The plane ride there was long and super bumpy (and I hate flying!!) and then I got strep throat and the plane ride back was a red eye that got into JFK at six a.m. (and I REALLY hate flying!!!!)…but other than that California was fantastic!
The APA was fun, though I got there on the last day of the conference and since I wasn’t feeling well (I was chaining Sucrets one after the other) I left after my talk. But I did see the session before mine, by Hanna Kim, on a proposed compositional semantics for metaphors which was interesting. She sketched an account that borrowed Jason Stanely’s idea of a hidden unarticulated variable that was context sensitive to metaphorical meaning. This would allow one to get the meaning of the metaphor in a way that was completely determined by the meaning of the parts (including the hidden, context sensitive variable). Marga Reimer responded with a couple of objections. One of which was the Gricean kind of objection one would expect. She invoked Grice’s modified Occam’s razor and asked why we need a semantic account of metaphor’s when we have a perfectly good account from Grice that appeals to speaker’s intentions and doesn’t posit all of these weird hidden variables? (Here! Here!) Kim’s answer, in part, was to point out that Grice’s account cannot take care of ‘impossible metaphors”. The basic idea behind impossible metaphors is that there are semantic and syntactic constraints on what kinds of sentences we can make metaphors from. I don’t recall any of her examples and I can’t find the handout…but still, I wonder about this kind of strategy. Why is an objection to Gricean theories to point out that sentence construction is constrained by syntax? A speaker is constrained by what she can reasonably assume will alert a hearer to her communicative intention and thereby fulfil that very intention. The syntax of a language is definitely one thing that would suggest itself as something which would constrain which utterances a speaker can reasonably expect a hearer to successfully infer what one is communicating. No problem.
My talk went well, I think. We had some interesting discussion. The commentator (Imogen Dickie) posed a dilemma for me. If we can have rigid designation in thought then either the problem of necessary existence reoccurs at that level and we haven’t solved the problem or we can have rigid designation without the problem of necessary existence (in thought) and so we shouldn’t be worried about it in language. This is especially pressing when we think that S5 is attractive because it is supposed to be a logic for thought. I responded that the problem of necessary existence is only a problem when we try to regiment our thoughts into a formal language. There is no problem with having a singular thought about Socrates, the problem is trying to formalize a sentence representing that thought. This is the evidence that we have that we need an separate account of the semantics of language. But S5 is still a logic of modal thought because we can formulate descriptions in it that ‘single out’ the object of thought without rigid designators. The absence of singular terms in our logic is nothing more than an inconvenience. She also mentioned, in passing, that Williamson thinks that necessary existence is not as terrible as one might think. One might argue that I exist in all possible worlds but in some worlds I exist without any properties. This was quite shocking to me, as I can’t really fathom what that would mean. Really, what does that mean? Anyone know?
From the audience I was asked several good questions. One was from Tim Lewis on how I felt about the fact that names on my account would fail the Church translation test. That is, we expect that ‘Richard’ and ‘Ricardo’ to be synonyms but if the really stand for ‘the bearer of “Richard”‘ and ‘The bearer of “Ricardo”‘ then pretty clearly they aren’t synonyms since they each have a separate quoted name in them. I thought that was a pretty nice objection. At the time I said that I would argue that names are not part of a language. So, in a complete dictionary of English there would be no ‘Richard’ or ‘Doug’ (forget about the dictionaries around now, they are half encyclopedia, I am talking about just a list of the words of a language and their conventional meanings, pronunciation guide, and syntactical/grammatical categories. That seems right to me, but then on the plane home, in a half trance, I started to think that maybe we could use Seller’s notion of ‘dot quotes’ to solve the problem if people don’t like the position on names. So instead of ‘the bearer of “Richard”‘ we could have ‘The bearer of *Richard*’ where ‘*P*’ is ‘dot-quote P’ and basically serves to single out all of the functional types that play the role that ‘Richard’ does in English. This would allow one to preserve the intuition that other language cognates of English names are synonyms. Or so it seemed on the plane…and besides I like the bit about names not being part of the language…
The other question that I remember was from Adam Sennet (there were a couple of others that I am forgetting). He echoed Williamson’s point that since we know quite well what a rigid designator is and how one would introduce them into a formal langauage it is then quite odd to say that there aren’t any. I responded that we know what it would be like for there to be all kinds of things that don’t exist. I know what it be like for there to be square circles (it would be for there to be one object that is both square anc circular at the same time), but that doesn’t mean that there are any. This is exactly what one would expect. We know what it would be like for there to be flogisten or tachyons or any other theoretical posit we come up with. It would be like finding the thing that we posited, but someimes we find out that they don’t exist. Interpreting that syntactical category proper noun as a rigid designator is a natural attempt at capturing what it is that we do when we think about some particular thing but when we do model that category that way we get the problems with necessary existence, which means that it is a mistake to model it in that way. I compared it to what happens when we try to mix quantuum theory with relativity theory. When we try to calculate the probabilities for things which we have well worked out answers for we get crazy results (like the probability of some event occuring being infinite). This let’s us know that there is a problem and then you get all of the different answers to solve the problem. Our finding the proofs for necessary existence in S5 are like the infinite probabilities in physics; it is an indicator that something needs to be done.
This is, by the way, why I disagree with Chappell’s charge that logic is over rated and that, in particular, my
employed logical apparatus merely serves to build in misunderstandings. The formal steps of the argument may be flawless, but that’s all for naught if the entire argument is based on a mistake — due to failing to understand precisely what all those formalisms really mean.
I understand what the formalisms mean and I am using them to apply pressure to a person who holds a certain kind of view. The proofs count as evidence that some assumptions don’t work. This is exactly what formal logic is good for…though I do agree that one needs to also make the argument in prose as well as symbols.
OK, well that’s enough for now, I gotta get to work on my Tucson presentation and grade some exams!!!!!!!