ABSTRACT
In this chapter we pose a number of challenges to and offer alternative formulations of the concept of identity within the framework of structuration theory. Our intention is to develop a positive critique of the concepts of reflexivity and interaction through an assessment of the theoretical role of language use in Giddens' ontology of modern social systems. Our aim, in part, is to take up the assessment of structuration theory where other critics have left off (notably Thompson 1989, Bauman 1989) and to analyse inter action as an 'ethnopolitical' as much as an ethnomethodological phenom enon. We suggest that Giddens' treatment of language, interaction and social practice is both innovative and perplexing at the same time. It is innovative in shifting sociology's approach to social reproduction from (Habermasian) system pragmatics to a concern with practice and in theorising the connec tions between agents' routine actions and institutional patterns as dynamic rather than systemic. It is perplexing because the traditions of natural lan guage philosophy on which Giddens draws do not provide the necessary sociological bridges between agents' and institutional knowledges, resources and practices. Our discussion of language, identity and institutions will take this claim as its starting point because it is in the terms of these traditions that Giddens' ontology of recursive social systems is formulated. Respond ing to Thompson's critical questioning of this philosophical dependence, Giddens writes:
In rejecting the formalism of the 'language' comparison, Giddens restates the centrality of an ethnomethodological approach to language use or 'talk' in his account of social systems. Although the qualifier 'in some ways' is introduced here, 'talk' or spoken language-in-use has always been the model of social practice in structuration theory, as we explain below. This qualifier is a further elaboration of his defence of the language metaphor in The Constitution of Society where Giddens (1984) acknowledges that people share rules of language in ways that they do not share rules of social action. We suggest, however, that the spirit of Thompson's critique remains important and that language-use, conceived as ethnomethodological 'talk', is less helpful in theorising the relationships between interaction, identity and institutions than structuration theory implies.
