs and preferences, and protects students’ linguistic and cultural identity. Meanwhile, Witte, Harden & Ramos de Oliveira Harden's edited collection (2009) strongly advocates translation activities as a means through which learners’ language awareness, intercultural competence and understanding of conceptual metaphors and literary texts may be developed (in addition to the benefits provided to learners’ fluency and accuracy). Similarly, Malmkjær (1998: 8) points out that translation, an important ‘fifth skill’ for language learners to develop in the contemporary world, depends upon, and is inclusive of, the other four skills of reading, writing, listening and speaking. Thus learners who engage in translation activities will inevitably practise other language skills, leading Malmkjær to suggest that translation is not, in fact, radically different from other language skills.
However, as G. Cook (Reference Cook2010: 52) observes, the distrust of translation became so deeply entrenched from the late nineteenth century onwards that the apparently small move from supporting own-language use to advocating translation is for many a step too far. He cites, for example, V. Cook's reluctance to advocate a return to translation in his seminal 2001 paper because of, according to V. Cook himself, ‘its pejorative overtones’ (p. 421); similarly, Stern's balanced account of cross and intralingual language teaching (1992) allows for only new to own-language translation (and then, not ‘to excess’, p. 293). Thus, recognising the potential contentiousness of the issue, G. Cook's discussion refers only to a ‘climate for revival’ rather than identifying a ‘revival’ in translation itself.
5. Examining the evidence
Writing in 2001, Macaro commented that there was ‘very little research’ into own-language use in language teaching and learning (p. 532) – a dearth which his own work has gone a long way to rectify; Liu et al. subsequently noted the particular lack of research in ELT contexts (2004: 610). But as section 2 illustrates, there is now a substantial literature which acknowledges the existence and extent of own-language use in classrooms, much written since Macaro's observation. A substantial proportion of this ‘flurry’ of scholarly interest (Levine Reference Levine2011: 72) documents the quantity and/or assesses the functions of own-language use, and it is to these two issues that we now turn.
5.1 The amount of own-language use
A number of studies have attempted to quantify the amount of own and new language use in the classroom, focusing mainly on the extent to which teachers use each code and the reasons for this. Although some research has focused on the teaching of English (e.g. Kharma & Hajjaj Reference Kharma and Hajjaj1989; Liu et al. Reference Liu, Ahn, Baek and Han2004; Copland & Neokleous Reference Copland and Neokleous2011), the majority of studies have investigated foreign language teaching in English-speaking countries (e.g. Duff & Polio Reference Duff and Polio1990; Macaro Reference Macaro1997; Rolin-Ianziti & Brownlie Reference Rolin-Ianziti and Brownlie2002; Edstrom Reference Edstrom2006; Kim & Elder Reference Kim and Elder2008; Levine Reference Levine2011).
In an early investigation, Guthrie (Reference Guthrie, Lantolf and Labarca1984, cited in Levine Reference Levine2003, Reference Levine2011) examined the own/new language practices of six university-level French teachers working in the same institutional context, all of whom reported teaching communicatively. Guthrie's analysis of ten hours of recorded teaching demonstrated substantial variation in the amounts and purposes of own-language use by teachers. Yet it also suggested that most teachers used the new language most of the time. Subsequently, however, Duff & Polio (Reference Duff and Polio1990) undertook a more wide-ranging study, asking ‘How much foreign language is there in the foreign language classroom?’ and observing, recording and transcribing the spoken discourse of thirteen foreign language classrooms within a university languages programme (their investigation also utilised learner questionnaires and teacher interviews to elicit attitudes and motivations towards own-language use, to be explored in section 7). Duff & Polio found much more variation than Guthrie in own-language use by teachers. Whilst the mean and median amounts of new language use across the study were 67.9% and 79% respectively (on average, therefore, the learners’ own language was used by teachers for 32.1% (mean) and 21% (median) of their utterances), within individual classes, the range of new language use ranged from 10% to 100% (correspondingly, therefore, own-language use within the study ranged from 0% to 90%). Turnbull (Reference Turnbull2001) identifies similar levels of variation in the amount of new language spoken by school French teachers in Canada (from 9% to 89% of classroom activities conducted in French). Similarly, Edstrom (Reference Edstrom2006), focusing on her own practices as a teacher of Spanish in a US university over the course of a semester, found ‘extreme fluctuations’ in own-language use, ranging from 7% to 70% of her talk for individual lessons, or, taken as monthly average, from 17% during one month's classes to 42% of her speech in another (see also Turnbull Reference Turnbull2001 and Kim & Elder Reference Kim and Elder2008 for similar evidence of variation in and relatively high levels of own-language use). Cai (Reference Cai2011), in her study of English teaching in a Chinese university, documented classes in which up to 80% of teaching was in Chinese, despite teachers’ self reports of much lower quantities.
Furthermore, even those studies which show generally less own-language use by teachers reveal variation both between instructors within the same institution and/or between lessons. For example, Rolin-Ianziti & Brownlie (Reference Rolin-Ianziti and Brownlie2002) found that own-language use by four French teachers directing the same listening exercise ranged from 0% to 18.15% of their total speech, whilst Macaro (Reference Macaro2001) observed that six student teachers’ own-language use between lessons ranged from 0% to 15.2%. Rolin-Ianziti & Brownlie (Reference Rolin-Ianziti and Brownlie2002) also reported that a teacher who had not employed the learners’ own language at all during a listening exercise used it for 55.51% of her speech during a grammar activity. Meanwhile, observations of own-language use in two ELT contexts suggest that teachers used the learners’ own language for an average of 20% in Kuwaiti beginner and lower intermediate classes (Kharma & Hajjaj Reference Kharma and Hajjaj1989) and, in a South Korean High School, for up to 32% of class time (Liu et al. Reference Liu, Ahn, Baek and Han2004). Moreover, Copland & Neokleous (Reference Copland and Neokleous2011) observed a range of own-language use in a Cypriot language school, from lessons conducted almost wholly in English to classes handled almost wholly in Greek.
To summarise, the discussion so far provides a general impression of own-language use in the classroom. Two general approaches to the analysis of teachers’ speech can be identified: own-language use measured in terms of time spent in new or own language (e.g. Duff & Polio Reference Duff and Polio1990; Edstrom Reference Edstrom2006) or of the number of words spoken in each code (e.g. Rolin-Ianziti & Brownlie Reference Rolin-Ianziti and Brownlie2002; Liu et. al. Reference Liu, Ahn, Baek and Han2004), the latter needing to take into account the differences in word counts between synthetic languages with a high morpheme-per-word ratio and isolating languages with a low morpheme-per-word ratio. Nevertheless, the studies confirm the extent to which teachers’ own-language use varies between and within contexts. A complicating factor, however, is the tendency for teachers to underestimate the extent to which they use the learners’ own language. For example, according to Polio & Duff, who in addition to the recorded and quantified classroom data presented here also obtained teacher perceptions of own-language use, the instructors in their study were unaware of ‘how, when and the extent to which they actually used English’ and the degree to which they ‘urged students to speak the L2, but then would not necessarily do so themselves’ (1994: 320; see also Edstrom Reference Edstrom2006; van der Meij & Zhao Reference van der Meij and Zhao2010; Copland & Neokleous Reference Copland and Neokleous2011). This may be because of the underlying negative attitudes towards own-language use which still prevail in many contexts, and individual teachers’ attitudes and beliefs about bilingual teaching. We shall examine teachers’ beliefs in more detail in section 7. Now, however, we shall examine why teachers code-switch: what are the functions (both observed by researchers and reported by teachers and learners) of own-language use in the classroom?
