A social realist approach to research in applied linguistics
Summary of key claims
The centrality of theory
Many approaches to research rest on an assumption that the social world is already constituted, and can be uncovered, or revealed, by the research process. From this perspective, it is in this process of un covering that theory has a role, and adjudication between theories is on the basis of how closely they correspond to an already constituted social reality. In contrast to this, we favor an approach which accords some relevance to the role of theory in the constitution of the social world. This means that our advocacy of theory is not of the sort whereby the researcher simply looks for empirical data to confirm a prior theory, since how we grasp the world theoretically will in important ways shape what is to count as empirical data. Moreover, as Layder has pointed out, '... if all fieldwork data is interpreted in terms of a prior framework favored by the researcher then it will lead to a blinkered outlook' (1993: 52).
The relationship between theory and observation in social research is a difficult one. In important ways our theories about the world favor certain observations and rule out others as the basis for confirming or disconfirming evidence; that is, what is to count as evidence is in part defined by the theoretical concepts we are employing. A greater role than is often recognized is therefore played by epistemology in the analysis of the social world (Layder 1990). This might seem to suggest a degree of indeterminacy which is inconsistent with our claim that the world is independent of our cognizing experience. However, we can both acknowledge an independently existing world and yet hold that the representation of it in theoretical terms is provisional and fallible, a matter, as we say, of discovery rather than revelation. There cannot then be a direct appeal to neutral observation as a means of 'proving' or 'disproving' a particular theory. This does not mean, of course, that all theories are equally accurate in their representations of the social world, nor that we cannot make judgements about their veridicality.
A further point to recognize is that the social world itself is partly constituted symbolically. Meanings cannot be observed directly, they have to be inferred from the observation of behavior, but such inference is always from a position within a theory or a set of theoretical concepts.
