Methodology week is now over: here are some photos. The conference was a success, with a number of interesting talks. Not all of the speakers put methodology front and center, but there were a lot of different approaches to philosophical methodology presented explicitly and implicitly. A recurring theme, present one way or another in all the papers, was the role of a priori methods in philosophy: their strengths and limitations, and their continuity or otherwise with empirical methods. In effect, two of the speakers challenged a priori methods, two of the speakers defended them, and two of the speakers questioned their distinctiveness.
Challenges came from Sally Haslanger and Steve Yablo. Sally considered pressure from the social and political spheres, arguing that a priori conceptual analysis can't do justice to concepts such as "race" and "parents", because their crucial social and normative role is not transparent to a priori reflection. Steve considered pressure from modal epistemology, arguing that a priori modal intuition can go wrong in ways that haven't yet been adequately explained e.g. by Kripke or by two-dimensionalists. The advocates of conceptual analysis and two-dimensionalism present argued in response that an appropriately broad and flexible use of these methods can accommodate all of the relevant phenomena.
Defenses came from Martin Davies and Frank Jackson. Martin considered the worry that a priori methods will "prove too much", allowing a priori arguments for apparently empirical conclusions, replying (in what was perhaps a slightly backhanded "defense" of a priori methods) that considerations about warrant transmission will prevent a priori premises from generating such powerful conclusions. Frank suggested that our ability to know what we mean with a word, combined with the "transparency" of language in referring to properties, will ground an ability to know what property a predicate ascribes, and went on, controversially, to use this method to argue against the "Australian" view (and his own prior view) that color predicates ascribe physical properties such as reflectance properties. The transparency claim led to a nice clear contrast with Sally's view, and led to some useful discussion of just what sort of transparency of language one can reasonably expect.
Doubts about distinctiveness came from Tim Williamson and Peter Godfrey-Smith. Tim argued that philosophical "intuitions" such as the Gettier thought-experiment rely on counterfactual reasoning, and that this sort of reasoning is continuous with counterfactual reasoning in empirical domains and doesn't require a distinctive sort of explanation. Peter applied ideas from the philosophy of science to the methodology of metaphysics, in effect arguing that a priori methods in metaphysics can be seen as model-based reasoning on a par with model-based reasoning in science, and evaluable by similar standards. In both cases, the moral could be taken as either supporting or as deflating the methods, depending on one's orientation.
In the end, I think everyone agreed that we should be pluralists about philosophical methodology: pretty obviously, there is no single method that's common to all good philosophical reasoning and that distinguishes it from inquiry in other domains. With that granted, the interesting questions concern the strength and limitations of various specific methods: what's the role for intuitions, for conceptual analysis, for formalization, for empirical observation, and so on. These questions are unlikely to be decisively settled anytime soon, so in the meantime, let a thousand flowers bloom.