Saturday, November 18, 2006
Acquisition Again
Exactly a month ago I posted a long post on Dummett's acquisition challenge, and I've been promising since then to have a go to get clearer on the issues I raised there. John Bengson convinced me in discussion that I'd come across in my original post too much as playing defence on the behalf of the acquisition challenger. In actual fact, my intention was just to register my confusion and to get other people's reactions to the point that was confusing me.
In what follows I'm going to take some familiarity with what I wrote in the previous post on this topic for granted.
I take it that Dummett's challenge is to explain how we could have acquired a particular concept, one which the realist credits us with possession of, but which the anti-realist finds problematic. Here's Dummett (Alex Miller calls this Dummett's 'canonical formulation' of the argument):
'[The anti-realist] maintains that the process by which we came to grasp the sense of statements of the disputed class, and the use which is subsequently made of these statements, are such that we could not derive from it [i.e. the process] any notion of what it would be for such a statement to be true independently of the sort of thing we have learned to recognise as establishing the truth of such statements.'
('The Reality of the Past', T&OE: 362. My italics)
It's important to note that Dummett is explicit here that the challenge is not to show that we understand statements in the disputed class; that's conceded by all sides. Nor is the challenge for the theorist (realist or otherwise) who holds that understanding one of the disputed statements consists in grasp of its truth-conditions to show how we could have come to have grasped those truth conditions given that our linguistic training is a training in use. The challenge as I see it is to explain in a manner consistent with the nature of our linguistic training how we could come to form a conception of what it would be like for such a sentence to be associated with realist truth-conditions, i.e. conditions that can obtain or fail to obtain undetectably. This is hardly a sharp, adequately spelled-out statement of the nature of the challenge that's being presented to the realist, but it hope it's clear enough for my present purposes.
What I found so surprising in the Wright quote from last time was that it basically concedes to the realist that it suffices to meet the challenge as I've put it above that the realist can show the following: she has some story about how we could have come by grasp of the truth-conditions of one of the disputed statements in a manner consistent with the nature of our linguistic training:
'Given that the understanding of statements in general is to be viewed as consisting in possession of a concept of their truth-conditions, acquiring a concept of an evidence-transcendent state of affairs is simply a matter of acquiring an understanding of a statement for which that state of affairs would constitute the truth-condition.' (Realism, Meaning & Truth: 16)
Once we've made this concession, it's clear how compositionality helps the realist meet the challenge.
Now, part of my confusion last time stemmed from the following. Wright thinks there are insuperable difficulties standing in the way of a semantic theory which replaces the notion of truth with that of warranted-assertibility (or something similar). So we should adopt truth-conditional semantics. Understanding a sentence consists in grasping its truth-conditions; understanding one of the disputed sentences consists in grasping its truth-conditions. The realist is to be accused of 'overdescribing' what understanding a sentence consists in; she's right to hold that it consists in grasp of truth-conditions, but unwarranted in supposing it ever consists in grasp of realist truth-conditions (since such grasp could never be manifested in linguistic behaviour). As Wright is prone to putting it (regarding epistemicism, but clearly the point is more general than that), the realist is to be accused of superstition rather than error.
My puzzlement was over this; why can't we offer the same response to the compositionality response to the acquisition challenge? We'd then concede that we understand the disputed statements, that understanding those statements consists in grasping their truth-conditions, and that one could come (via compositionality) to grasp their truth-conditions in a way consistent with recognising that our linguistic training is a training in use. But what forces us here to say that the truth-conditions we've arrived at are realist truth-conditions? (I've discovered since writing my first post on this that Miller explicitly considers such a response to the compositionality response in his 'What is the Acquisition Argument?'. See pp. 492-4)
Actually, there seems to be a very simple reason to think we are forced to regard the truth-conditions of the disputed statements as realist truth-conditions, namely that (as is conceded on all sides) these statements are undecidable (See Miller p491 and p7 of Williamson's 'Must Do Better'.) I think that there's a serious challenge there, but setting that point aside for now, we still have something of a mystery as far as I can see. Why did Wright think that it sufficed to meet the acquisition challenge - formulated as a challenge to explain how we could have acquired a conception of what it would be for a statement to be associated with realist truth-conditions - to explain how we could come to grasp (in the right sort of way) the truth-conditions of one of the disputed statements, given that he holds that nothing warrants us in regarding those truth-conditions as realist truth-conditions?
In what follows I'm going to take some familiarity with what I wrote in the previous post on this topic for granted.
I take it that Dummett's challenge is to explain how we could have acquired a particular concept, one which the realist credits us with possession of, but which the anti-realist finds problematic. Here's Dummett (Alex Miller calls this Dummett's 'canonical formulation' of the argument):
'[The anti-realist] maintains that the process by which we came to grasp the sense of statements of the disputed class, and the use which is subsequently made of these statements, are such that we could not derive from it [i.e. the process] any notion of what it would be for such a statement to be true independently of the sort of thing we have learned to recognise as establishing the truth of such statements.'