5.2 The functions of own-language use
As Edstrom (Reference Edstrom2006) observes, whilst there is considerable variation in the reported quantities of own-language use by teachers, the reported functions of own-language use are relatively constant. When interviewed, for example, the teachers in Duff & Polio's study (Reference Duff and Polio1990) suggested that they were more likely to use the learners’ own language when it was particularly ‘different’ from the new language (in terms of, for example, writing systems or grammar); similarly, departmental policies, lesson objectives and tasks, and training and qualifications were said to affect the amount of own-language use.
Subsequently, Polio & Duff (Reference Polio and Duff1994) returned to their transcription data to establish in more detail the particular pedagogic functions of own-language use, finding that teachers used their learners’ own language for grammar instruction, classroom management and administration, to demonstrate empathy or show solidarity with the learners, to provide translations for unknown words and to compensate for learners’ apparent lack of understanding and in response to learners already speaking in their own language. Comparable rationales for own-language use are identified in a range of similar studies, such as Atkinson (Reference Atkinson1987), Franklin (Reference Franklin1990), Macaro (Reference Macaro1997, Reference Macaro2001), Rolin-Ianziti & Brownlie (Reference Rolin-Ianziti and Brownlie2002), Carless (Reference Carless2004), Liu et al. (Reference Liu, Ahn, Baek and Han2004), Copland & Meokleous (Reference Copland and Neokleous2011) and Levine (Reference Levine2011), as well as in the broader approaches of V. Cook (Reference Cook2001) and Butzkamm & Caldwell (Reference Butzkamm and Caldwell2009) outlined above (in section 4). Meanwhile, Ustunel & Seedhouse (Reference Ustunel and Seedhouse2005), asking ‘why that, in that language, right now?’, employ a Conversational Analysis methodology to trace in detail the relationship between teacher code-switching and pedagogical focus and sequencing in lessons. They find, for example, that teachers switch from the new to the learners’ own language after a pause or hesitation by learners, or switch code in order to prompt a similar switch by learners.
Meanwhile, Rolin-Ianziti & Varshney (Reference Rolin-Ianziti and Varshney2008) draw upon Ellis's distinction between ‘medium-oriented goals’ and ‘framework goals’ in classroom interaction to classify teachers’ reasons for own-language use: respectively, the teaching of the new language (‘the medium’) itself and ‘goals associated with the organization and management of classroom events’ (Ellis Reference Ellis1994: 577–578). According to Rolin-Ianziti & Varshney (Reference Rolin-Ianziti and Varshney2008), teachers might use the learners’ own language for medium-oriented goals, such as explaining vocabulary items or teaching grammar (see, for example, Polio & Duff Reference Polio and Duff1994; Castellotti Reference Castellotti, Castellotti and Moore1997; Kim & Elder Reference Kim and Elder2008) and framework goals, such as giving procedural instructions and assigning homework (e.g. Polio & Duff Reference Polio and Duff1994; Macaro Reference Macaro2001). Littlewood & Yu (Reference Littlewood and Yu2011) similarly draw upon Kim & Elder's (Reference Kim and Elder2008) distinction between ‘core goals’ (teaching the target language), ‘framework goals’ (managing the classroom situation) and ‘social goals’ (expressing personal concern and sympathy) to explore a number of ways in which teachers strategically employ learners’ own languages in class. Citing a number of illustrative works, they focus upon the role of the learner's own language during the presentation, practice and production of new language, such as Dodson's (Reference Dodson1967/1972) and Butzkamm's (Reference Butzkamm2003) ‘bilingual sandwich’ technique, Butzkamm's (Reference Butzkamm2001) ‘mirror translation’ exercises and Duff's (Reference Duff1989) and Deller & Rinvolucri's (Reference Deller and Rinvolucri2002) bilingual learning materials.
Littlewood & Yu (Reference Littlewood and Yu2011) also emphasise the ‘reassuring’ role the learners’ own language can play in the classroom and the potentially alienating effects of monolingual teaching, Allwright & Bailey similarly observing that ‘banishing the learners’ first language. . .deprives [them] of their normal means of communication and so of the ability to behave fully as normal people’ (1991: 173). For example, learners in Chen's (Reference Chen2003) study ascribe their in-class reticence to the demands placed on them by communicative language teaching, whilst Brooks-Lewis (Reference Brooks-Lewis2009) documents stress experienced by adult learners in a monolingual language classroom and their sense of disadvantage in relation to their teacher. Similarly, Auerbach's more overtly political take on the issue emphasises how teacher power and authority can be reinforced by monolingual teaching and, consequently, how own-language use can ‘reduce anxiety and enhance the affective environment for learning’ (1993: 20). (See also Stables & Wikeley (Reference Stables and Wikeley1999) with reference to the negative effect of monolingual teaching on British teenagers’ attitudes to language learning.)
Meanwhile, Stibbard (Reference Stibbard and Malmkjaer1998) highlights the affective-humanistic benefits of mother tongue use with beginner learners in Hong Kong, and Levine (Reference Levine2003) suggests that ‘principled’ and ‘meaningful’ own-language use can contribute to a reduction in learner anxiety. Canagarajah (Reference Canagarajah1999: 132) observes how own-language use puts Sri Lankan learners ‘at ease. . .and creates a less threatening atmosphere’, especially when discussing local events; similarly, in her study of her own practice, Edstrom (Reference Edstrom2006) uses the learners’ own language to ‘connect with students’ and to deal light-heartedly with cultural stereotyping which, she felt, could not be discussed adequately in the new language.
A similar positive effect of own-language use on in-class relationships is documented by Nikula (Reference Nikula2007), who observed Finnish learners employing their own language for affective and interpersonal functions in an otherwise monolingual content-based class. Own-language use has been likewise identified as promoting class unity and identity in a variety of ELT classrooms and contexts, such as in Maltese secondary schools (Camilleri Reference Camilleri1996), South African high schools (Adendorff Reference Adendorff, Bailey and Nunan1996) and Swedish primary schools (Cromdal Reference Cromdal2005). Meanwhile, Schweers (Reference Schweers1999) identifies a role for own-language use in encouraging learner motivation and positive attitudes towards the language being learned.
