What happens to a language when it goes global? We do not really know. It may look like any other language, only bigger. Or scale may do something new to it. What we do know is that when languages get very small, below a certain level they disappear fast, because there is no community to sustain them. When they get very large, they tend to spread themselves thinly and start diversifying. When they get enormous, and a lion’s share of the use is as a lingua franca – this is uncharted territory.
This book addresses English as a lingua franca (ELF) – its characteristics, processes of use, and the linguistic consequences of its unprecedented spread. Clearly, this is a vast area of study, and one book can but scratch the surface. What I have chosen to do is to put one territory into special focus, and criss-cross that area with a few paths so as to get an overview of the landscape and what more there might be to be discovered. My province is academia, and in this domain I analyse samples and endeavour to see what they might tell us about the wider scene of which they form a part.
Academia is a good choice for exploration because it is inherently international, its domain spans the globe, and it has settled on English as its common language. The academy relies on verbal communication, exchanging concepts and ideas that are abstract, novel, and complex; academic discussion makes sophisticated demands on verbal skills. How speakers manage in a lingua franca in such circumstances tells a very different story from studies describing the spontaneous development of pidgins or lingua francas between speakers or groups to cope with rudimentary communicative needs. The growing global higher education market brings a vast proportion and a typologically diverse selection of the world’s languages into contact with English. This gives us a rare opportunity to investigate the intricacies of lang-uage contact of unforeseen complexity.
The exploration of ELF has potential for uncovering general, perhaps universal, features of language use – processing, change, and language in interaction. In this lies its theoretical contribution. It helps gauge phenomena that are similar in language use across the board, whether speakers and environments are monolingual or multilingual, what aspects may be shared by second language use and language contact, and what might be specific to lingua francas, or just to ELF itself.
Among the general theoretical and descriptive issues are matters of processes: how do speakers go about dealing with the contingencies of communicating in a non-native language? We can pose questions about multilingual processing as a cognitive phenomenon, taking into account the contingencies of online processing, or we can look at it in terms of communicative strategies, such as discourse reflexivity. Both are taken on board in this book, with the focus distinctly on language and discourse.
A wider perspective of language change as a consequence of lang-uage contact presents questions arising from today’s increased geographic and social mobility: although it is known that language change accelerates in periods of mobility, there are also centripetal forces counteracting change, seeking commonalities across uses and users, in order to secure a sufficient level of comprehensibility. Language change may also be affected by numbers of speakers. As Croft (2000: 204) points out, a possible scenario where native speakers accept and adopt non-native features is when native speakers are outnumbered by non-natives. Like many sociolinguists discussing language contact, he speaks of ‘shifting’ speakers, which does not directly apply to ELF: lingua francas are used by speakers who maintain their first languages. Thus, even though speakers who use English as an additional language outnumber native speakers (speakers of English as a native language, ENL), their influence is even harder to predict than the direction of language change in general. The question of numbers is nevertheless important, as are the shifting parameters of prestige. Standard English is the unquestioned prestige variety at the moment, but since the status of languages and varieties normally follows that of their speakers, alterations in social and political power on the international scene may well affect the balance between different Englishes as well.
ELF is, then, also specifically about English. As we go about describing characteristics of ELF, the questions that arise are particularly interesting for probing variability and regularity: how do they relate to other Englishes? English has diversified as it has spread around the world, giving rise to a large number of indigenised varieties. Over the last couple of decades, World Englishes have attracted lively research (Kachru 1990; Kortmann et al. 2004; Schneider 2006; Mesthrie & Bhatt 2008; papers in Kirkpatrick 2010, to name but a few). Even though it is out of reach of the present book to make comparisons beyond some sporadic observations, developments in second-language varieties (Mukherjee & Hundt 2011) are significant to ELF and changing English. Similarities and parallels with present or past developments in native English (such as the recent increase in -ing form, see Smith 2005; Mair 2006; Ranta 2006) raise further issues about changes in language and the constraints that a given matrix language imposes. Features of lexis, grammar, and phraseology are likely to be affected, as is pragmatic usage. The consequences of the massive volume of ELF to English taken as a whole, as the aggregate use of the language, is essentially a descriptive issue: what new features of English are emerging and taking hold?
