ABSTRACT
Sociologists and social anthropologists build their understandings of the social world primarily upon data that takes the form of language, yet the nature and status of that language is infrequently considered. Research methods such as interviews, focus groups, textual study, participant observation and surveys generate data that is primarily linguistic, in the form of what research subjects say or write.1 Even quantitative data is linguistic in that it is generated through research subjects responding to written or spoken questions by selecting the answer that best fits them. New methods have developed that are oriented towards non-linguistic data – such as in the study of images, music, material objects and the built environment – but these tend to be interpreted through a consideration of how people talk about them. While this is a correct approach, since it guards against the sort of de-contextualized analysis prevalent in some areas of cultural studies,2 it leaves us with the fact that sociological data is linguistic.3 A more promising route out of this reliance upon linguistic data is a thorough study of bodily gestures and of the positions and interactions between bodies more generally.4 Even so, much interpretation of uses of the body is shaped through the linguistic accounts of those observed or who are present in the same situation, or the accounts of those who are teaching others how to act appropriately. Given the intractability of having to rely primarily upon linguistic data, it is essential to consider the role that language plays for those we study. It certainly cannot be assumed that how people speak or write in research settings is an absolute, once-for-all representation of how they perceive or interpret themselves and their world. Language is far more particular than that – it needs to be seen in terms of social context and practice. While ethnographers, by generating or recording linguistic data in situ, are well placed to recognize the contextualized nature of language, the danger of not recognizing that is sharper when other research methods are used (especially without extensive prior contact with research subjects). This is because methods such as interviews take place in an artificial social setting which may mistakenly suggest that research subjects’ words are detached,
abstract and thus somehow more objective (and true). These considerations require that sociologists reflexively consider the social role of language with the aim of enabling a better understanding of the status of research data and hence the use they make of it.
