|Graduate Seminar: Content without structure
|234 Moses Hall
Our standard ways of representing and attributing mental states and speech acts involves two components, which Frege called force and content. Believing that snow is white and imagining that snow is white are attitudes that share a content but differ in force; similarly, believing that snow is white and believing that mud is red share a force but differ in content. This approach gives us the efficiencies of a division of labor. A single account of content can be combined with accounts of various kinds of forces to yield accounts of various kinds of attitudes and speech acts.
We will focus on the question: What should we take contents to be, given their roles in our psychological and linguistic theories? The classic accounts given by Frege and Russell take contents to be structured complexes of senses (on the Fregean view) or objects and properties (on the Russellian view). We will be primarily interested, however, in unstructured conceptions, which take contents to be functions from circumstances, possibilities, or conditions to truth values.
We will begin by looking at Stalnaker’s development of such a conception in Inquiry. We will ask what motivates Stalnaker’s conception of content, as against structured alternatives, and what explanatory work it does. We will then look at several objections to unstructured conceptions:
The problem of logical omniscience. Intuitively, one need not believe all of the necessary consequences of one’s beliefs, but the unstructured conception seems to entail that one does. Relatedly, there are plausible psychological generalizations – for example, that if one believes a conjunction, one must believe both conjuncts, but need not believe all of the necessary consequences of the conjunction – that it seems cannot be stated in a framework that takes contents to be unstructured.
Aboutness. Intuitively, beliefs are individuated in part by what they are about (by their topics or subject matters), but it is not clear how a notion of aboutness can be defined that makes sense for unstructured propositions.
Frege’s puzzle. Intuitively, believing that Hesperus is visible is not the same as believing that Phosphorus is visible, but on an unstructured conception, the contents would seem to be the same.
De se attitudes. Intuitively, I could be omniscient, in the sense of knowing which world is actual, without knowing which object in that world I am. So standard unstructured views do not seem to provide resources for describing de se attitudes.
Although some of these objections have been taken to motivate structured theories of contents, we will be particularly interested in seeing how unstructured theories can evolve to meet them. In addition to classic treatments of these problems, we will look at some very recent (and in some cases, future) work on them.