Consequently, outlining her own reasons for using the learners’ own language, Edstrom (Reference Edstrom2006) argues that the debate transcends concerns about language acquisition processes or ways of managing classroom activities; own-language use is a moral issue. Drawing on the work of Johnston (Reference Johnston2003), who explains morality in terms of the value-laden decisions that teachers make on a moment-by-moment basis and their consequences for learners, Edstrom suggests that teachers have a ‘moral obligation’ to recognise learners as individuals, to communicate respect and concern and to create a positive affective environment (which, in turn, benefits learning). She therefore argues that concerns for learner affect outweigh her belief in maximising new language use, although she notably points out that, whilst this may seem subjective, she is not arguing for ‘purposeless’ or ‘lazy’ own-language use (Edstrom Reference Edstrom2006: 289). Instead, she suggests that ‘thoughtful [and] honest self-analysis’ can help teachers use the learners’ own language ‘judiciously’ (we shall return to the search for ‘judicious’ or ‘optimal’ levels of own-language in section 6.5).
In summary, therefore, there is overwhelming evidence of widespread own-language use and code-switching in language classrooms, so much so that the amount of own-language use is often underestimated or under-reported by teachers. It is also clear that code-switching fulfils a number of clear pedagogic functions, and is employed in similar ways and for similar reasons across a range of differing classroom contexts; as Widdowson (Reference Widdowson2003) points out, given that own-language use is an inevitable and ‘natural’ part of classroom life, it should and can be turned to pedagogic advantage, no longer viewed as an impediment to remove or avoid, but as a positive resource (p. 152). V. Cook (Reference Cook2001), Butzkamm & Caldwell (Reference Butzkamm and Caldwell2009) and G. Cook (Reference Cook2010), among many others, make the same case. However, beyond ‘just’ the pedagogic advantages of code-switching in the classroom, how might own-language use support the process of acquiring a new language? It is to theories of learning, and of language learning, that we now turn.
6. Theoretical frameworks and research findings
The growing theoretical diversity of SLA research has led to cognitive and sociocultural perspectives on language learning being characterised as ‘parallel worlds’ (Zuengler & Miller Reference Zuengler and Miller2006: 36). Yet both these approaches to second and additional language learning provide substantial arguments in support of own-language use in English language teaching, based upon current understandings of both how the bi- and multilingual mind functions and how people learn (Cummins Reference Cummins2007).
6.1 The bilingual mind: psycholinguistic and cognitive perspectives
We have already noted (section 4) the role of the learners’ own language as a ‘natural’ reference system and ‘pathfinder’ (Butzkamm Reference Butzkamm and Weinstock1989) for learners, which they inevitably draw upon as the new language is learned, and also the case for compound bilingualism, in which psychological connections are maintained between languages in the learners’ minds (see V. Cook Reference Cook2001; Widdowson Reference Widdowson2003; Butzkamm & Caldwell Reference Butzkamm and Caldwell2009). From a psycholinguistic perspective, V. Cook's multicompetence model, ‘the knowledge of more than one language in the same mind’ (2008: 231; also 1995, 2002), reconceptualises language learners as bilingual language users who have different uses for language from monolinguals, have a different knowledge of both the new language and their own language from monolingual speakers, and even have different minds from monolinguals. Indeed, brain and behavioural research by Bialystok et al. (Reference Bialystok, Craik, Grady, Chau, Ishii and Gunji2005) and Bialystok & Feng (Reference Bialystok and Feng2009) has found significant differences in the ways monolingual and bilingual language users perform linguistic tasks (see also Grosjean Reference Grosjean1989, Belz Reference Belz2002 and Herdina & Jessner Reference Herdina and Jessner2002 for discussions related to multicompetence). Cummins (Reference Cummins2007: 299) also highlights the ‘enhanced metalinguistic awareness’ that bilingual learners develop as a result of processing two languages, suggesting that learners are likely to benefit from focusing upon the similarities and differences in their two or more languages (for example, focusing on cognates or working on dual language projects). Thus learners should be encouraged ‘to see the first language as something that is part of themselves whatever they do and appreciate that their first language is inextricably bound up with their knowledge and use of the second’ (V. Cook Reference Cook and Cook2002: 339).
Similarly, focusing upon the complex ways in which languages interact in the minds of language learners (i.e. bilingual language users), Cummins (Reference Cummins1981, Reference Cummins2007) suggests that, because of interdependence across languages, the development of a skill or proficiency in one language assists in the development of that same skill in the other language(s). Thus, learners have a common underlying proficiency that is interdependent across languages and which allows for ‘the transfer of cognitive/academic or literacy-related proficiency from one language to another’ (2007: 232). Citing a number of studies from bilingual and immersion programmes in Canada and the US (e.g. Thomas & Collier Reference Thomas and Collier2002; Genesee et al. 2006), Cummins (Reference Cummins2007) identifies five types of cross-lingual transfer which might operate in a variety of ways, depending on the sociolinguistic and educational context (i.e. the transfer of phonological awareness; pragmatic aspects of language use, such as the ability to take risks in communication; metacognitive and metalinguistic strategies, such as vocabulary acquisition strategies; specific linguistic elements, such as the meaning of ‘photo’ in ‘photosynthesis’; conceptual elements, such as understanding the concept of photosynthesis). For Cummins, learning is likely to be more efficient if teachers draw learners’ attention to the similarities and differences between their languages, coordinating and reinforcing learning strategies across languages.
6.2 Tracing own-language use in SLA research
Although, as noted (section 1), ‘mainstream’ SLA has shown remarkably little interest in the ways learners’ own languages might positively influence new language learning, there are clear links between Cummins’ discussion (2007) of the way teachers might draw learners’ attention to the similarities and differences between languages and concepts which are central within current SLA research, such as noticing (Schmidt Reference Schmidt1990; see also Robinson's (Reference Robinson, Doughty and Long2003) discussion of attention) and focus on form (Long Reference Long, De Bot, Ginsberg and Kramsch1991). Indeed, Laufer & Girsai (Reference Laufer and Girsai2008) make the case for explicit contrastive analysis and translation as part of form-focused instruction after finding that learners taught unfamiliar vocabulary items via translation fared better in a subsequent retention test than those taught solely through meaning-focused instruction. They suggest that translation activities stretch learners’ linguistic resources and result in pushed output (Swain Reference Swain, Gass and Madden1985) as, to produce good translations, learners cannot avoid problematic words or structures, a point also emphasised by G. Cook (Reference Cook2010: 136); this ‘elaborateness of processing’ consequently leads to more durable memory encoding (Hummel Reference Hummel1995). However, although there has been some interest in the effects of contrastive form-focused instruction and translation on learner uptake, the area remains significantly under-investigated by SLA research. The few studies there are in this area include Kupferberg & Olshtain (Reference Kupferberg and Olshtain1996), Källkvist (Reference Källkvist, Ortega and Byrnes2008) and Scott & de la Fuente's (Reference Scott and de la Fuente2008) experimental approaches to grammar acquisition; Snellings, van Gelderen & de Glopper's (2002) use of translation tasks to establish vocabulary retrieval speed and Hummel's (Reference Hummel2010) comparison of translation and rote-learning in vocabulary learning.
