Dear Sir,
Many thanks. It is a pleasure a receive a letter from East Moseley especially about the philosophy of Charles Travis.
You write:I see what you mean about the Witttgensteinian approach - ie one looks to see how vocabulary is being used, rather than thinking that the use should be predicted from a general account of the meaning of the descriptive terms used. But then given that pragmatic move, Travis still wants to provide a theorisation of the semantic machinery involved: ie
What holds all the cases together is precisely and only this: in each of them it is not too stupid to suppose that the (candidate) name bears an understanding that allows us to make sense, for any general concept, of the idea that what fits that understanding does not fit that concept, and vice-versa.
I’m not sure I understand what exactly Travis means by the terms in bold, but maybe that’s not important, and what he is trying to do is more therapeutic... ie it helps us become comfortable (no longer perplexed) with the language use, and we can throw away the ladder once we've achieved that. What holds all the cases together is precisely and only this: in each of them it is not too stupid to suppose that the (candidate) name bears an understanding that allows us to make sense, for any general concept, of the idea that what fits that understanding does not fit that concept, and vice-versa.
I think that Travis thinks he has spoken plainly. I don't think it’s supposed to be philosophy by pointing. In this case, I think he means that when we understand a name - ‘Scott’, say - to be a name - when we’re using it that way rather than as a shorthand for a description (cf Evans’ example of Julius as a conventional name for whoever invented the zip) - then for any Russell-style description - the author of the Waverley novels, say - we can understand the idea, or possibility, that Scott may not have written those novels. That possibility doesn't disbar ‘Scott’ from being a genuine name and, more radically, a name explained by using some general descriptions (unlike Kripke) which might have included, until the discovery, “You know the one, the author of the Waverley novels, not Captain Scott”.
That’s not to say that there are no bits that I don't follow. As I may have said before, he seems to support an idea he ascribes to Frege that once one has articulated a bit of the conceptual realm in the form of a judgement, then its truth or falsity depends on nothing other than how the world is. That’s a condition on objectivity which I think he accepts. It fits all his anti-McDowell stuff about the conceptual reaching as far as the non-conceptual realm that instances it. If, by contrast, truth turns also on the thinker, too, then it isn’t an objective matter. It might be an expression of subjective preference.
And yet, at the same time, following both Austin and his own ideas on occasion sensitivity, he also seems to think that there is some slippage. He quotes Austin: ‘Truth and falsity are . . . names for . . . a dimension of assessment—how the words stand in respect of satisfactoriness to the facts, events, situations, etc., to which they refer.’ Somehow, that idea of a ‘dimension of assessment’ doesn’t sound as unmediated as the Frege story.
I’d like it to be the case that Travis thinks that occasion sensitivity operates on words (used with standard meanings) to arrive at a specification of a bit of the conceptual realm - after which Frege's view holds - given the basic idea that words, even with their English meanings, do not specify what is being said. They can bear different understandings until a context is provided to know the language game being played. (Is it true that the room is dark? Well it depends which game is being played. Too dark to find Zettel but not dark enough to remove film from a canister.) But he also seems to suggest that the conceptual just is words used with standard meanings. So the slippage occurs between the conceptual and what was actually said not between the words and what was actually said ie the conceptual.
Having wrestled with this for about a month, I think I've now decided that I should say what I think he should say and accept that sometimes he's not consistent with that.
There’s another tension in his discussion of knowledge. Here what I think he should say is that an assessment of what alternatives need to be ruled out in saying that someone knows there's a sheep behind the barn on the basis of the noise heard - so there’s no risk of that it could have been, though it wasn’t, a similar sounding goat really making the noise - is occasion sensitive. In a later paper, this seems to be all he says. I‘m happy with that. In effect he replaces precise statements of the range of possible worlds a knowledge producing mechanism has to be reliable in with a rougher notion of relevance assessed in context. ((The danger of even this is that it might collapse into an account of when it is reasonable to ascribe knowledge to S rather than an account of what S's knowing consists in, or what conditions have to be met for S to know.))
But in ‘A sense of occasion’, he compares ‘S knows’ with ‘is blue’ where the latter might speak of a bucket of water looking blue, as though dyed, or a lake reflecting the blue of the sky. This implies that he thinks ‘S knows’ might be used in different language games. In this case, however, he doesn't tell us what those different games might be (unlike for ‘is blue’). But also, he rejects philosophical accounts of 'S knows' in which S merely has very good evidence. Now I think he’s right to reject that idea - it’s a bad philosophical account of knowledge - but I'm not sure that he has the right given the suggestion that ‘S knows’ might be used in different language games. Why not in this one, too?
Tim
