To: Stuart B.
From: Geoffrey Klempner
Subject: Essays on Santa Claus, proper names, and truth
Date: 12th August 2012 14:11

Dear Stuart,

Thank you for your email of 31 July, with your first three essays in response to questions from the University of London Logic paper.

Essay 1. 'Is there a satisfactory account of the truth of the sentence, 'Santa Claus does not exist?''

My first thought about this is, Why does the question ask for an 'account of the truth' rather than an 'account of the meaning' (i.e. in terms of truth conditions)? The examiner's thought here was that someone might be led to give a theory of meaning according to which the sentence can be false but cannot be true. So the question is whether an account of the meaning of this statement has to allow for the possibility of truth as well as falsehood.

Of course, we *say* things like, 'SC does not exist' but if you are sufficiently confident in your theory that 'SC does not exist' cannot be true, then you will have some way to explain this away, an 'error theory' of everyday discourse. Anyone who rejects any form of description theory of proper names has to bite the bullet and take this option, on the assumption that 'Santa Claus' is indeed being used (or, we are attempting to use it) as a proper name. 'No object, no thought.' The fact that you *think* you are expressing a thought doesn't entail that you are doing so.

So, I would object to your associating the Searleian theory you put forward with Kripke's view here. On Kripke's view, or any similar theory which rejects a descriptive account, empty reference = no meaning, no thought, no truth or falsity.

Contrary to what you say, Frege did not put forward the sense/ reference distinction 'as a response' to the problem of names with no referents. Frege did hold that the possible non-existence of a referent was a fault of natural language, and made no attempt to explain how a sentence can still have meaning (= possess truth conditions) in this case. (Russell does, however, use Frege in his 'On Denoting' as a contrasting theory to his own theory of descriptions. But this is Russell, not Frege.)

What you are confusing (and Russell deliberately conflates) with the sense/ reference distinction is Frege's account of second-order concepts, and of existence as a second-order concept, according to which 'SC does not exist' or 'unicorns do not exist' can be analysed in terms of the existential quantifier and the property '...is SC' or '...is a unicorn'.

However, for Frege (as Dummett argues) there is no necessary assumption that sense can be explained in conceptual terms. Gareth Evans says more about this in his 'Varieties of Reference'. Knowing where to find the hut in the forest is a perfectly good sense = mode of presentation which cannot be explained in terms of descriptions. Sense can be 'knowing how' as well as 'knowing that'.

(P.S. Don't mention a 'Randian' theory of this or that without at least stating what the theory is, or what you take it to be. The examiners have probably never read Ayn Rand, and more likely than not hold that she has not said anything worth reading on this or any other topic. But if you want to enlighten them, go ahead.)

Essay 2. 'Defend what you think is the best account of proper names.'

OK, so now we have Rand's theory of concepts. This looks to me like a version of Locke's distinction between 'nominal' and 'real' essence. For example, I may believe that gold is a 'yellow metal' but I am prepared to allow for the possibility that the yellow colour is due to impurities. My intention is to refer to that which integrates with our knowledge of the physics and chemistry of the element gold, knowledge which is continually expanding.

However, 'Aristotle' differs from 'gold' in that we are making a very specific assumption: the existence of a spatio-temporal particular. Suppose that there were two philosophers in ancient Greece who were responsible for the body of work which we associate with 'Aristotle', but that this fact has been lost over the course of time. So, 'There was a philosopher called Aristotle' is false (because it implies that there was one individual) yet 'Aristotle put forward a theory of four causes' is true (understanding 'Aristotle' to mean the Aristotle team, or partnership).

I don't see that it is enough to say that names are just an example of a concept word: we need to explain the assumed uniqueness the bearer of a name. Did Ayn Rand even attempt to put forward a theory of proper names? I would be surprised if she did. (It is a pretty hard problem to explain as anything other than scholastic quibbling to someone who has not previously studied philosophy).

An important point needs to be made here: we can use a name or concept word without fully understanding it -- in the manner of a 'tape recorder' (see Dummett's appendix on Kripke in 'Frege Philosophy of Language'). We are interested in what full understanding consists in. Putnam's so-called 'division of linguistic labour' obscures that fact. This is the basis of Evans' critique of Kripke and of his own previous causal theory. The fact that I can use the name 'Feynman' without knowing how to distinguish Feynman from Gell-Mann merely shows that I am not fully competent with the use of this word. You can be a link in the transmission of knowledge without fully understanding the information that you are transmitting.

Your reference to pigeon holes looks like something I once heard in a paper given by Strawson, according to which names function as cards in a card-index system. This may be true of the way we collate much of our knowledge of individuals, but we still have to explain how it is that we come to know *objects* at all.

Although you have taken some pains to explain Rand's theory, I don't really feel that I have a firmer grasp of it. The quotes didn't help me at all.

Essay 3. 'Can there be a satisfactory account of the notion of correspondence employed in the claim that 'a proposition or statement is true if it corresponds to the facts'?'

I liked this essay more, although it is not clear to me that you have really succeeded in identifying what is crucial to the notion of 'correspondence'.

Tarski, as you know, claimed that his semantic account of truth is a 'correspondence theory of truth'. However, what is most significant about Tarski's formal definition is that the notion of a 'fact' never gets a look in.

In Tarski, connection 'to the world' is accounted for in terms of assignments of objects to names and satisfaction conditions to predicates. End of story. The 'correspondence' in question is purely formal, as given by the Tarski schema 'P' is true if and only if P. The correspondence is formal in the same way as Aristotle's definition of truth: To say X when X obtains or not-X when not-X obtains is to say something true. To say X when not-X obtains or not-X when X obtains is to say something false. (End of story.)

One of the bones of contention between Austin and Strawson was precisely over this conception of correspondence -- i.e. a 'fact' for every true proposition. Austin not only wanted to reject Strawson's claim that the term 'fact' is eliminable in favour of 'true proposition', but also didn't like the idea that there the correlation between facts and true propositions is 1-1, which would imply that we have to admit, for example, negative facts (a problem which seems to have bothered Russell) or general facts. (See Austin's subsequent paper, 'Unfair to Facts' which is included in his Philosophical Papers OUP.)

You seem to side with Strawson over the second point but want to say (as many would) that a fact isn't *just* a true statement. A fact is out there in the world. A statement is something we say. It is true *because* of the existence of the associated fact, and for no other reason.

But the problem is still, what you are actually claiming when you say this. I agree that the issue is, ultimately, about the truth of realism (as contrasted with anti-realism). In Dummett's characterisation of realism, correspondence is only mentioned as a metaphor or picture which we are gripped by, not a constituent part of a realist theory. According to Dummett, realism is just the view that truth and not verification is the 'central concept in a theory of meaning'.

However, I think Dummett is wrong about this. An anti-realist has as much right to a Tarski-style definition of truth/meaning as a realist.

The bottom line is that belief in 'correspondence' reduces to belief that 'something out there makes our statements true, when they are true', as the realist believes, but every attempt to explain the alleged 'relationship of correspondence' reduces to things an anti-realist would, apparently, accept as happily as a realist.

All the best,

Geoffrey