Number 96 : April 1999 |
Note : I’m not sure whether I’m really at liberty to print this, but as it was handed out at a public lecture, I presume I am. This article is an extract from one of Stephen Mulhall’s books (forthcoming, I believe, as the handout looked like a photocopy of a galley-print). Stephen Mulhall is a fellow of New College Oxford and read a paper (though the above extract was the only hand-out) entitled Cavell, Murdoch & the Fact-Value Distinction at the Royal Institute of Philosophy lecture on Friday 26th February 1999. Yours truly was present in body if not spirit, and didn’t get as much from the lecture as he’d hoped, despite having an interest in the topic. The pace of presentation was too fast, especially for a Friday evening, for amateurs like me, who’d never heard of Stanley Cavell and never read Iris Murdoch. Apparently, says Stephen Mulhall, Cavell's criticisms were first published as chapter XII of The Claim of Reason (Oxford University Press: Oxford, 1979 - hereafter CR); but that chapter, essentially as it stands in CR, formed part of Cavell's PhD dissertation, submitted in 1961.
For Cavell, this thesis depends upon its being obvious that we cannot infer 'We ought to do X' [O] from 'We promised to do X' [R2] unless we also grant the ethical proposition 'We ought to keep promises' [R1]. It is true that O cannot be derived from R2 alone; and let's assume that it can only be derived from R2 with the addition of R1. This will only establish the thesis if the step from R2 to O is inferential; and Cavell denies that it is.
He utilizes a non-ethical example to make his point. Suppose that, whilst I am playing chess, a friend advises me that I ought to castle now [O], and, when I ask why, replies that castling will neutralise my opponent's bishop and develop my rook [R2]. Does O follow from R2 ? It is surely absurd to say that it would follow if we added the premise that: Whenever castling will neutralise a bishop and develop a rook, then you ought to castle. My friend is not saying anything like that; she's just saying that castling is the best move here. And she can display why she thinks that in the form of further premises: With my rook in play I can trap my opponent’s queen in two moves [R3]; with the queen gone I can win in four moves [R4]. From R2, together with R3 and R4, it follows (in terms of ordinary logic and a knowledge of chess) that: castling now will win [R5]. Does O follow from R5 ? Or is my friend making the further assumption that I want to win ? Does O not follow from R5 without accepting the major premise: If you want to win, then you ought to castle ? But again, my friend didn't mean anything so absurd: her advice is that castling will win here, not just any time. Does logical rigour then require the premise that: If you want to win, you should do whatever will win ? But is that intelligible ? Is it chess ?
Cavell concludes that there is no major premise that could bind R2 to O: but that is not because a special sort of logic or psychology is needed to bridge the gap between them, it is rather because there is no gap to be filled - the step is not the conclusion of an inference.
The gap which looks open on paper is closed in the act of confronting the player. To get a hint of what function 'ought' has, it should be noticed that any of R2-R5 could directly have been used to advise the player to castle: which of them you say depends upon what you think the person doesn’t see, and how many steps you take in your proof will depend on how much it will take for him to see what you see... If he cannot counter these 'factual premisses', whose point, in that context, is to advise him (and it wouldn't (grammatically) be advice unless you thought they would further his position), then unless he actually makes the move, or gives a good reason why he does not, then he either has not understood you, or has not mastered the game. (CR, p 317)
Cavell's claim is that the content of the original claim 'You ought to castle now is completely specified by the reasons you are prepared to give in its support; any proof available for any of the reasons R2-R5 is simultaneously a proof of O. Choosing O, choosing to say 'You ought to..', adds nothing whatever to the content those reasons specify, but rather determines the mode of that content’s presentation.
Theo Todman à I thought the above extract might be interesting, given our previous discussions on this topic and the references to chess-playing.
The point here seems to be that there’s no gap between "you promised" and "you ought to keep that promise" to be filled by "promises ought to be kept", as though the latter principle might be true of some societies but not others. The reason (and I don’t know whether you found the chess analogy interesting or helpful) is that it is part of the meaning or grammar of promising that a promise ought to be kept. So we have a statement of fact "he promised to do X" that entails an ought, "he ought to do X". The chess-playing analogy is that "wanting to win" is part of the grammar of chess-playing, and that someone who doesn’t want to win doesn’t understand the game. That’s my reading of the matter, anyway !
Well, we can object to both of these (the promising and the chess-playing), but the question is whether or not these objections are relevant in the context of the is/ought distinction. Firstly, objections to the "promising" assertion that, in general, promises ought to be kept. The principle that promises ought to be kept might well be intrinsic to the meaning of promising, but whether any particular promise ought to be kept at the time of its potential fulfilment is not known at the time of promising. It will depend on the state of affairs and the moral outlook of the individual & society when the promise comes up for payment. It will depend on whether or not the promise is over-ridded by some other contingency, some higher priority duty that it’s not possible to fulfil together with the promise. That prioritisation is not fundamentally factual because it requires bringing in a hierarchy of values. So, while we shouldn’t make a promise without some general intention of keeping it, societies might differ as to what would be allowed as an excuse for a person to renege on a promise.
As for chess-playing, Russian Grand Masters were well-known not always to play to win against their compatriots, who might thereby win the tournament. Similarly, its not always good policy to thrash one’s spouse or aspiring young sprog off the board. What is necessary as a minimum is to take the game reasonably seriously and pay attention, but even that might be over-ridden in the case of a very strong player playing a very weak one, or where training or encouragement for a protege is required. After all, when playing a computer we can set its playing level to sub-optimal. This doesn’t thereby mean that its not playing chess on levels 1 - 8 whereas it is on level 9. There are also degrees of wanting to win.
Theo