Thursday, March 16, 2006

Propositions and States of Affairs

I like blogging because it gives me a medium to "test drive" ideas and arguments. One topic that I've been mulling over of late has been the relation between propositions ("props") and states of affairs ("sofas"). My working theory over the past couple years has looked something like this:
  1. Props are assertoric units of meaning that we express by means of "statements", i.e., declarative sentences.
  2. Sofas are conceivably instantiable situations typically expressed by a gerundive noun phrase.
  3. Props posit sofas. For example, the statement "Fluffy the cat is on the mat" expresses the prop Fluffy the cat is on the mat, which posits the sofa Fluffy the cat's being on the mat.
  4. A prop is true if and only if the sofa it posits obtains (is actual). Thus, Fluffy the cat is on the mat is true iff Fluffy the cat's being on the mat obtains, such that Fluffy the cat exists and really is on the mat and not elsewhere.
So far matters seem pretty clear to me, however, issues start arising once we try to figure out how to square 1-4 with the popular view that takes sofas to be actually instantiated situations (a la David Armstrong). On this view of sofas, it is redundant to say that a sofa obtains, since all sofas do so by definition. Those who take this view may charge that I'm introducing needless complications. I need three things in my ontology: props, conceivably instantiable sofas (a type of possibilia), and actually instantiated sofas. Armstrongians seem to need only two: props and sofas. Thus, they will say that a prop is true iff it corresponds to a sofa.

Richard Fumerton expresses this concern in his book Realism and the Correspondence Theory of Truth:
If ... we had in our ontology states of affairs, some of which obtain and some of which do not, then we could let all propositions represent states of affairs - true propositions represent states of affairs that obtain; false propositions represent states of affairs that fail to obtain.... However, with an ontology of such possibilia there is no need for propositions in addition to states of affairs. We would have available a much more straightforward and elegant version of realism. One could simply identify propositions with states of affairs and analyze truth as "obtaining." ... But while [that] view is, perhaps, dialectically attractive, the metaphysical cost is prohibitive. In any event, it is surely desirable to characterize the essence of a correspondence theory in such a way that it is not committed to possibilia. It is desirable because there aren't any such things as states of affairs that do not obtain, and it is desirable because even if there were, it would be a mistake to suppose that the plausibility of the correspondence theory stands or falls on the possibility of defending such a problematic metaphysical claim. (p. 40)
I presume Fumerton's confident that there are no non-obtaining sofas stems from the popular Fregean theory of 'existence'. The Fregean conflates existence-as-actuality with existence-as-class-nonemptiness; hence, to say that there are non-obtaining sofas looks like affirming a contradiction - there exist sofas that do not exist. I've suggested in another post, however, that the Fregean theory might well be wrong. So for my purposes the objection that counts is Fumerton's charge that on my view "one could simply identify propositions with states of affairs and analyze truth as 'obtaining'."

I'm still trying to formulate a clear response to this kind of objection. Here's a rough outline of what I've got so far. Props are assertoric units of meaning; hence, they point to something beyond themselves; they posit something. Sofas, on the other hand, are not assertoric at all. They don't point to anything else. They posit nothing. Hence, only props and not sofas are suited to be truthbearers - they are true iff what they posit obtains. Fumerton's version of the correspondence theory abstracts from the assertoric character of a prop and by doing so deprives him from being able to equate a prop with the meaning of a statement (or assertion).

The confusion I think Fumerton is guilty of (and it is rampant in analytic philosophy) is the same one exemplified by Fregean theories of existence. Peirce calls it 'nominalism', a failure to appreciate 'thirdness as thirdness' by objectifying thought ('thirdness as secondness'). Lonergan calls it 'conceptualism', a conflation of understanding with judgment; a conflation of the meaning of a merely entertained thought (S's being P) with the meaning of an assertion (S is P). Polanyi would call it a failure to appreciate the tacit, personal dimension of all human thought and language. If those thinkers are right, then semantics must not be divorced from pragmatics. Our speech acts don't just color the meaning of our speech, they fundamentally transform it.


At 3/16/2006 6:20 PM, Blogger Tom said...


Nice to have you back. Sounds like you had a relaxing time. We had our share of snow this week here, enough to close schools in our area (which almost never happens in MN).

