Thursday, November 7, 2013

Theory of descriptions: Strawson vs Russell

To: Emelie G.
From: Geoffrey Klempner
Subject: Theory of descriptions: Strawson vs Russell
Date: 27th January 2011 14:14

Dear Emelie,

Thank you for your email of 20 January, with your essay for the University of London BA Logic module, in response to the question, ''The present King of France is bald' is neither true nor false. So Russell's Theory of definite descriptions is mistaken.' Discuss.'

You give a good, clear exposition of the bare bones of Russell's theory, as given in his article 'On Denoting', followed by a brief account Strawson's response to Russell in his article 'On Referring'.

The crux of your argument is that (i) Strawson's claims regarding our intuitions about which statements have or fail to have a truth value are inconclusive, in the light of examples which tend to bring out contrary intuitions. (ii) There are cases where Strawson's view cannot be applied, e.g. 'The greatest prime number does not exist' and 'The centre of mass of the Solar System is constantly changing.'

Let's start with (ii). I don't think Strawson would feel that these are particularly impressive or difficult counterexamples.

'The greatest prime number does not exist,' is a very odd way of saying, 'There is no greatest prime number,' which is explicitly a quantified statement.

'The centre of mass of the Solar System' may be constantly changing, but the presupposition of making this statement, Strawson would say, is that the Solar System has a centre of mass, which successively occupies different spatial positions. If in giving a paper to a conference of astrophysicists, I stated, 'The second centre of mass of the Solar System moves in the opposite direction to the first centre of mass,' the audience would express the same perplexity as someone who is told that the Present King of France is bald. As any first-year physics student knows, there can only be one centre of mass, that's just what centre of mass is, and anyone who assumes the opposite isn't saying something false, they are talking complete and utter nonsense.

On the other hand, if in my paper I had said, 'I believe that I have discovered a second centre of mass of the Solar System', then the focus would be on whether my thesis was true or not. Once again, my statement is now explicitly a quantified statement, as in the example of 'the greatest prime number'.

Strawson isn't denying that there is a role for quantification in natural language: his case is that this role is far less than Russell would have it.

So, what is the case that Strawson is making? As you present it, his argument relies exclusively on our intuitions about whether a statement can be said to 'true' or 'false', or lacking in truth value. And as you point out, intuitions cut both ways. It all looks very inconclusive. (A popular nickname for Strawson amongst my generation of students was 'Strawman'. I think in this case your 'Strawson' is indeed a straw man. His 'objection' to Russell seems to carry very little weight and is easily defeated.)

I would like to give Strawson a run for his money. First off, we need to decide a big question, which you don't discuss, namely why it is so important to fill truth-value gaps. (There is an excellent 1959 paper by Michael Dummett on this question, 'Truth', reprinted in 'Truth and Other Enigmas'. Although the paper is short, it is quite difficult. However, I am mentioning it because of its importance.) The short answer is that if we allow truth value gaps, logic gets rather complicated and difficult to do, and there is a premium, as Russell recognized, in being able to apply first-order predicate logic to natural language.

However, this doesn't mean that we should be prepared to pay any price to fill truth-value gaps. So what is the real motivation for Strawson's view? Strawson believed that he had discovered an important concept which is pervasive in natural language. The example of non-referring singular terms is just one case of this phenomenon: the phenomenon of presupposition. First-order predicate logic, as described by Frege and Russell fails to take account of this important phenomenon.

Is he right? How useful is the notion of presupposition? Where else can you find examples of the application of this notion? Can you think of any?

Another point made by Strawson concerns the difference between a Rusellian 'proposition' and what we would refer to, in ordinary language, as a 'statement'. It is not implausible to argue that the primary unit so far as understanding a language is concerned, is a statement, a form of words which can be true on some occasions, and false on other occasions. E.g. 'The snow has melted.' Yesteday, the statement was false because there was still snow on the ground. Today, the statement is true. In a country where snow never falls, the statement cannot be used to state something true, or to state something false -- in other words, it doesn't have a use, because the presupposition fails. By contrast, Russell insists on identifying 'meaning' with the particular use of a sentence on a particular occasion. So, whereas for Strawson a statement sometimes has a use and sometimes not -- which seems intuitively plausible -- for Russell, every occasion on which a person 'asserts a proposition' should be evaluated in the same way, regardless of the circumstances.

All the best,

Geoffrey