language as a cultural emergent property
What's real about real language?
We start with this question for two reasons. Firstly, applied linguists are explicitly concerned with 'the real', although we shall argue that the concept of the real which concerns them warrants further elucidation. Secondly, the non-applied, non-social dimensions of language are often seen to be what constitute its 'reality' in the Chomskyan tradition, so we need to address that line of argument here as well. Indeed, we might suggest that a key contrast between linguistics and applied linguistics concerns their respective concerns with different kinds of 'reality'.
For Chomsky, the concept of language which locates it beyond the brain, and thus 'externalizes' it as 'a "construct" that is understood independently of the properties of the mind/brain' is 'too far removed from reality' (Chomsky 1986: 19, 20). This 'externalized' version of language, or 'E-Ianguage', is criticized for being epiphenomenal to 'the data and mechanisms' (Botha 1992: 84) of the internalized, '1 language', the proper concern of linguistic science. The I-language is 'some element of the mind of the person who knows the language, acquired by the learner, and used by the speaker-hearer'; 'the system of knowledge of language attained and internally represented in the mind/brain' (Chomsky 1986: 22, 24). This internal, unobservable, generic version of language is what is 'real' in this version of linguistics.
As an applied linguist, Brumfit (2001) demonstrates a different interpretation of what is meant by 'real', when he re-presents his definition of applied linguistics as 'the theoretical and empirical investigation of real-world problems in which language is a central issue' (p.169). Cook (2003) makes a similar link: applied linguistics is defined as 'the academic discipline concerned with the relation of knowledge about language to decision making in the real world' (p.5). Cook prioritizes 'the facts of language use' (ibid.: emphasis added), while Brumfit claims that '''real-world problems" are investigated partly as a counterweight to the disconnection of autonomous and idealized disciplines' (p.186), thus contrasting 'real' with 'idealized'.
When the word 'real' is used, as it is in several publications, to denote a characteristic of the language with which applied linguistics is concerned, it is clearly intended to invite approval (ef. Carter 1998). COBUILD, which uses the Bank of English as a source of data for its dictionaries and grammars, includes on its publications the strap-line 'Helping learners with real English'. The Longman book series entitled 'Real Language' describes itself as a 'sociolinguistic series about language in the real world' (e.g. Milroy and Milroy 1993: back cover), whose titles (one of which is Real English) 'draw on natural language data from a wide range of social contexts'.
So can we begin to answer the question 'what is real language?'? Two contrasting positions can perhaps be summarized thus:
1. Real language is an unobservable property of the human mind. Understanding it requires analytical skill and an ability to deduce generic and generalizable patterns, rather than observations of people communicating in any specific language variety in particular circumstances.
2. Real language is what people actually say. Understanding it involves collecting examples of this observable language behavior. Describing it means describing this behavior, rather than accounting for some hidden generic mechanism.
Of course, there are many shades of commitment to and modified versions of these contrasting positions, which to some extent echo those held in different traditions of linguistics through the twentieth century. It is possible that misunderstandings based on the connotations of each are responsible for some of the divisions between 'autonomous' and 'applied' linguists today, and for the failure of each to recognize potential common ground between them. Our own position distinguishes us both from the 'ontologically driven' rationalist version of realism which we believe characterizes adherence to hidden generative mechanisms and a derogation of empirical data, and also from the naIve empiricism which privileges observable behavior over theory. We accept that some aspects of the conceptualization of language may depend on the purposes of the person who is adopting it. It is thus quite conceivable that a literary critic, a translator, a curriculum designer, a teacher, a speech therapist and a neuroscientist (not to mention all of them- and us- in our roles as routine language users) might hold very different concepts of 'language' without anyone of these being 'wrong'.
Lest we be accused of side-stepping theory and taking refuge in pragmatism, however (d. Carr 1990: 71ff), we should make it clear that our position here is not that 'anything goes'. Theory plays a part in conceptions of language, and language itself has properties which limit the ways in which it can legitimately be described. Those who analyze language, whether for 'pure'-ly descriptive or for 'applied' (e.g. pedagogic) purposes, will necessarily have a different understanding of it from those who use it as a medium to get things done. This means that we can rule out the exclusively hermeneutic position, summarized by Pateman (1987) as the requirement for a theory- in this case, a theory about language- that it 'could in principle be recognized as true by those to whom it applies' (p.12). But it also means that we need to add a third conceptualization of 'real language' to the two already identified:
3. Real language is what ordinary speakers produce and encounter as they go about their business. In the normal course of existence, language users only intermittently have the need or motivation to analyze the properties and patterns of language, and are likely when they do to turn to 'authorities' such as dictionaries (to establish what words 'really' mean).
This means that there is a potentially greater contrast between what the lay person perceives as significant and what the analyst chooses to highlight, than between the accounts provided by different kinds of analyst. There is a potential paradox here. The 'real' concerns of the 'non-scientist, non-specialist' may turn out to be incompatible with the characteristics of 'real' language as identified by those who analyze it (as Cameron (1995), among others, has pointed out). Or, as Brumfit (2001: 153) puts it, 'Teachers work in a world of real people, real motives, and conflicting interests' so will often see things differently from researchers, whose job is to be reflective about language in particular and explicit ways. Applied linguists thus have to ask themselves whether their version of 'real language' can accommodate position 3 above, including, for example, commonsense definitions of who in the world speaks 'real English', or prescriptivist attitudes to 'correct' language use.
This is one way in which the applied linguistic enterprise of engaging with 'real language' in 'the real world' invites further attention, and it leads us to another one. Even the most practical of applied linguists, whose principal concerns are with helping language learners to make more successful progress in their studies, for example, have to make use of some theoretical constructs in conceptualizing language. As Brumfit (2001: 178) points out, 'any theoretical and empirical study involves principled idealization'. In other words, we are arguing that no applied linguist (when being an applied linguist, that is, and thus, by our definition, a social scientist) can take 'real language' as given and unproblematic. Some theorizing and analysis inevitably goes with the territory. This is not to denigrate either the practical abilities or reflective capacities of 'lay' speakers; it is simply to acknowledge that using language and researching language use are not identical activities.
