The workshop resumed after the vacation this evening with our postponed discussion of Perry's 'The Essential Indexical' and Shoemaker's 'Self-Reference and Self-Awareness'. Nic Koziolek provided very helpful summaries of both articles.
Perry wants to account for the psychological change he undergoes when he realizes that the shopper making the mess is he himself. Indeed, there seems to be an important difference even between "JP is making a mess" and "I am making a mess". But according to "the doctrine of propositions", these sentences express the same proposition, and if beliefs have propositions as their objects, we can't explain the psychological change as a change of belief. We wondered why Perry felt it necessary to give up on "the doctrine of propositions" so easily -- he gives up on trying to distinguish between distinct propositions he would express with "JP is making a mess" and "I am making a mess", ultimately in favour of his distinction between objects of belief (the same for those two utterances) and "belief states", which track the differences connected to action, etc. He recognizes that the "missing conceptual ingredient" can't be captured in non-indexical descriptive terms, but what's wrong with supposing that the missing ingredient is indexical, and that the context of utterance partially determines the proposition? David objected, against Perry's "belief state" view, that if Perry believes that JP is making a mess, and them remembers that he himself is JP--and thus believes that he is making a mess--his belief (what he believes) doesn't change, just his belief state (the way in which he believes what he believes). But this seems odd, and contrary to the goal of the paper as it's set out on the first page. Jason pointed out that if Perry believes he's making a mess but doesn't believe JP is making a mess, then he believes both that JP is making a mess and that he's not, even if his belief states are consistent. But this seems problematic.
We wondered about Shoemaker's argument for the claim that the use of 'I' "as subject" has a certain kind of priority; the argument (about getting from ascribing M-predicates to oneself to having the capacity to self-ascribe P*-predicates) doesn't seem to work unless more is built in to the capacity to ascribe M-predicates in the first place. David suggested that this argument was a proto-version of an argument Shoemaker has made on several subsequent occasions against the conceivability of "self-blindness", and which appeals to considerations pertaining to being rational. Jason worried about Shoemaker's account of the difference between immunity to error through misidentification relative to the first person pronoun and the kind of IEM there is in demonstrative reference. The upshot of the discussion was that Shoemaker's definition of IEM relative to the first person pronoun doesn't work. Here's David's example: if David is crazy and thinks he's Lincoln, he might be believe "I was a great president".
The IEM stuff was interesting but difficult. So we agreed to read Jim Pryor's paper on it for next time, Wednesday Jan 28th. Jacob Swenson will present. See you then!
Comments, criticism, and further discussion encouraged!