The possible benefits of own-language use and translation as language learning strategies has also been discussed (e.g. O'Malley & Chamot Reference O'Malley and Chamot1990; Oxford Reference Cohen, Mercer, Ryan and Williams1996; Cohen forthcoming). Indeed, O'Malley & Chamot (Reference O'Malley and Chamot1990) argue that translation is one of the most used learning strategies, whilst Hummel (Reference Hummel2010) reviews a range of evidence which emphasises the value of translation as a potentially effective cognitive strategy for learners.
Meanwhile, there is evidence that code-switching and own-language use facilitates learning by reducing the processing load for learners during cognitively challenging tasks. The extent to which many learners use their own language internally in the classroom is well documented. For example, Blyth (Reference Blyth and Kramsch1995), Antón & DiCamilla (Reference Antón and DiCamilla1999), Centeno-Cortés & Jiménez Jiménez (Reference Centeno-Cortés and Jiménez Jiménez2004) and Alegría de la Colina & García Mayo (Reference Alegría de la Colina and García Mayo2009) all report own-language private verbal thinking (also termed private speech) and mental translation during new language tasks and activities. Whilst many studies adopt a sociocultural approach to private speech (which we shall examine shortly, in section 6.4), Kern (Reference Kern1994; see also Macaro Reference Macaro and Llurda2006) suggests that own-language use lightens the cognitive load of tasks by reducing the demands on learners’ working memory and facilitating cognitive processing; Carless’ (Reference Carless2002) and Scott & de la Fuente's (Reference Scott and de la Fuente2008) studies, of young EFL learners in Hong Kong and college-level learners of Spanish and French respectively, also report learners’ own-language use increasing in line with task complexity. Consequently, Macaro (Reference Macaro and Llurda2006) notes that code-switching and immediate translation during breakdowns in classroom discourse can also lighten the cognitive load on learners, facilitating the processing of other input and providing ‘an efficient shortcut, more related to the learning processes the students are using’ (V. Cook Reference Cook and Cook2002: 340). For Macaro (Reference Macaro and Llurda2006), code-switching and own-language use in the classroom fulfils the role of both a learning strategy and a communication strategy (confirming, perhaps, that the distinction between language use, communication strategies and learning strategies is conceptually unclear; for further debate on this issue, see Oxford Reference Oxford2011; Cohen forthcoming).
The role of code-switching in vocabulary teaching and learning has also been widely documented. Alongside substantial empirical support for the use of bilingual dictionaries (e.g. Prince Reference Prince1996; Nation Reference Nation1997, Reference Nation2003; Bruton Reference Bruton2007; Laufer & Girsai Reference Laufer and Girsai2008), Celik (Reference Celik2003) suggests the insertion of a single word from the own language as an effective way of introducing new vocabulary (see also Macaro et al. Reference Macaro, Chen, Tian and Richards2009). Indeed, Nation, summarising a number of studies which compare the effectiveness of learning strategies, suggests that translation is invariably ‘the most effective’ way of learning vocabulary (2003). Meanwhile, examples of cross-linguistic awareness-raising activities which develop learners’ metaphorical competence (Low Reference Low1988) and their ability to understand and create metaphors in their new language are outlined by Lazar (Reference Lazar1996), Deignan, Gabryś & Solska (Reference Deignan, Gabryś and Solska1997), Harden (Reference Harden, Witte, Harden and Ramos de Oliveira Harden2009) and Thorpe (Reference Thorpe, Witte, Harden and Ramos de Oliveira Harden2009).
6.3 Building on prior knowledge
The idea that learning is most effective when it builds upon the prior knowledge and understandings of learners has support from a number of theoretical traditions. Brooks-Lewis (Reference Brooks-Lewis2009) draws upon John Dewey's humanistic and democratic educational philosophy (1939) to point out that ‘recognizing a person's prior knowledge is another manner of recognizing the person and that is demonstrated with the incorporation of the L1 in foreign language education’ (2009: 228). With clear links to our earlier discussion of the affective constraints learners may experience when trying to behave as ‘normal people’ (see Allwright & Bailey, Reference Allwright and Bailey1991, above), Brooks-Lewis highlights the way in which adult Spanish-speaking learners of English in her study found own-language use motivating as they could utilise their ‘lifetime of investment in prior knowledge and communicative experience’ (2009: 228), making use of their existing language skills and understanding of grammatical concepts.
Similarly, constructivist accounts of learning suggest that ‘new knowledge and understanding is based on what learners already know and believe’ (Bransford, Brown & Cocking Reference Bransford, Brown and Cocking2000: 10). Effective learning therefore depends upon the engagement of prior knowledge, which includes not only previously taught information or skills, but ‘the totality of the experiences that have shaped the learner's identity and cognitive functioning’ (Cummins Reference Cummins2007: 232). Cummins (ibid.) develops this thesis in the context of language learning, suggesting that, if prior knowledge is encoded in the learners’ own language, then the engagement of this knowledge is necessarily mediated through their own language. Language teaching should consequently aim to activate learners’ prior knowledge, but will need to draw upon the learners’ own language to achieve this.
There are clear links between the incorporation of learners’ prior knowledge, in the form of their own language and, for example, contrastive analysis, ‘noticing’ (whereby learners consciously ‘pay attention’ to the gap between their current linguistic performance and the new language) and the development of language awareness (Widdowson Reference Widdowson2003; see also Butzkamm & Caldwell Reference Butzkamm and Caldwell2009). More broadly, however, prior knowledge ‘makes learning significant’ as learners fit new information to the knowledge they already possess (Brooks-Lewis Reference Brooks-Lewis2009: 228); in effect, prior knowledge and the learners’ own language provide a cognitive framework through which new knowledge is constructed and regulated. Studies drawing upon sociocultural theory have further explored this perspective.
6.4 Sociocultural approaches to own-language use
Sociocultural accounts of language learning assert that cognitive development, including language development, is a collaborative process ‘driven by social interaction’ (Levine Reference Levine2011: 24; see also, for example, Vygotsky Reference Vygotsky1978; Lantolf Reference Lantolf2000; Swain & Lapkin Reference Swain and Lapkin2000). Language is seen as a cognitive tool through which learners mediate their mental processing, such as their planning, noticing or reasoning (Swain & Lapkin Reference Swain and Lapkin2000: 253), but this mediating role is derived from the social role which language fulfils as learners regulate themselves and others in socially situated activities (Brooks, Donato & McGlone Reference Brooks, Donato and McGlone1997; Alegría de la Colina & García Mayo Reference Alegría de la Colina and García Mayo2009). Thus ‘psychological processes emerge first in collective behaviour, in co-operation with other people, and only subsequently become internalised as the individual's own “possessions”’ (Stetsenko & Arievitch Reference Stetsenko and Arievitch1997: 161, cited in Swain & Lapkin Reference Swain and Lapkin2000: 254). Language is said to mediate learners’ cognitive activity on both the external (interpsychological) and internal (intrapsychological) planes (Antón & DiCamilla Reference Antón and DiCamilla1999).
