Researching language learning: theories, evidence, claims
Evidence and ILL research: claims and warrants
Ethnographic evidence in ILL research
The method most typically contrasted with variables-based studies is ethnography: extended research involving interviews, observations and some participation by the researcher in the lives of those being studied. Again, it is neither feasible nor appropriate to review in detail here the characteristics of the ethnographic method, but it is relevant to our discussion to consider one of the claims often made for its epistemic authority. It is something of a truism that context-specific case studies, ethnographies and actors' own accounts receive criticism at the methodological level, failing to persuade skeptical readers because of the reliance 'on the vividness and logic of their descriptions and inferences as confirmation of the validity of the descriptions ... [which] can hardly be justified merely by asserting the "interpretive" or subjective nature of the qualitative endeavor' (Chaudron 1988: 23). As we have noted, dialogue on this issue is unlikely to be productive if the epistemological positions of the protagonists are in commensurable. Van Lier (1990: 42) identifies 'two basic principles that are generally agreed to underlie ethnographic enquiry'. One of these is 'a holistic treatment of cultural factors, or, in other words, a concern with context', a point which relates to the 'bundle' metaphor and the question of the size and nature of the unit of analysis, both of which we have discussed above. But it is the other core principle on which we focus now, namely 'an ernic viewpoint', glossed as 'the rules, concepts, beliefs and meanings of the people themselves, functioning within their own group' (p.43).
Paradoxically, there is a strand in 'qualitative' and 'interpretivist' research approaches which could lead to the replacement of one kind of empiricism with another, an approach we might label 'experiential empiricism'. By this we mean that there are potential dangers for research in relying too heavily on people's accounts of their experience. Quite apart from the technical difficulties inherent in accessing people's experience by means of recording what they say about that experience, we need to be clear how far a respondent's claim to 'be', say, 'a Spanish speaker' can illuminate our understanding of, say, the learning of Spanish among a particular social group. Now, some theorists claim that participants' discourses are the very stuff of social reality. Thus Shotter (1993: 21), for example, believes that ', .. new ways of talking "about" social relations, new forms of debate, work to constitute and to establish the very relations to which the words uttered within those ways seem to refer'. While there may be some truth in this, we find it unconvincing that social relations can be only the products of discourse, so that although, as Pawson recognizes, in the everyday world we all engage in the practice of 'get[ting] hold of a bit of the world and assign[ing] it a value' (Pawson 1989: 70), and although research must take account of 'the significance of, and the variation in, cultural meanings and natural discourse' (ibid.: 20), the social scientific interpretation of the world, and the classification of its constituents, demands some explicit theorizing. As Byrne puts it, '... classification is perhaps the most important way in which we can understand a complex nonlinear world' (2002: 5). To be sure, conventional analytic categories may be inadequate or inaccurate, as some of the critics we cite have claimed, but the use of participants' categories will not necessarily serve us any better.
This is why we view with some skepticism claims such as those made by Norton (2000: 58) in her presentation of her informants' accounts of their experiences as adult immigrants to Canada learning and using English there in differentiated social contexts. 'It is important to note', she says:
. , . that the stories these women have told are their stories. I have taken seriously their interpretations of events and their understanding of cultural practices in Canadian society. I have sought to understand the world as they have understood it, and have not questioned whether their interpretation of events was the correct or true interpretation.
Our point is not that researchers should reject participants' own accounts of their experiences, but there is no intrinsic reason why these accounts should be assumed to provide reliable information about, for example, the social structures which maybe constraining or facilitating language learning processes. Indeed, Norton herself provides some highly theorized sociological commentary on the experiences of her informants, in addition to the extensive extracts she reproduces from their diaries and interviews. This in itself, together with the fact that the extracts appear, edited, in an academic monograph, recontextualizes these 'authentic' accounts, and suggests that the analyst has, quite properly in our view, brought to bear on her interpretation of the topic being researched (adult immigrants' experiences of language learning) knowledge and understanding additional to that of the participants themselves.
This is not to argue that respondents' accounts are in some way 'inferior' but rather that they are inescapably partial, and this in two senses: the purpose of everyday understandings is practical adequacy in the context of daily life and not the practice of social scientific enquiry; respondent accounts, insofar as they remain grounded in this practical adequacy, also remain tied to a less objective view of the social world than that sought by the researcher. It is the very opacity of social life that requires us to develop theoretical accounts if we are to grasp those features of it that are not given in our everyday, phenomenal experiences. If the way the social world appeared to us was the way the social world was, then there would be no need for social science and no basis for claims to epistemic privilege (and far fewer human errors!). It is partly because the social world is not transparent in this way that we experience its force in the form of constraints- and its power in the form of enablements- and that we need to develop novel, frequently counter-intuitive, theories about what it consists of and how it works. As Brumfit (1997) acknowledges, theory-making is a social practice, to be sure; however, as he puts it, if 'the claim is made that theory-making is merely a social practice, with an implication that it can make no claims to privileged usefulness', then it is both unconvincing and somewhat disingenuous. Researchers lay claim to epistemic privilege even as they write to champion the perspectives of the researched. As Brumfit continues, '... publishing a book or an article is a claim to have ideas worthy of consideration and is precisely the same kind of claim that any "expert" may make' (p.93).
Rejecting the scientist's variable categories only to embrace the 'emic' categories used by those who participate in the processes we are researching can thus lead to a different set of epistemological problems, which suggests that interpretation of complex, nonlinear, emergent systems and processes cannot be achieved simply through thick description and holistic accounts.