The timescale of language change is variable, with some changes occurring rapidly and others slowly, and in many cases old forms will linger on alongside innovations. Some strata or systems of language are more susceptible to change than others, and lexis has usually been found to be the site of rapid spread of influences, in contrast to structure, which is more resistant. Fast changes may nevertheless be more ephemeral, and it seems that the robustness of linguistic elements is in part dependent on the timescale that we adopt in assessing it. It is not an easy task to set an appropriate timescale for assessing change triggered by the use of ELF, or to give definitive answers concerning the features that have changed or will change. We can observe features that are common to ELF in different contexts, but their possible influence on English as a whole takes longer to observe. This issue is taken up in Chapter 2, and discussed further in connection with empirical analyses in later chapters. What seems clear at this point is that the worldwide spread of English, carefully described in several recent treatises (e.g. Crystal 1997; Brutt-Griffler 2002; Jenkins 2009), has continued from the early seventeenth century through different phases and under different circumstances, but that the most recent, explosive expansion could perhaps most reasonably be timed to coincide with the rise of the Internet. This took place in the mid-1990s and had an enormous effect on global communication in ELF: on this count, we are living the first generation of ELF.
English must have been used as a lingua franca from very early days (as far back as we can speak of ‘English’), most likely long before any written records survive. However, as a phenomenon of a global scale, ELF is relatively new. Its nature and position is only gradually making its way to common awareness, where well-entrenched conceptualisations operate in terms of native speakers and language learners.
A question that often arises when discussing ELF with people who are new to it, or have only vaguely heard about it, is: how is ELF different from learner language? This topic is not very prominent in the rest of this book, although comparisons to learner language come up here and there. Therefore, to clarify the difference and at the same time to illustrate certain basic issues about ELF, I draw some distinctions between learner language and lingua francas at this point. The discussion is also intended to illuminate the approach to ELF adopted in this book, namely a three-level view on lingua francas: a macrosocial, a cognitive, and a microsocial or interactive perspective.
Using a lingua franca means being a user of a second language (L2) but not a learner. We can draw a line between second language use (SLU) and second language acquisition (SLA), with important implications. They have of course much in common: fundamental bilingual processes must be essentially very similar whenever the simultaneous use of more than one language is being dealt with. Learners, translators, and other bilingual users such as early bilinguals, initial monolinguals with their later languages, or plurilinguals with their complex mixture of language resources can be assumed to share some processing mechanisms. It is also a reasonable guess that a speaker’s later languages are less well entrenched than their first. Similarities found across bilingual language use are therefore not surprising. All speakers probably have common processes in their first, second, and nth languages, such as memory storage and retrieval in terms of neural pathway formation, information chunking, or simultaneous processing and monitoring. One would nevertheless expect that things like ease and speed of retrieval, access to alternative expressions, or mapping linguistic and social repertoires routinely onto each other operate differently in a speaker’s first and other languages. By and large, processes that diverge from monolingual first language use are likely to have much in common among all bi- or multilinguals, and therefore research on any group is of interest to the study of all the others. Together they deepen our understanding of the fascinating multifariousness of human languages in use: monolingualism is neither the typical condition nor the gold standard.
Much of what separates learners and users boils down to the peculiar social environment of the classroom. A classroom is a social environment of its own kind which imposes particular social positions on learners that do not hold outside its confines. The learner position outweighs any other social parameters in a classroom setting, whereas outside the classroom other social parameters override learner status. Educational and classroom goals are relevant from a learner position, and they regulate the norms of interaction in every respect. This is particularly obvious in giving and receiving instructions or feedback, providing and following models of behaviour, and in practices of assessing performance. Some pedagogical genres are entirely confined to educational environments, such as particular kinds of question–answer sequences, fill-in exercises, essays, or ‘composition’. Out of the bounds of educational settings, interactional parameters do not follow classroom rules. While it is possible to negotiate and transcend such borderlines at times, this does not often happen; even when a non-native speaker is assuming a learner position relative to a native-speaker interlocutor, native speakers tend to orient to them as speakers, attending to the contents of what is being said, not correcting their language (see Kurhila 2003, 2006).
It is often pointed out that people can alternate in these roles – in the classroom they are learners, but as soon as they get outside they may turn into users of the same language. Thus, the argument goes, the identities are inseparable because the people are the same. However, identities are not simple and constant, but assumed situationally. Different identities and their elements are foregrounded, backgrounded, and drawn on in response to the social environment. When people enter an educational context as learners, their position shifts from that of a user.