From this perspective, therefore, own-language use by learners is regarded as a cognitive tool for learners through which learning is scaffolded. At the interpsychological level, Antón & DiCamilla found that learners use their own language for collaborative talk during tasks, such as jointly explaining the nature of tasks, solving problems and maintaining focus. Similarly, Brooks & Donato (Reference Brooks and Donato1994) also acknowledge that own-language use during language-learning tasks is ‘a normal psycholinguistic process that facilitates L2 production and allows the learners both to initiate and sustain verbal interaction with one another’ (p. 268); Swain & Lapkin (Reference Swain and Lapkin2000), studying language use by English-speaking French immersion students, also in collaborative tasks, suggest that learners’ own language helps them understand task content, focuses attention on form, and helps establish and maintain interpersonal collaboration and interaction. Likewise, Vilamil & Guerrero (Reference Vilamil and de Guerrero1996) report that their ELT learners engaged in the peer review of writing via their own language for similar reasons, whilst Thoms, Liao & Szustak (Reference Thoms, Liao and Szustak2005) also observe own-language use by English language learners to ‘move the task along’ in online chat activities. Thoms et al. (ibid.) additionally suggest that when performing solely in the new language is beyond the learners’ ability, own-language use may allow less proficient learners to maintain interaction with more proficient language users and even access their higher-level knowledge. Although there are parallels with Macaro's (Reference Macaro and Llurda2006) observation (see section 6.2) that code-switching enables communication to continue and lightens the cognitive load on learners, understood from a sociocultural perspective, own-language use may enable learners to work with ‘expert others’ at a level which would otherwise be beyond their reach, thereby working in their zone of proximal development (Vygotsky Reference Vygotsky1978). Further relevant studies of own-language use for interpsychological purposes include Behan & Turnbull with Spek (Reference Behan, Turnbull and Spek1997), Storch & Wigglesworth (Reference Storch and Wigglesworth2003) and Swain & Lapkin (Reference Swain and Lapkin2005).
At an intrapsychological level, Antón & DiCamilla's (Reference Antón and DiCamilla1999) study also notes learners’ use of their own language in private speech to direct their own thinking. Similarly, Centeno-Cortés & Jiménez Jiménez (Reference Centeno-Cortés and Jiménez Jiménez2004) observed own-language private speech by learners of Spanish which demonstrated learner reasoning at the initiation, progression and conclusion of tasks. They also found that there was a shift from new to own-language private speech as learners encountered more difficult tasks.
Pedagogically, therefore, it seems logical to suggest that teachers can facilitate learning by allowing the ‘judicious’ use of learners’ own language. For Swain & Lapkin (Reference Swain and Lapkin2000), this entails neither prohibiting nor encouraging own-language use (in order to avoid own-language use substituting for, rather than supporting, new language learning). Likewise, Centeno-Cortés & Jiménez Jiménez (Reference Centeno-Cortés and Jiménez Jiménez2004) argue that, whilst prohibiting own-language use deprives learners of a key cognitive and metacognitive tool, this does not mean the learners’ own language should be used for all classroom activities, but for specific linguistic or communicative functions in the classroom (Thoms et al. Reference Thoms, Liao and Szustak2005). What, however, might this mean in practice? What is meant by the ‘judicious use’ of the learners’ own language?
6.5 Searching for ‘optimal’ own-language use
The notion of the ‘judicious’ use of the learners’ own language has been touched upon regularly throughout this review. Yet, from Stern's (Reference Stern1992) search for the ‘appropriate’ combination of intra- and crosslingual activities to Edstrom's (Reference Edstrom2006) call for principled and ‘purposeful’ own-language use, the question of what this might entail has remained somewhat vague. Consequently, there have been several calls for more research to find an ‘appropriate ratio’ of own to new language use (Crawford Reference Crawford2004) and ‘when and why’ the learners’ own language might be used (Turnbull & Arnett Reference Turnbull and Arnett2002; see also Stern Reference Stern1992, Macaro Reference Macaro1997, Reference Macaro and Llurda2006, Reference Macaro, Turnbull and Dailey-O'Cain2009a; Meiring & Norman Reference Meiring and Norman2002; Turnbull & Dailey-O'Cain Reference Dailey-O'Cain, Liebscher, Turnbull and Dailey-O'Cain2009). Yet whilst acknowledging the reality and beneficial effects of code-switching in the classroom, this debate also recognises V. Cook's (Reference Cook2001) concern that new language input and use is ‘maximised’ (see section 4.1). Turnbull & Arnett (Reference Turnbull and Arnett2002), for example, cite Ellis’ (Reference Ellis1984) argument that own-language use deprives learners of target language use; citing MacDonald (Reference MacDonald1993), Macaro (Reference Macaro1997) and Turnbull (Reference Turnbull2001) also argue that using the new language in the classroom is particularly motivating, as it enables learners to see its immediate usefulness (as opposed to at some point in the future) and provides opportunities for immediate success. Butzkamm & Caldwell (Reference Butzkamm and Caldwell2009: 150) also make the point that a structured and principled deployment of the student's own language – as opposed to the chaotic way it tends to appear, despite restrictions, in monolingual teaching – can increase rather than decrease the use of the new language for communication.
As Turnbull & Dailey-O'Cain (Reference Dailey-O'Cain, Liebscher, Turnbull and Dailey-O'Cain2009) suggest, therefore, there is some concern on the part of researchers that in the absence of clear research findings, teachers may be making up their own arbitrary rules concerning code-switching. Yet in contrast, there is also a compelling argument that teachers and learners are in fact best placed to decide what is appropriate for their own classroom (McMillan & Rivers Reference McMillan and Rivers2011). Thus whilst there is some evidence that teachers resort to the learners’ own language when they are tired, and that learners use their own language when they go off-task in class (Macaro Reference Macaro and Llurda2006) – examples perhaps of Edstrom's ‘unprincipled’ own-language use (2006) – Macaro also finds that teachers hold coherent sets of beliefs concerning the balance of own and new language use in the classroom. Investigating ‘predominantly CLT-based’ foreign language classrooms in UK schools, Macaro (Reference Macaro1997) identified three distinct personal theories amongst teachers. The ‘virtual position’ posits that the classroom is a ‘virtual reality’ which mirrors the environment of first language learners and migrants in a new country; this was acknowledged by Macaro to be an ‘unattainable ideal’ reflecting perfect learning conditions which do not exist in instructed language learning. Consequently, he found that many teachers aspire to a ‘maximal position’ of new language use; from this perspective, however, use of the learners’ own language was ‘tainted’ (2009a: 36) and led to feelings of guilt amongst teachers. Finally, Macaro identified an ‘optimal position’ in which the value of own-language use was recognised as enhancing learning at certain points during lessons, more so than using only the new language. As we have seen, there is considerable evidence in support of ‘optimal use’ of code-switching in class, leading Macaro (Reference Macaro, Turnbull and Dailey-O'Cain2009a: 38–39) to suggest that it involves teachers making a judgment about:
the possible detrimental effects of not drawing the learners’ attention to aspects of their first language, or not making comparisons between the first and second languages. It involves a principled decision regarding the effects of not conveying important information simply because this might be too difficult for the learners to understand in the second language – a teacher avoidance strategy. It involves decisions about the relative merits of second language input modification as opposed to activating first language connections.