I like your 1-4. I certainly agree that truth is an attribute of props and not of sofas. I also find it very helpful to define sofas as you do—conceivably instantiable situations. It seems unhelpful (even objectionable) to limit sofas to actually instantiated situations. We certain can and do conceive of instantiable situations that are not actual. What are we to call these if not sofas? All one needs to do is qualify herself and make clear whether or not the particular sofa she’s talking about is actual or non-actual. So an adequate ontology needs the three you mention I think.

The issue I’m not quite on board with is your 4, your saying that a prop is true iff “what it posits” (viz., the sofa it posits) obtains. Not sure I’ll describe this well. But let me try. I’m more inclined to say a prop is true iff there is some sofa that grounds the truth of the prop, or some sofa that could not obtain without the prop’s being true [see Crisp's truth axiom below]. This is slightly different than saying the prop is true iff “what it posits” obtains (or at least “what it posits” doesn’t appear to me to do the job in some cases). I’m thinking of past-tense props for example: “Caesar crossed the Rubbicon.” What’s this posit? If you answer by saying it posits a particular memory in God, then you’ll be expressing my view. And I think (given your paper) that you’ll agree to this, since you elsewhere argue that a past-tense prop is true iff “God remembers the past” in the required way (e.g., God remembers the past in the way the prop describes that past). But then I’m confused as you how the words of prop assert something about God. Some will certainly wonder where God is in “Caesar crossed the Rubbicon.” They seem to assert something about the history of our word (that it included the event in question) and nothing about God.

What I’m suggesting is that it may not be such a simple thing to deduce a prop’s asserted sofa from the words of the prop, so that saying a prop is true iff “what it posits” obtains may not be an adequate explanation. The sofa that grounds a prop’s truth may not show up in the actual words that form the prop’s assertion and constitute its meaning. So now we’re distinguishing between the sofa described by the prop (non-actual in the case of past- and future-tense props) and that actual sofa that grounds the truth-value of the prop. I said some time ago that for SOME props there are at least TWO distinct sofas at play—one observable in the words of the prop or described by the prop, and another that appears nowhere in the words of the prop but which we must posit to account for its truth. My point: a prop’s truth-grounds is that actual sofa that obtains at the time the prop is asserted and which may or may not appear in or be described by the actual words of the prop. We have to be more specific about what we mean when we say, in 4, that a prop is true iff “what it posits” obtains. Most are going to look in the actual words and sofas describe by the words for “what a prop posits,” in which case they’ll come up empty handed. The more important sofa may nowhere appear in the words of the prop. “Caesar crossed the Rubbicon” is true iff, as you argue elsehwere, the sofa “God remembers Caesar’s crossing the Rubbicon” obtains in God. This sofa is posited by the prop, though not described by the prop.

We either have to go with your (4) [a prop is true iff “what it posits obtains”] and then develop a strategy for getting from the sofa described by “Caesar crossed the Rubbicon” (namely, Caesars’ crossing the Rubbicon) to that sofa not described by the prop but which is the one that’s got to obain if the prop is to be true (namely, God’s remembering that Caesar crossed the Rubbicon), OR just redefine (4) to read “a prop is true iff there is some actual sofa that accounts for its being true” (or however you prefer to word it), which is what Crisp does:

Ap(p --> Ex(x --> p)) [Invert A and E to get the operators; couldn't figure out how to insert them in html],

In other words:

Necessarily, for any proposition p, if p is true, then there exists some condition x, such that necessarily if x obtains then p is true.


At 3/17/2006 4:05 PM, Blogger Alan Rhoda said...

Hi Tom, thanks for your comments. I agree with you that "it may not be such a simple thing to deduce a prop’s asserted sofa from the words of the prop....The sofa that grounds a prop’s truth may not show up in the actual words that form the prop’s assertion and constitute its meaning." Here I make a distinction between the phrases we used to denote sofas and sofas themselves. The sofa posited by "Caesar crossed the Rubicon" may be appropriately denoted by the phrase "Caesar's having crossed the Rubicon", but that tells us little to nothing about what it is for that sofa to obtain. And here different systems of metaphysics give different accounts. I think that sofa consists in a divine memory. Crisp thinks it consists in a primitive 'earlier than' relation holding between two abstract 'times'. Etc.

By the way, to represent the logical symbols for ∀ and ∃, simply type an ampersand (&) + "forall;" and "exist", respectively.


Post a Comment

Links to this post:

Create a Link

<< Home