On Mon, 31 May 2004, Carrol Cox wrote:
> The limitations of communication in e-mail posts. Human agency is there,
> and there is no question of us being pawns of some mechanism. But it
> would take me about 18 months and a final draft of 30 to 150 pages to
> work it out.
I think this agency/structure dichotomy is a misleading way to analyze this sort of thing. I like the biological analogy: organisms have systemic, emergent features that cannot be reduced to the separate parts that make up the organism (one of Richard Lewontin's fav themes). Thus simply understanding (say) gene mechanisms is not adequate to understand the development of complex organisms.
--And just so with complex social systems: the features of those systems are not simply an aggregation of the features of the individuals who participate in and create the social system. Social systems, like organisms, have emergent features as a result of their complex structures; as the cliche goes, the whole is greater than the sum of the parts.
Seen this way, the dichotomy between agency and social structure falls apart: people actively create social structure, and a society is in turn is a higher level "agent" whose effects can be assessed and analyzed. (The old canard about how "society doesn't do things-- people do" is about as silly as the claim that "organisms don't do things--genes/cell do". There are different levels of analysis here, and given the emergent features of social systems, it's important to emphasize the real, agentic impact of social systems above and beyond the thoughts and activities of any specific individuals.)
I've just read Gould's Structure of evolutionary theory, and his emphasis on hierarchical analysis appeals to me: as Darwin argues, individual selection can cause evolutionary change; however, it is also true that species selection (and other higher levels of entities) can lead to evolutionary change as well. --Thus the evolutionary agent is not simply the individual organisms or the individual genes. I think applying this kind of hierarchical reasoning to the social world is useful: there are different types of agents working at different levels, with various effects on agents at each level.
Miles