However, in arguing that these should be informed judgments, Macaro (ibid.) also acknowledges that there are ‘virtually no studies’ which have considered whether code-switching into the learners’ own language while maintaining new language discourse actually results in more effective learning; clearly more research is needed to address this issue.
6.6 Complementary frameworks? From psycholinguistics to sociolinguistics
This broad review of empirical studies has taken in a range of findings which support the use of the learners’ own language in the classroom. The discussion has moved from cognitive and psycholinguistic to social and sociocultural perspectives on language learning, although it is clear that, in terms of informing real world practices, these two theoretical approaches cannot be so simply separated. Thus whilst, as already noted, SLA research has traditionally overlooked own-language use or regarded it as an unwelcome barrier to new language input and use, recent conceptualisations of bi- and multilingual competence and cognition have suggested that own-language use is not only inevitable within the language classroom, but contributes positively to new language development. Approaching the issue from a very different theoretical perspective, sociocultural approaches to learning offer similarly substantial support to these claims.
Undoubtedly, learners still require significant exposure to, and practice of, new language. Equally, many researchers and teachers are still extremely cautious about encouraging own-language use (we shall further examine teachers’ attitudes and beliefs, beyond those already summarised, in more detail in section 7). We should also note that several of these studies were undertaken in bilingual and immersion contexts in Canada, or with learners of foreign languages other than English, albeit in generally communicative contexts. Yet the relevance of psycholinguistic and sociocultural arguments for ELT, favouring principled own-language use and code-switching, seems clear; and these approaches can be brought together via a sociolinguistic perspective of the language classroom as a ‘bilingual space’ in which learners are developing into bilingual or multicompetent language users whose own language complements the development of their new language (V. Cook Reference Cook and Cook2002; Edstrom Reference Edstrom2006; Turnbull & Dailey-O'Cain Reference Dailey-O'Cain, Liebscher, Turnbull and Dailey-O'Cain2009). From this perspective, therefore, the language classroom can be conceptualised as a bi- or multilingual community of practice (Wenger Reference Wenger1998) which is characterised by ‘mutual engagement’, ‘joint enterprise’ and ‘shared repertoire’. Consequently, own-language use is understood as a legitimate practice which contributes to the classroom's ‘conceptual architecture for learning’ (p. 230), that is, alongside other classroom procedures and practices, code-switching is a resource through which ‘the conditions for learning’ are created (Levine Reference Levine2011: 42). Furthermore, learners’ bi- and multilingual identities are accepted which, in turn, makes a positive contribution to learning (Liebscher & Dailey-O'Cain Reference Liebscher and Dailey-O'Cain2005; Unamuno Reference Unamuno2008).
Thus the extent to which own-language use occurs in a language classroom will in many ways depend on the teacher's and learners’ perceptions of its legitimacy, value and appropriate classroom functions. It is to these attitudes and beliefs that we now turn.
7. Teachers’ and learners’ attitudes
As outlined in the early sections of this paper, own-language use has been regarded as ‘the skeleton in the cupboard’ of English language teaching (Prodromou Reference Prodromou, Deller and Rinvolucri2002: 5), with negative perceptions of the issue maintaining a ‘stranglehold’ (ibid.) on teachers’ attitudes and beliefs. Consequently, many studies have reported teachers’ sense of ‘guilt’ when learners’ own languages are used in the classroom (e.g. Mitchell Reference Mitchell1988; Macaro Reference Macaro1997, Reference Macaro and Llurda2006, Reference Macaro, Turnbull and Dailey-O'Cain2009a; Butzkamm & Caldwell Reference Butzkamm and Caldwell2009; Turnbull & Dailey-O'Cain Reference Dailey-O'Cain, Liebscher, Turnbull and Dailey-O'Cain2009; Littlewood & Yu Reference Littlewood and Yu2011); as Macaro (Reference Macaro and Llurda2006: 69) observes, teacher guilt ‘is not a healthy outcome of pedagogical debate’.
Yet teacher attitudes towards own-language use are more complex than just feelings of guilt. For example, Macaro reports that ‘the majority of bilingual teachers regard code-switching as unfortunate and regrettable but necessary’ (2006: 68; emphasis added), whilst, as we have seen, numerous studies have elicited a more positive and reasoned justification for own-language use in the classroom. Examples include classroom management and administration, grammar instruction, and to demonstrate empathy with learners (Duff & Polio Reference Duff and Polio1990, see section 5.2), ‘moral obligation’ (Edstrom 2008, also section 5.2) and Macaro's discussions concerning ‘optimal position’ for own and new language use (1997, 2009a, 2009b, see section 6.5; also van der Miej & Zhao 2010; McMillan & Rivers Reference McMillan and Rivers2011). Summarising the literature, therefore, Macaro notes the ‘overwhelming impression that bilingual teachers believe that the L2 should be the predominant language of interaction in the classroom. On the other hand,. . .[we do not find] a majority of teachers in favour of excluding the L1 altogether’ (2006: 68, original emphasis).
Teacher beliefs are derived from and shaped by a range of sources, including teachers’ own experiences as language learners, their practical experience of what is and is not successful in class, and the perspectives of others (including colleagues, teacher trainers and educators, managers and policy-makers and academic research and researchers). Additionally, teacher beliefs can change over time; for example, Atkinson (Reference Atkinson1993), Auerbach (Reference Auerbach1994), Burden (Reference Burden2000), Mattioli (Reference Mattioli2004) and Butzkamm (n.d.) all document how, having originally advocated a monolingual approach to teaching, they later came to view the learners’ own language as a resource for language learning (cited in McMillan & Rivers Reference McMillan and Rivers2011).
Frequently reported teacher beliefs include, for example, the notion that the balance between own and new language use is most consistently affected by learners’ ability (Mitchell Reference Mitchell1988; Macaro Reference Macaro1997; Crawford Reference Crawford2004). Meanwhile, Liu et al. (Reference Liu, Ahn, Baek and Han2004) identify teachers’ beliefs about their own new language proficiency as a factor affecting own-language use (also Carless Reference Carless2004; Bateman Reference Bateman2008; Kim & Elder Reference Kim and Elder2008; Nagy & Robertson Reference Nagy, Robertson, Turnbull and Dailey-O'Cain2009).