('The Reality of the Past', T&OE: 362. My italics)
It's important to note that Dummett is explicit here that the challenge is not to show that we understand statements in the disputed class; that's conceded by all sides. Nor is the challenge for the theorist (realist or otherwise) who holds that understanding one of the disputed statements consists in grasp of its truth-conditions to show how we could have come to have grasped those truth conditions given that our linguistic training is a training in use. The challenge as I see it is to explain in a manner consistent with the nature of our linguistic training how we could come to form a conception of what it would be like for such a sentence to be associated with realist truth-conditions, i.e. conditions that can obtain or fail to obtain undetectably. This is hardly a sharp, adequately spelled-out statement of the nature of the challenge that's being presented to the realist, but it hope it's clear enough for my present purposes.
What I found so surprising in the Wright quote from last time was that it basically concedes to the realist that it suffices to meet the challenge as I've put it above that the realist can show the following: she has some story about how we could have come by grasp of the truth-conditions of one of the disputed statements in a manner consistent with the nature of our linguistic training:
'Given that the understanding of statements in general is to be viewed as consisting in possession of a concept of their truth-conditions, acquiring a concept of an evidence-transcendent state of affairs is simply a matter of acquiring an understanding of a statement for which that state of affairs would constitute the truth-condition.' (Realism, Meaning & Truth: 16)
Once we've made this concession, it's clear how compositionality helps the realist meet the challenge.
Now, part of my confusion last time stemmed from the following. Wright thinks there are insuperable difficulties standing in the way of a semantic theory which replaces the notion of truth with that of warranted-assertibility (or something similar). So we should adopt truth-conditional semantics. Understanding a sentence consists in grasping its truth-conditions; understanding one of the disputed sentences consists in grasping its truth-conditions. The realist is to be accused of 'overdescribing' what understanding a sentence consists in; she's right to hold that it consists in grasp of truth-conditions, but unwarranted in supposing it ever consists in grasp of realist truth-conditions (since such grasp could never be manifested in linguistic behaviour). As Wright is prone to putting it (regarding epistemicism, but clearly the point is more general than that), the realist is to be accused of superstition rather than error.
My puzzlement was over this; why can't we offer the same response to the compositionality response to the acquisition challenge? We'd then concede that we understand the disputed statements, that understanding those statements consists in grasping their truth-conditions, and that one could come (via compositionality) to grasp their truth-conditions in a way consistent with recognising that our linguistic training is a training in use. But what forces us here to say that the truth-conditions we've arrived at are realist truth-conditions? (I've discovered since writing my first post on this that Miller explicitly considers such a response to the compositionality response in his 'What is the Acquisition Argument?'. See pp. 492-4)
Actually, there seems to be a very simple reason to think we are forced to regard the truth-conditions of the disputed statements as realist truth-conditions, namely that (as is conceded on all sides) these statements are undecidable (See Miller p491 and p7 of Williamson's 'Must Do Better'.) I think that there's a serious challenge there, but setting that point aside for now, we still have something of a mystery as far as I can see. Why did Wright think that it sufficed to meet the acquisition challenge - formulated as a challenge to explain how we could have acquired a conception of what it would be for a statement to be associated with realist truth-conditions - to explain how we could come to grasp (in the right sort of way) the truth-conditions of one of the disputed statements, given that he holds that nothing warrants us in regarding those truth-conditions as realist truth-conditions?
Labels: Anti-Realism, Dummett, Philosophy of Language
# posted by Aidan @ 2:15 PM 
Comments:
<< Home
So, if I have it right, the issue is the following: there is some set of sentences which are disputed, call it D. It is agreed that members of D are understood. Furthermore, these sentences seem to have evidence transcendent truth-conditions. But, you note, nothing Wright says requires us to regard the truth-conditions of those sentences as realist truth-conditions. The question then is whether showing how we can come to grasp the truth-conditions of members of the disputed class helps or doesn’t help meet the acquisition challenge. I think it does (though I confess I’m not at present familiar with the Wright you cite, so excuse me if I rehash some of what you read). Here is what I take to be the key. Although we are not warranted in regarding the truth-conditions of members of the disputed class as evidence-transcendent, we are so warranted in many other cases. Let’s call this class of sentences R. [Note, though, that these sentences are such that we are probably warranted in believing both that they have realist truth-conditions and that they have irrealist truth-conditions (a la Dummett); nothing in these cases seems to require choosing only one type of truth-condition.] But, if we can show that using only the syntactic resources available to one who understands members of R in the realists’ sense and also that a recursive semantic machinery capable of generating the members of R can also generate members of D, then we will have met the acquisition challenge. Although, according to Wright (from your post), we are not warranted in ascribing realist truth-conditions to members of D, it seems we would be now. So it looks like we have a bit of a contradiction on our hands. The way out, as I see it, is to maintain that we are not warranted in ascribing realist truth-conditions to members of R taken in isolation (preserving Wright’s claim). Not taken in isolation, however, but as consequences of a compositional meaning theory whose base clauses are understood as a realist claims to understand them, we are warranted in regarding members of D as having realist truth-conditions (preserving Wright’s claim about the acquisition challenge).
The question now seems to be whether there is some set of sentences that for independent reasons we can regard as having realist truth-conditions and whether their semantic constituents are simple enough to figure in the base clauses of a compositional meaning theory. Enough for now, though, I want to mull this over.
Post a Comment
The question now seems to be whether there is some set of sentences that for independent reasons we can regard as having realist truth-conditions and whether their semantic constituents are simple enough to figure in the base clauses of a compositional meaning theory. Enough for now, though, I want to mull this over.
# posted by 
: 11/26/2006 4:37 PM
: 11/26/2006 4:37 PM << Home
