To: Bogdan P.
From: Geoffrey Klempner
Subject: Proper names as rigid designators
Date: 22nd January 2010 13:01

Dear Bogdan,

Thank you for your email of 15 January, with your essay for the University of London BA Logic module, in response to the question, 'A proper name designates the same object in any possible world in which that object exists.' Discuss.

This is a good essay which shows a confident grasp of the issues around the question of proper names and rigid designation.

Your answer can be improved. The main issue concerns the resources of the description or cluster of descriptions theory for accounting for our intuitions regarding statements about individuals in other possible worlds.

Around 1976, when I was an undergraduate at Birkbeck, I was fortunate to attend a lecture by Kripke held at the University of London. An old lady in the audience asked a question which Kripke didn't seem to get the point of immediately, and there was some tittering from the audience. The 'old lady' was the philosopher G.E.M. Anscombe, and her question was, 'What do you think of this joke: 'Scholars have discovered that the Odyssey was not written by Homer, but by someone else with the same name.''

What do you think of the 'joke'? On Kripke's theory, isn't it perfectly correct to say that in some other possible world, the Odyssey was not written by Homer but by some other author who (coincidentally) was called 'Homer'?

Let's put a bit of flesh on this. First as you say, 'I imagine putting [Homer] in an imaginary situation'. Then I add details to this 'possible world'. Homer is too lazy to write the Odyssey. Instead he gets a job as a scribe to the local corn merchant. However, another writer, coincidentally called 'Homer', does write the Odyssey.

The alternative view would be, given that all we know (or ever expect to know) about 'Homer' is that he (or she) wrote the Odyssey, then the name 'Homer', in any possible world in which Homer exists, designates the author of the Odyssey in that world.

Is this a substantial dispute? I don't think it is. Don't we have a choice, whether to use a proper name as Kripke and Evans describe, or as equivalent to a description or set of descriptions, as suits our communicative purposes? Who is Kripke to lay down the law about how names should be used?

I think Kripke has a reply to this, along the lines of remarking that 'Homer' is a special case, and that he is interested in the central cases of proper names, that is to say, the point of having proper names in our language.

But there is more to say. Other critics of Kripke, for example Michael Dummett in his Appendix to the chapter on 'Sense and Reference' in 'Frege Philosophy of Language' (1973) argues that it is perfectly possible to account for our modal intuitions regarding names like 'Aristotle' (Dummett uses the example 'St Anne') by making scope distinctions. Russell was fully aware of the possibility of taking definite descriptions as having 'broad' or 'narrow' scope. That's all 'rigid designation' amounts to.

Let's define 'Aristotle' as 'the pupil of Plato and teacher of Alexander who founded the Lyceum'. When we say, 'Aristotle might not have taught Alexander' what we mean is:

'There is an x such that x is the pupil of Plato and teacher of Alexander and founder of the Lyceum, and in some possible world w, x is not the teacher of Alexander in w.'

The only point where Kripke's view of proper names diverges from the 'scope distinction' analysis is in the case (similar to 'Homer'!) where we want to say that in some possible world, Aristotle might not have been the pupil of Plato, taught Alexander or founded the Academy. In that case, according to the description theory (assuming that we define 'Aristotle' by these three descriptions) what we mean is that 'Aristotle might not have been Aristotle'. Whereas on Kripke's account, the infant baptised with the name 'Aristotle' in the actual world might not have been the pupil of Plato etc.

However, the description theorist can reply that there are far more complex and subtle ways of defining 'Aristotle' by means of descriptions.

Again, it's not clear that we need to decide this. Because, in the above analysis, we have already assumed an interpretation of modal predicate logic according to which it makes sense to quantify over a domain of objects which exist in different possible worlds. 'There IS an x such that...and in some other possible world w, x...'.

Which brings us back to the topic of the essay: does the notion of 'trans-world identity' make sense? If your view is just that talk of possible worlds is mere 'make believe' (as you seem to interpret Kripke) then there's no great difficulty. But then what kind of attitude should we take to the proposal that modal statements can be true, and have truth conditions? How can we be talking about truth, if it's all just make-believe?

On the much more robust view of David Lewis, possible worlds are every bit as 'real' as the actual world. There is no trans-world identity. In assessing the truth of counterfactual statements, we look for the most 'similar' world or set of worlds.

On the face of it, this account is open to the objection that, in world w, 'Aristotle' is the object most similar to Aristotle in the actual world. But I don't think Lewis has to say this. Another way of evaluating 'similarity' is in terms of imagining deviations from the actual world. If you want to imagine a different life for Aristotle, you take the infant called 'Aristotle' in the actual world and run a scenario where different things happen to him. In other words, Lewis can just help himself to Kripke's intuitions about origin and essence.

All the best,

Geoffrey