Obviously, however, not all teachers hold the same beliefs. In contrast to the studies cited in the previous paragraph, for example, van der Meij & Zhao (Reference van der Meij and Zhao2010) find that teachers of English working in two Chinese universities disagree with the notion that learner or teacher proficiency should affect own-language use. This raises the possibility that attitudes and beliefs towards code-switching vary according to cultural background and educational tradition. Similarly, in their study of Japanese-origin and British-origin teachers of Japanese in UK secondary schools, Hobbs et al. (Reference Hobbs, Matsuo and Payne2010) found, perhaps counter-intuitively, that the non-native speakers of Japanese had a more positive attitude towards own-language use than the native speakers, a finding they ascribe to the teachers’ different cultures of learning. Furthermore, Canada, with its particular history of bilingual education, provides a specific environment in which teachers are apparently sympathetic to own-language use and code-switching in class (as shown by the numerous studies cited in this paper, such as Duff & Polio Reference Duff and Polio1990; Rolin-Ianziti & Brownlie Reference Rolin-Ianziti and Brownlie2002; Cummins Reference Cummins2007).
Meanwhile, the potentially differing attitudes to code-switching of teachers who do or do not share the learners’ own language has also been touched upon. Harbord (Reference Harbord1992: 250), for example, refers to ‘frequent differences of opinion’ between native and non-native English speaker teachers (nests and non-nests in Medgyes’ (Reference Medgyes1994) terminology) over whether to use the learners’ own language in class. In their more recent study, however, McMillan & Rivers’ (Reference McMillan and Rivers2011) found that NESTs working in a broadly CLT environment in Japan were evenly divided between those with a positive and those with a negative attitude to teacher and learner own-language use. Interestingly, not all NESTs with limited Japanese proficiency (i.e. the learners’ own language) regarded judicious own-language use by teachers negatively; and teachers with lower Japanese proficiency tended to have a more positive view of learners’ own-language use than higher Japanese-proficiency teachers. In justifying their views, teachers suggested several arguments in favour of own-language use which have already been identified in this paper, such as to save time, build rapport with learners and engage in contrastive analyses of their own and the new language (and participants’ arguments against own-language use similarly reflected points touched on in this review, such as learner motivation and the need for exposure to language input). Consequently, echoing Macaro's suggestion that teachers’ judgement is central to the establishment of ‘optimal’ levels of own and new language use (2009a, see above), McMillan & Rivers call for an ‘English-mainly’ policy rather than ‘English only’, based on the reflections of teachers as professional decision-makers in their own local context.
McMillan & Rivers’ discovery of relatively little difference between the views of NEST and non-NESTs in a specific Japanese context may therefore be typical of contexts around the world; certainly the many studies cited in this paper which examine teachers’ experiences and perspectives of own-language use do not seem to reveal a wide gap between NEST and non-NEST perspectives of the issue. This is encouraging if NESTs and non-NESTs, and monolingual and bilingual language teachers, are to work together in more effective and complementary ways to teach English through cross-lingual approaches. However, more research is clearly needed in this area.
A small number of studies have also looked at the beliefs of student teachers (as opposed to in-service instructors) with regard to own-language use. Orland-Barak & Yinon (Reference Orland-Barak and Yinon2005) explored the perspectives of 14 Arab and Jewish EFL teachers in Israel, finding that, over the course of their initial teaching practice, they ‘exhibited new insights regarding the different purposes for which L1 can be used in a communicative lesson’ (p. 91), challenging the ‘myth’ of new language only in the classroom (p. 97); in their classes, the learners’ own languages were used for explanation, rapport-building and classroom management. Interestingly, although the study aimed to investigate the ways in which cultural differences might affect the beliefs of Arab and Jewish student teachers, the teachers in this study exhibited ‘strikingly similar’ attitudes towards own and new language use in the classroom (see the discussion of the relationship between teachers’ attitudes and cultural/educational background, above). Meanwhile, Macaro's (Reference Macaro2001) study of the practices and beliefs of six English-speaking teachers of French found that, although the learners’ own language was used surprisingly infrequently, it was utilised for the (by now familiar) functions of giving instructions, classroom management and discipline. Exploring in detail the reasoning of two of the participants, Macaro also found that whilst what one teacher allowed was heavily influenced by the perspectives of policy-makers and statements, the other continued to follow her personal beliefs, with implications for the success (or otherwise) of initial teacher education programmes.
Finally, learners’ attitudes towards code-switching will also clearly affect the extent and function of own-language use in the classroom, and its potential contribution to learning; do learners, in fact, prefer monolingual teaching and learning, as is often asserted?
Both Turnbull & Arnett (Reference Turnbull and Arnett2002) and G. Cook (Reference Cook and Beaven2008) highlight the relative absence of research into learner perceptions of own-language use. A notable exception, however, is Duff & Polio's study of 13 language classes (1990; see also section 5.4), which found that a large majority of learners were satisfied with the amount of own-language use by their teachers, irrespective of whether this was a large amount or not. Rolin-Ianziti & Varshney (Reference Rolin-Ianziti and Varshney2008), meanwhile, found that the majority of beginner learners of French in an Australian university preferred own-language use for classroom management. Their study also confirmed the positive affective role code-switching can play in the language classroom, participants suggesting that own-language use often reduces learner anxiety, though it may also result in a lack of challenge and subsequent learner demotivation. Meanwhile, Pease-Alvarez & Winsler (Reference Pease-Alvarez and Winsler1994) and Brooks-Lewis (Reference Brooks-Lewis2009) uncovered positive attitudes towards bilingual classroom practices amongst young and adult Spanish learners of English in the US and adult learners of English in Mexico, whilst Chavez's (Reference Chavez and Blyth2003) wide-ranging study of 330 US college learners of German indicated that most learners considered and preferred their classroom to be ‘diglossic’; indeed, Chavez reports that teachers tended to use the new language more than their learners actually wished. Beyond these studies, in which learners were directly asked for their opinions, a generally positive picture of learner engagement with own-language use also emerges when we reflect upon evidence for the ‘humanistic’ and ‘reassuring’ function of own-language use (Harbord Reference Harbord1992; Littlewood & Yu Reference Littlewood and Yu2001, respectively) which was highlighted earlier in this review (see section 5.2).
8. Entrenched monolingualism in ELT
As noted at the beginning of this article, new directions in the language learning literature are never universally adopted. Despite the overwhelming force of the arguments and evidence in favour of bilingual language teaching in a globalised multilingual world, many curricula, institutions, syllabus and materials designers, as well as teachers, parents – and, of course, students – remain committed to monolingual teaching.
A notable manifestation of diehard monolingualism, strangely posing as a new approach, is content-based language teaching, in which school curriculum subjects are taught through the medium of a new language on the assumption that this simultaneously furthers both student proficiency in that language and their knowledge of the subject in question. It has been energetically promoted in a number of countries, including the UAE, Mexico and South Korea for the teaching of English, and in Europe notionally for the teaching of any language, under the name content and language integrated learning (clil) (Marsh Reference Marsh2002; Coyle, Hood & Marsh Reference Coyle, Hood and Marsh2010).
In practice, however, CLIL is most frequently used to extend the teaching of English. Despite running counter to the EU commitment to maintaining linguistic diversity, English-medium CLIL in Europe has received intensive support, in the shape of investment by the EU in research, development and implementation, and, surprisingly – counter to the trends surveyed in this article – endorsement by language teaching theorists and applied linguists. Meanwhile, in many postcolonial contexts such as South Africa, where the ex-colonial language (i.e. English) is used as the primary language of instruction, the effects of CLIL have been at times disastrous, maintaining the dominance of English and acting as a barrier to multilingual and multicultural socialisation (Chick & McKay Reference Chick and McKay2001).