A learner identity can also be seen as reductive and limiting, as has been pointed out by Firth and Wagner (1997) who criticise the SLA research paradigm for precisely that: seeing learners as deficient communicators who struggle with difficulties. Firth and Wagner called for a broader basis of SLA studies that would embrace the everyday use of a second language outside classroom settings and include L2–L2 communication. Lingua franca research is one way of broadening the horizon for SLA in this way, expanding it to include SLU.
Sociocultural corollaries of the classroom vs. lingua franca difference include cultural assumptions and the relationship with ‘target culture’. ELF speakers do not share a cultural background or a first language, whereas in a vast number of classrooms around the world one or both are shared. Much pedagogical effort rests on the assumption that learners of the same native language will have similar problems with a target language and are therefore offered similar remedies. In an environment of shared linguistic and cultural assumptions, orientation to the new language is also shared, and along with it, cultural identities and expectations relative to target language speakers. English-speaking countries, or more precisely in Kachru’s (e.g. 1985, 1990) terminology, the ‘inner circle’, constitute ‘target cultures’ which learners are to view against their own cultural background for comparison, contrast, and models of target appropriateness. In contrast, a lingua franca is chosen as a matter of convenience or necessity, and interlocutors may know little of each other’s cultural backgrounds or be unfamiliar with Anglo-American cultures. Conventions of English-speaking cultures may be quite inappropriate when communicative effectiveness hinges upon dealing with cultural mixes of different kinds. A definable, let alone national, target culture for norms of appropriateness or politeness is irrelevant and unreliable as a repertoire of shared assumptions.
A sociolinguistic question of a more macrosocial kind is the relationship to Standard English. It is perhaps useful to distinguish standards, as ‘imposed norms’, from natural, or spontaneous, norms. Norms can be imposed from outside, and standard languages have been a typical case of such societal regulation, particularly pronounced in nineteenth-century Europe with the development of the nation-state. Such imposed standards are different from the natural norms that arise in groups and communities primarily in face-to-face interaction to regulate interaction in the interests of mutual intelligibility and smooth communicative progress. Natural norms arise from what a speech community adopts, tolerates, or rejects. ‘Norms’ can thus refer either to those imposed in standard languages or to the much more variable norms of language communities, which include non-standard varieties and variant forms. The distinction is not always clearly maintained in discussions where ‘native-speaker norms’ are confusingly referred to in either meaning. The issue is perhaps further blurred in corpus-based language teaching, where the preferred usage as manifest in native-speaker data is taken as a model. This is obviously a more accurate reflection of natural speech community norms than prescriptive reference works, including as it does some variability. However, reference corpora are not compiled with wide sociolinguistic representation in mind; they are biased towards standard language and ‘good usage’.
It is standard language that guides learners and teachers. Learner language cannot influence the target language by definition, because learners are committed to acquiring proficiency in terms of given standards. Learners get corrected for their errors, and accept the correction because it is in line with the goals of the whole enterprise of learning – and the target language remains intact. ELF does not have a ‘target language’ because it is a vehicular language itself, an instrument for achieving communication. Although ELF is typically associated with fleeting encounters between strangers, it is also the working language of more long-lasting communities, for example business, trade, or academia. Spontaneous norms arise in communities of these kinds; they can thus become endonormative for their own duration and purposes. In the absence of linguistic authority other than communicative efficiency, group norms are negotiated internally. Communities of practice can share members, individuals in them change and move across communities; in this way these mobile individuals become agents of spreading the practices of one community to the next. In brief, the diffusion of usages initiated in ELF communities resembles the mechanisms of diffusion of innovations or dialect features in ENL communities. Communities of this kind, with some duration, are likely sites for innovations that may spread to the language beyond their own boundaries, as will be discussed in Chapter 2.
In cognitive terms, lingua franca speech orients to achieving mutual comprehension. Many ELF scholars have noted a strong orientation to content over form in ELF discourse (e.g. Karhukorpi 2006; Ehrenreich 2009), in line with researchers working on L1–L2 interaction in real-life conversations (Kurhila 2006). In contrast, learners’ cognitive orientation is far more towards language form. This is a consequence of the pedagogical setting, where learning things like grammar or phonology are necessarily in focus. Feedback and evaluation are based on mastering elements of the language, and it is hard to imagine an educational setting where this would not be so. Although long-term objectives of SLA curricula are often defined in real-life communicative terms, those objectives remain outside the classroom. They can only be simulated in class, but not performed there. Real-life communication cannot be assessed in class for success or effectiveness outside it. Communicative simulations may be pedagogically useful, as Widdowson has often pointed out (e.g. 2000), even realistic and meaningful, but not authentic in its basic sense of being real (cf. Mauranen 2004a). Success in SLA and SLU contexts thus depends on different criteria, and the situated cognitive orientation in SLA and SLU consequently diverge.
