Tuesday, June 12, 2012

The cluster of descriptions theory of proper names

To: Mark S.
From: Geoffrey Klempner
Subject: The cluster of descriptions theory of proper names
Date: 29 October 2007 12:52

Dear Mark,

Thank you for your email of 20 October, with your University of London Logic essay in response to the question, ''A proper name is an abbreviation of one or more definite descriptions.' - Discuss.'

This is (or would be, apart from one serious glitch) an impressive piece of work. Not only have you shown that you have digested the literature and understood the main lines of argument, you have also contributed a theory of your own. Although the theory, as stated, as serious shortcomings, it is not the worst theory that has been proposed as a solution to the problem of defining a proper name, and your explanation has some plausibility.

First, the glitch. On p. 4 when stating Kripke's modal argument, you state, 'For example, Richard Nixon would according to Kripke, have the proper name 'Richard Nixon' in all possible worlds, but he may have different definite descriptions in different possible worlds.' Then on p. 7, you say, 'I do not regard proper names as rigid designators. For example, if the lump of protoplasm that was Richard Nixon could, in an alternate world, have lost the 1968 election, I do not see why, in the same alternate world, his parents could not have named this lump, 'John Smith'.'

This is just wrong. 'Rigid designator' and a 'no-rigid designator' are terms which refer to referring expressions in our language -- the language we use in the actual world. 'Richard Nixon' is a rigid designator, according to Kripke, because when we apply the terms of our language to other possible worlds, it always picks out the same object. Whereas, 'Winner of the 1968 US election' is non-rigid because it picks out different objects in different possible worlds.

It is perfectly possible, on this model, that Richard Nixon might have been called 'John Smith'. The first name is a term in our actual language. The second name, in quotes, is not the name of any particular individual. Understanding the quoted expression does not require that one attribute to it any particular referent.

I have a smaller disagreement with what you say about Wittgenstein's comments on Russell's theory, in Philosophical Investigations §79. It seems clear from the context that Wittgenstein is not criticizing Russell but rather applying his theory and elaborating on it. He takes Russell to hold that a *particular use* of a proper name is intended to invoke a particular description. On another occasion, it might invoke a different description. Russell's response to Searle would be that it is a fault of natural language that we can't just all agree on what definite description or descriptions a name stands for (as one can, e.g. in mathematics).

Dummett wrote his response to Kripke in 'Frege Philosophy of Language' at the last minute just before the manuscript was due to go to press. His argument about scope is unconvincing because it clashes with the intuitions that Kripke appeals to: The actual writer of the Odyssey might not have written the Odyssey, but we do not say, on that account, that 'Homer might not have been Homer'.

However, Dummett has a much stronger case against Kripke, which you don't develop. There has to be a 'route to the reference'. If a name has currency in a language, then we need a philosophical account of how that is possible, the semantic facts by virtue of which the name links to that object. This is what Frege intended with his notion of 'sense'. The problem with Kripke's chain of communication theory, in Dummett's view, is that it fails to distinguish between someone who really grasps a name and someone who is merely 'acting like a tape recorder'.

The 'linguistic division of labour' idea muddies this point, because although it is true that linguistic labour is divided, we nevertheless require an account of what someone who fully grasps a term (rather than deferring to the 'experts') knows. This is the reason why Gareth Evans ditched his earlier 'causal theory of names' in virtue of the much more subtle account in his 'Varieties of Reference'.

Kripke does consider the theory (or a version of the theory) which you put forward. The problem, which Kripke and Dummett both appreciate, is that viewing a name N as equivalent to 'the person named N' doesn't answer the philosophical question of how it is possible to give anything a name, and then subsequently use that name to refer to that thing. If you read Evans' book, you will see just how hard it is to explain this.

Still, considering the difficulty of the topic you have done a good job.

May I also recommend the article by John McDowell, 'On the Sense and Reference of a Proper Name' Mind, New Series, Vol. 86, No. 342 (April 1977), pp. 159-185. (You'll find it in JSTOR.) This is a quite difficult and subtle paper but very well worth taking the time to study.

All the best,

Geoffrey