The applied linguistic arguments advanced in its favour (Dalton-Puffer Reference Dalton-Puffer2007: 257–277, ICRJ 2008 onwards) draw upon the old SLA view that exposure and attention to meaning will be sufficient factors for language learning success, and also on analogy with content-based instruction in bilingual second language learning contexts such as Canada rather than foreign language teaching programmes. Advocates broadly subscribe to the notions that ‘the requisites of success lie in exposure’ (Marsh Reference Marsh2002: 9) and that ‘early introduction (4–12 years) is now. . .advantageous’ (ibid.: 10) Bilingual teaching (‘trans-languaging’ in the CLIL jargon) is seen as something which will – and should – wither away (Marsh Reference Marsh2002: 98). The utilitarian rationale (ibid.) virtually ignores the complex impact on diversity and identity both of this major extension of English into classrooms and subject areas where students’ own languages previously held sway, and in contexts where the ex-colonial language maintains its dominant position.
While there is surprisingly little academic criticism of English-medium CLIL, those asked to implement it sometimes express strong opposition and suspicion of the motives behind it; it is also seen as detrimental to the maintenance of linguistic diversity (G. Cook Reference Cook2010: 115). In South Korea, for example, the Korea Times (2008) reports that the government's English immersion programme seems to have been halted, partly because of teacher resistance. Meanwhile, in South Africa, increasing attention has been given to the ‘judicious’ introduction of African languages (i.e. own languages) as complementary languages of instruction (see, for example, Bloch Reference Bloch, Torres-Guzmán and Gómez2009; Wildsmith-Cromarty Reference Wildsmith-Cromarty, Torres-Guzmán and Gómez2009).
9. Practical applications and pedagogical approaches
Over the course of this article, we have outlined the ways in which the changing academic and political climate surrounding English language teaching has led to an increased acceptance of own-language use in the classroom. We have also documented a wide range of empirical evidence which demonstrates the importance and functions of own-language use and code-switching in language teaching and learning. Presenting a state-of-the-art ‘review of recent and current research’, our discussion has reflected the ways in which theoretical approaches and empirical studies have conceptualised and modelled own-language use, focusing on, for example, the overall goals of classroom interaction (e.g. ‘core’, ‘framework’ and ‘social’ goals – see section 5.2) rather than specific classroom goals or activities. Yet a substantial body of research and research-informed teaching materials highlights the ways in which learners’ own language might be utilised in specific classroom practices and activities. It is to this that we turn in the final section of our review.
For mainstream ELT publishers, a tension exists between the desire to produce materials for global distribution and the increasingly important demand for coursebooks which meet the needs of teachers and learners in a particular country or region (Bolitho Reference Bolitho2003). Clearly, it is easier for localised learning materials to take account of local traditions of learning, including the degree to which the learners’ own language is used in class and in published coursebooks. Consequently, many major ELT publishers now produce country- or region-specific materials which incorporate to some extent the learners’ own language (e.g. through translated word lists, own-language rubrics or grammatical explanations). Interestingly, however, several widely-distributed global textbooks now also integrate translation into activities, albeit on a relatively minor scale (e.g. Soars & Soars Reference Soars and Soars1986 onwards; Swan & Walter Reference Swan and Walter1990 onwards).
Beyond this, however, a number of teacher resource books outline a range of practical ideas and classroom tasks, for example Duff (Reference Duff1989), Deller & Rinvolucri (Reference Deller and Rinvolucri2002), González Davies (Reference González Davies2004) and Butzkamm & Caldwell (Reference Butzkamm and Caldwell2009). They include, for example, activities which explicitly examine the role of the learners’ own language in learning and in the classroom; contrast own and new language forms; develop role-play and drama via own-language preparation; emphasise the efficiency of a bilingual approach to vocabulary development; and focus on the value of translation as a ‘fifth skill’ in language learning.
Although these resources are generally framed within a ‘communicative’ approach which emphasises interaction, collaboration and meaning, they often involve a more ‘traditional’ focus on form, accuracy and individual study. Interestingly, they differ in the extent to which they accommodate monolingual teachers (i.e. teachers who do not speak the learners’ own language); are appropriate to beginners, intermediate and advanced learners; take account of younger learners and learners with different learning styles, experiences and preferences; and are suitable for classes in which learners share the same language or for mixed-language classes. Thus Butzkamm & Caldwell's (Reference Butzkamm and Caldwell2009) bilingual approach calls for teachers who can code-switch effectively in the learners’ own language. In contrast, Deller & Rinvolucri (Reference Deller and Rinvolucri2002: 10) identify a number of activities in which teachers who do not know the learners’ own language (or who are teaching mixed-language classes) may cede ‘full autonomy. . .[to learners] using their mother tongue’. Similarly, G. Cook, reviewing ‘what [kind of own-language use], when, what for, by whom, and with whom’ (2010: 125–153), notes a place for own-language use in mixed-language classes, and makes specific suggestions for translation activities which are possible for teachers who do not share their students’ own language(s). Although the possibilities for own-language use in these contexts are more limited than those available to bilingual teachers working with single language classes, ‘they help to establish the presence and relevance of learners’ own languages in the classroom’ and the reality of a bi- and multilingual world (p. 152).
At the start of the twenty-first century, therefore, now that ‘the long silence’ (G. Cook Reference Cook2010: 20–37) about bilingual teaching has been broken, and its merits are no longer routinely ridiculed and dismissed, the way is open for a major ‘paradigm shift’ in language teaching and learning (Maley Reference Maley2011). The literature reviewed in this article is no doubt only a beginning.
Graham Hall is Senior Lecturer in Applied Linguistics at Northumbria University, UK, where he coordinates and teaches on the university's M.A. Applied Linguistics for TESOL and M.A. TESOL programmes. He has been involved in English language teaching/TESOL for twenty years, working as a teacher, teacher educator and researcher; his research interests range from classroom discourse to the cultural politics of TESOL. He is the Key Concepts editor for the English Language Teaching Journal and has recently published Exploring English language teaching: Language in action (Routledge, 2011).
Guy Cook is Professor of Language in Education at King's College London, UK. He was formerly head of TESOL at the London University Institute of Education (1991–1998), Professor of Applied Linguistics at the University of Reading (1998–2004) and Professor of Language in Education at The Open University (2005–2012). He has published extensively on applied linguistics, English language teaching and discourse analysis. He was co-editor of the journal Applied Linguistics 2004–2009, and is current Chair of the British Association for Applied Linguistics. His books include Translation in language teaching (Oxford University Press, 2010) (winner of the International House Ben Warren Prize), Genetically modified language (Routledge, 2004), Applied linguistics (Oxford University Press, 2003), The discourse of advertising (Routledge, 2001) and Language play, language learning (Oxford University Press, 2000) (winner of the MLA Kenneth Mildenberger Prize).