The cognitive load in ELF is unusually heavy on account of the variety and unpredictability of language parameters: interlocutors’ accents, transfer features, and proficiency levels. Even in multilingual classrooms such variability is more confined, and speakers adapt to their classmates’ speech fairly soon on account of regular contact (cf. Smit 2010).
At the interface of the wider social context and an individual’s cognitive determinants of processing is social interaction, by means of which speakers go about negotiating language and the norms of interaction itself. ELF speakers typically find themselves in situations where discourse norms are not clear or given: group norms are negotiated within ELF groups by participants, none of whom can claim the status of a linguistic model. Even if ENL speakers are present, they can decline the role of language expert (Hynninen forthcoming). ELF speakers engage in various interactive strategies to achieve mutual comprehensibility; they seem to be prepared for the possibility of misunderstanding and take steps to pre-empt that, which in effect results in misunderstandings being rare (Mauranen 2006a; Kaur 2009). In co-constructing comprehension, any solutions that serve the purpose may be adopted by common consent – the best solutions need not be the most standard-like or native-like (see Hülmbauer 2009). Practices ruled out in L2 classrooms, like language mixing, can be effective strategies in ELF communication (e.g. Klimpfinger 2009). In mediating norms, speakers may even deliberately appeal to practices that are international although not in accordance with British or American norms (Hynninen 2011). For learners the appropriate discourse conventions are those of the target culture, just like lexicogrammatical standards. Native-speaker authority and superior expertise are axiomatic for foreign-language learners, and native preferences at all levels of language use are to be emulated for improved proficiency. Clearly, classroom tasks can require learners to attend to comprehensibility. The difference here is one of emphasis: for users it is a constant determiner of behaviour, for learners it may be desirable but not vital, since the classroom safety net will prevent major disasters ensuing from communication breakdown.
The consequences of these differences are reflected in appropriately different research design and data compilation principles for learner and ELF studies, as will be discussed further in Chapter 3. Here it is enough to point out that the two fundamental differences relate to speaker proficiency and participants’ native language. SLA data rests on controlling for learner proficiency as carefully as possible, whereas for an ELF study this would be an untenable solution, because it is in the nature of lingua franca communication that speakers’ proficiencies are variable. For learner research, first-language backgrounds are also usually kept separate, and each represented in comparable proportions. In contrast, realistic data for lingua franca research should include first languages in different mixes.
Learners are primarily identified by reference to their role in the education system. ELF speakers must be identified with respect to a definition of a lingua franca. This book adopts a general definition compatible with the traditional concept of a lingua franca as a vehicular language used by speakers who do not share a first lang-uage. This is a situation-based definition, and speakers may take on the identity of a lingua franca speaker in response to situational demands, irrespectively of whether in other situations they approach the same language as a target language to be learned. However, on a more stable basis, the prototypical ELF speaker is one who uses English habitually without either being a native speaker or a learner.
Our definition excludes contexts where English is the object of study, therefore English courses and CLIL (Content and Language Integrated Learning) are ruled out. What about the native speaker? Can we speak of ELF with ENL speakers present? Since ENL speakers do not share a first language with those who speak English as an additional language, they fit the definition.1 However, in interactions where native speakers are the majority in a group situation, language use tends to be determined in ENL terms. Although there is no principled theoretical basis for rejecting such instances, they are of marginal interest in investigating ELF. Native-/non-native dyadic interaction is the limiting case: it is a linguistically special, asymmetric situation, where the interlocutors have a different relationship to the language. It has been studied extensively in SLA, even if not much in authentic environments, and its special features render it relatively uninteresting as ELF interaction. Prototypical cases are more important than borderline instances for charting a new research domain.
