Wednesday, April 25, 2007


It’s always seemed to me that there is an obvious weak point in Kripke’s discussion of Wittgenstein. I don’t know the literature well at all so I may be making an elementary mistake or pointing out the obvious – either way I’d like to know.

Kripke argues that there are no meaning facts because there’s nothing to determine which of infinitely many rival meaning hypotheses are correct. Nothing about the world determines that I mean addition rather than quaddition by ‘addition’, for example. In arguing for this Kripke tries to show that various ways of choosing between the rival hypotheses won’t work. One way he considers is that it is the simplest hypothesis that is the correct one, and it is here I think he makes a mistake. He says

“[A]n appeal [to simplicity] must be based either on a misunderstanding of the sceptical problem, or of the role of simplicity considerations, or both. . . [S]implicity considerations can help us decide between competing hypotheses, but they obviously cannot tell us what the competing hypotheses are. If we do not understand what two hypotheses state, what does it mean to say that one is ‘more probable’ because it is ‘simpler’? If the two competing hypotheses are not genuine hypotheses, not assertions of genuine matters of fact, no ‘simplicity’ considerations will make them so.” (Kripke, Wittgenstein on Rules and Private Language, p38)

Kripke's point, I take it, is this. Grant for the sake of argument that if we have multiple hypotheses about what is meant by a term then the simplest of those hypotheses is the best one. Still this is of no help, because the conclusion proper of the sceptical argument is that there are no such competing hypotheses because any hypothesis about the meaning of a term is literally contentless. As Alex Miller puts it, “if two ascriptions of meaning do not have truth-conditions, what does it mean to say that one of them is more probably true because it is simpler?” (Miller, Philosophy of Language, p173)

But consider the dialectic that was intended to establish the sceptical conclusion. We were given a challenge by the sceptic to account for why it was we meant addition by ‘plus’. And, the argument went, if we can not answer the challenge, then we are stuck with a non-factivity about meaning. The simplicity considerations are raised as an attempt to answer that challenge, so we cannot argue against that attempt by appealing to a result which is only established if that attempt fails. Kripke is illegitimately assuming the conclusion of his sceptical argument at a point in the dialectic where it is still to be established. If simplicity considerations are such that they can adequately choose between rival hypotheses then the sceptical argument fails and the non-factivity of meaning does not follow. One cannot argue against this attempt at a straight solution by arguing that the rival hypotheses lack content; that is to assume what, at this stage in the dialectic, is still very much up for debate: the truth of the sceptical conclusion. So Kripke’s argument against the appeal to simplicity here simply begs the question.

What conclusion should you draw? I’m not saying the appeal to simplicity is the way to answer Kripke; there are still the problems with this appeal that Kripke lays to one side: i.e. “that simplicity is relative, or that it is hard to define, or that a Martian might find the quus function simpler than the plus function” (ibid). But if what I’ve said above is correct is lets us respond to an objection against Kripke’s sceptical solution.

Kripke embraces the non-factivity of meaning and argues that language use and communication is something like a community wide game. The claim is that there need be no facts about meaning for language to be useful. So instead of looking for meaning-constitutive facts – a futile effort, for there are no such things – Kripke suggests we look instead to the role meaning ascriptions play in the ‘langauge game’: to when ascriptions of meaning are justified in the discourse in question. Kripke says

“All that is needed to legitimize assertions that someone means something is that there be roughly specifiable circumstances under which they are legitimitately assertable, and that the game of asserting them under such conditions has a role in our lives. No supposition that ‘facts correspond’ to those assertions is needed.” (ibid. p78)

Kripke’s sceptical solution is criticised by José Zalabardo ('Rules, Communities, Judgements' Critica 63, 1989), who says that the ‘solution’ only tells us how to choose between rival hypotheses about what are the correct meaning ascriptions, and that if Kripke’s sceptical conclusion is right – and the rival ‘hypotheses’ are contentless – then the sceptical solution is no solution at all.

Now Zalabardo might be right that if Kripke’s sceptical conclusion is correct then the sceptical conclusion can’t get Kripke what he wants. But if what I said above about the simplicity argument is correct, then we can take Kripke’s ‘sceptical’ solution to be a perfectly good straight solution instead. If the ‘sceptical’ solution does let us choose between the rival hypotheses then why not let the meaning facts be what the ‘sceptical’ solution says they are? One can’t complain that there are no hypotheses for the sceptical solution to choose between, because that presupposes the sceptical conclusion; and if the ‘sceptical’ solution really does pick out the correct meaning hypothesis then the sceptical conclusion is not established. So to make that complaint against using the sceptical solution as a straight solution would simply be to beg the question in exactly the same way Kripke begged the question when dismissing the appeal to simplicity.


Aidan said...

Hey Ross,

Crispin makes precisely this point in footnote 6 of 'Kripke on the Argument against Private Language'. See page 109 of 'Rails to Infinity':

'Whatever criterion of preferability among competing hypotheses we come up with, its application can be appropriate only if we do genuinely have competing *hypotheses*, only if there is some 'fact of the matter' about which we are trying to arrive at a rational view. Therefore--or so Kripke's thought presumably runs--we beg the question against the sceptic in appealing to any such criteria at this stage. But this surely gets everything back to front. It is only *after* the sceptical argument has come to its conclusion that the sceptic is entitled to the supposition that there is indeed no such fact of the matter. In the course of the argument, *he* cannot assume as much without begging the question.'

Ross Cameron said...

That pesky Crispin - always pre-emptively stealing my work! I thought up neo-Fregeanism at the age of four you know!

(Still - it's hard to be annoyed when you find out you had the same thought Crispin had.)

jorgensens said...

Ross, is it not possible Kripke has a different conception of how simplicity bears on the issue? What do you make of his claim that "an omniscient being would have neither need nor use for simplicity considerations" (p.39)? Presumably he says this because omniscient beings deal in certainties and (he's thinking of) the simplicity of a hypothesis as just something that makes it more probable. I admit I don't really understand how he thinks simplicity considerations bear on the topic here.

He must mean something different by 'simplicity' anyway, because in footnote 25 on p.39 he talks about the bearing of a different use of simplicity, ``not that by which we evaluate competing theories'' where we compare all the different programmes a machine might run and where there is some objective way to measure and compare the simplicity of the programmes. Against this he says that we have at this point no good reason to believe that a human brain would run the simplest programme, and he thinks this suggestion doesn't provide a convincing response to his normativity objection to the dispositionalist solution.