Michael E Smith provides a link in a comment to a paper of his, on the uses of urban theory in archaeological research on ancient cities – his argument is that this work serves as middle-range theory in contrast to grand theory of the sort developed by Latour, or Giddens, or Bourdieu. I like the idea that theories are always best when they are marked by a certain sort of empirical modesty – although it’s interesting that the sorts of social theory that Smith thinks of as ‘grand’ in his field would in geography these days appear to be much more ‘middling’ than the grandly philosophical styles associated with current work on spatial ontologies, affect, events, the post-political, and related themes.