ELF was contrasted to learners along three dimensions: the macrosocial, the microsocial, and the cognitive. This three-way perspective is adopted as the general approach in this book. The rationale for adopting a three-pronged framework is to contextualise ELF in the field of language study. A research domain that is primarily defined by its object of study necessarily touches upon many traditions. Any one of them can be taken as the point of departure, but the need to integrate more than one is perhaps strongest when a new territory is being explored. The perspectives chosen here have different relations to the data at hand, as will be discussed more thoroughly in Chapter 2. The macrosocial view provides a general backdrop to the whole and takes note of the tendencies in the overall quantitative findings, the cognitive perspective can be inferred from close observation of some of the data, and the interactional viewpoint is involved in the analyses of samples of discourse.
Research on ELF is still in its early stages. A few general introductions to English in the world mention it (e.g. Graddoll 1997, 2006; Kirkpatrick 2007) but do not go into detail. Change in research activity has been fast over the last few years, and a dramatic difference is seen between the time of writing this and the turn of the millennium, when ELF could boast only a handful of mainly small-scale exploratory studies (e.g. Knapp 1987; Firth 1996; Firth & Wagner 1997; Meierkord 1998, 2000). A landmark of change was Jenkins’s (2000) extensive study of phonology in ELF. This book inspired more research into ELF, along with fuelling the ongoing debates around the ownership of English and English Language Teaching (ELT). The debate was much shaped by disputes between Quirk (1985) and Kachru (1985), as well as an influential paper by Widdowson (1994), which stated the case for international users of English and the need to consider their English as not merely an imperfect form of ENL but language use in its own right. For the most part, the still ongoing argument around teaching and ELF generates yet more debate rather than serious research.
This has not prevented major empirical research taking off, resulting in monographs investigating attitudes and ideologies (Jenkins 2007), general processes in ELF (Seidlhofer 2011), ELF in Asia (Kirkpatrick 2011), English in education (Smit 2010), in ‘outer circle’ and ‘expanding circle’ contacts (Guido 2008), and collections of research papers (e.g. Knapp & Meierkord 2002; Mauranen & Metsä-Ketelä 2006; Mauranen & Ranta 2009; Mauranen & Hynninen 2010), as well as a large number of individual articles in journals and collected volumes. Many new PhD studies have also been completed in different countries within the last few years (e.g. Lesznyák 2004; Dewey 2006; Cogo 2007; Björkman 2010; Mortensen 2010; Dröschler 2011; Pitzl 2011), and more are in progress. The doctoral studies have focused on pragmatic, interactional, and morphosyntactic features of ELF, laying a good foundation for an emerging field.
ELF studies have predominantly been concerned with spoken lang-uage. This is an interesting reversal of the usual order in linguistic research, where written modes have been analysed before spoken. It is undoubtedly speech that more readily adapts to change than written text, especially in its published form. Writing is more oriented to standard language and more conservative, and publications tend to go through editing cycles, which lose some of the original voice of the author. Things are changing fast on this front, though, even in the academy, where a wealth of online academic publishing, science blogging, and other new forms of reporting and discussing findings materialises almost by the day (see e.g. Myers 2010). Editing, traditionally performed by native speakers, is losing its hold on academic publication as a consequence. Web-based forms of study make further inroads to ELF in academia, and discussion groups as part of university courses with international students offer interesting ELF data (Karhukorpi 2006). The relations between spoken and written modes are thus changing, presumably manifesting themselves in writing above all, since writing is quite flexible in incorporating spoken features, while speaking seems to be more similar across registers (Biber 2009), as it is subject to conditions of real-time processing that determine its shape. Online discussion groups, chatting, etc. mix features of speech and writing. This hybrid mode has been much investigated, and while it is interesting as ELF, most of it falls outside the scope of this book.
One of the approaches used in ELF research over the last few years has been corpus linguistics, which is based on searching large electronic databases (see Chapter 3), although the first investigations were based on fairly small amounts of data and were qualitatively oriented. Many individual scholars have compiled their own databases, not always small, but a qualitative orientation still predominates in ELF studies. The completion of two large corpora used by a number of researchers enables quantitative studies, which can supplement qualitative analyses, as is common in corpus linguistics. The first large corpus to be finished was the one used for the analyses in this book, the English as a Lingua Franca in Academic Settings (ELFA) corpus (www.helsinki.fi/elfa), which is described more fully in Chapter 3, and soon after it the VOICE corpus (www.univie.ac.at/voice/index.php) was finished in Vienna. More large corpora are being compiled in different parts of the world, including Asia (the ACE Project).