To: Mark S.
From: Geoffrey Klempner
Subject: Understanding the notion of necessity
Date: 14 February 2008 11:58

Dear Mark,

Thank you for your email of 3 February, with your University of London Logic essay written under exam conditions in response to the question, 'How can we best understand the notion of necessity?'

On the face of it, this question looks like an open invitation to say whatever you like on the topic of theories of necessity. Accordingly you start off by stating that you will restrict the context to alethic modality, and in particular the de re interpretation of the modal operator.

The result is that you have written a very good essay in response to the question, 'Elaborate the main ways in which the notion of possible worlds may be employed in analysing the notion of de re necessity.'

Given the time constraints you have done very well. I do have a horrible memory from my second year as an undergraduate at Birkbeck, giving a talk on one of the essays from Quine's excellent collection 'Ways of Paradox' on the notion of necessity. I'd just discovered Lewis and Kripke, and proudly pulled Kripke's theory like a rabbit out of a hat as a 'solution' to the problem of transworld identity posed by Lewis's theory. The lecturer, Mark Platts (who now teaches at the Institute of Philosophical Investigations, Mexico City) had a bad throat that day and was not very pleased at having to explain, at painful length, why I had dismally failed to address Quine's concerns.

To this day, I don't know whether I had any part to play in Mark's decision to leave the country.

Modal fictionalism looks like the perfect rabbit to pull out of a hat. Ontologically parsimonious, based on a strong intuition (e.g. Sherlock Holmes in the 'world' of Conan Doyle's novels), the answer to all our prayers.

The problem for me is that statements about necessity, contingency and impossibility claim truth. You can argue endlessly about the character of Sherlock Holmes and never reach a definitive conclusion: that is the way with questions of literary interpretation. Sure, he lives at Baker Street. That only means that the famous street gets mentioned in the books. What if, for fictional purposes, Conan Doyle switched the locations of Baker Street and Wigmore Street (admittedly a dumb thing for a novel writer to do; as a Londoner I knew these streets well). Does Holmes live in Wigmore Street or Baker Street? I don't think there is a 'truth' there. You can say what you like, as long as you say it clearly.

Whereas with statements about necessity, I don't think that we are prepared to allow truth to reduce to a species of interpretation. There are necessary facts. Leaving aside nomological necessity, what kind of fact is it that a surface cannot be red and blue all over, or that a man cannot spring from dragon's teeth?

This bears on the original question, which looked to me like a perfect opportunity to use the kind of analysis which you employed with the notion of knowledge, I mean the Wittgensteinian family resemblance or Aristotelian focal meaning kind of analysis which explains there is one word, while at the same time distinguishing between different but related applications or 'meanings'.

Lewises mad dog realism looks the best candidate for accounting for truth. The problem is believing it. It gives necessities which seemingly have nothing to do with language alongside those which can be accounted for in linguistic terms -- including the kind of distinctions that Kripke makes about the necessities of origin, which are arguably analysable in terms of the logical criteria for spatio-temporal continuity. (The individuals I might have been are restricted to the set of possible life histories terminating at this present point, where a life history is defined in terms of the sortal concept 'human'.)

In other words, although my intuitions are less frugal than Quine's, I am with Quine on the desirability, wherever possible, to account for necessity in logical/ linguistic terms. From this perspective, the hardest cases are those like the red and green.

If I was writing this essay, I would go even further back, starting with Wittgenstein's attempt, in the Tractatus, to account for all necessity in terms of strict logical necessity. Why does that idea still seem so appealing? why was it destined to fail? what does that failure show about the nature of language, and the nature of reality?

I'm not saying you have to do this. The question is less restrictive than others that you have answered and so you do have a certain degree of choice. However, I do think that you need to at least throw a sop to the logician/ semanticist who thinks that given we have single word, 'necessity', there is at least a strong presumption that in some sense it indicates a single 'notion', whatever that notion may be.

All the best,

Geoffrey