Propositions and States of Affairs - II

I'm blogging again on this issue of the relation between propositions ("props") and states of affairs ("sofas") because I'm not quite sure I'm got a firm handle on things yet. My understanding of these matters is still somewhat shifting and uncertain. So let's take a fresh look at things from a different angle.

Consider a prop, P. We regularly say things of the form "S asserts / knows / believes / hopes / doubts / fears / etc. that P (is true)." The different verbs (knows, believes, etc.) are commonly said to denote different "attitudes" toward the prop, P. Hence, these are called "propositional attitudes."

Now consider a sofa, X. We regularly say things of the form "S asserts / knows / believes / hopes / doubts / fears / etc. that X (obtains / is actual)." The different verbs (knows, believes, etc.) may be said to denote different "attitudes" toward the sofa, X. Hence, these may be called "statal attitudes."

But is there any real difference between these two modes of discourse? It seems that we can easily translate prop-talk into sofa-talk and vice-versa. So why not just eliminate one of them? It seems gratuitous to posit both props and sofas - where the latter are understood in the Chisholmian (abstract) not the Armstrongian (concrete) sense.

Something like the foregoing line of thought is what motivates Fumerton's argument that

with an ontology of such possibilia there is no need for propositions in addition to states of affairs. We ... could simply identify propositions with states of affairs and analyze truth as "obtaining." (Realism and the Correspondence Theory of Truth, p. 40)

What's key here is the identification of both props and sofas with the common content of the that-clause that is the object of the different "attitudes." And there seems to be something right about this. After all, there clearly is some common content to, say, "S asserts that P" and "S believes that P", and we can only isolate that by abstracting from S's "attitude" toward P. On this account, then, props are attitudinally neutral.

In contrast, I have proposed that props be thought of as intrinsically assertoric in nature. In other words, they are not additudinally neutral. Why think that? Well, my main reasons have to do with truth-conditions and the meaning of assertions. Assertions posit something about the world, and it is this feature that makes them suitable truth-bearers. For example, consider the string of words "Caesar crossed the Rubicon." Does this express something that has a truth-value? Maybe. If the words express the claim that Caesar crossed the Rubicon then they posit something definite, namely, the world's being such that Caesar has crossed the Rubicon, and the claim is true iff the world is as claimed. However, imagine that those words are uttered as part of a query ("Caesar crossed the Rubicon?"). This has no truth-value because it doesn't assert anything to be the case about the world. Similarly, if those words were uttered by an actor in a play who has no idea whether Caesar even existed or whether he is just a fictional invention of the playwrite. Again, the utterance asserts nothing, makes no claim about the world. Hence, it does not have a truth-value.

In general, a given sentence-type is capable of bearing a truth-value only if it is capable of expressing an assertion (i.e., capable of expressing a prop). Furthermore, a given sentence-token bears a truth-value only if it is actually employed to make an assertion (i.e., express a prop).

But if props are fundamentally assertoric, then we need something besides a prop to serve as the common content of the various attitudinal that-clauses. That's where abstract sofas may come in handy, since sofas are attitudinally neutral. Props posit sofas. Sofas posit nothing.

Perhaps, but I'm still not certain I've got this right. More tomorrow...