Analytic Philosophy Reexamined

ANALYTIC PHILOSOPHY REEXAMINED AALYTIC PHILOSOPHERS COMMONLY hold three theses which, taken collectively, put them in a hopeless bind as regards the identification of the bearer of truth. These are: a) that analytic statements are not about any extralinguistic fact, b) that truth consists in the correspondence of statement to fact, and c) that analytic statements are necessarily true. By accepting a), b) and c) our analytic philosopher is saddled with the following dilemma from which escape appears to be impossible. Either he holds that any proposition, analytic or synthetic, can be called true, or else he holds that only synthetic statements can be said to be true (or false). But given his commitment to a), b) and c) neither alternative is open to him. If he takes the former option he must either deny a) or b) above. If all statements, analytic and synthetic alike, can be the bearers of ' true ', then either analytic statements are about extra-linguistic facts or else truth is not definable as the correspondence of statement to (extra-linguistic) fact. On the other hand, if he takes the latter alternative, namely, that synthetic but not analytic statements can be called true (or false) then, while he may then consistently hold both a) and b), he must abandon c). Unless, therefore, our analytic philosopher gives up either a), b) or c) above, he is ensnared in a dilemma from which there is no escape. But the trouble is, since a), b) and c) are central to the programme of philosophical analysis, abandoning either a), b) or c) proves to be unpalatable for most analysts. First, as regards a), to admit that analytic statements are about facts would resurrect metaphysics in the old and grand style, as a science of necessary truths about reality. It would then have to be conceded that philosophy does not begin with Kant, that since there is a realm of' transcendent eternal facts which are ex251 252 JOHN PETERSON pressed by analytic truths, pre-critical dogmatic metaphysics is thereby reinstated in full dress. Second, as regards b) , to deny the correspondence view of truth is for our analyst to affirm that the fact, say, that it is now snowing is not itself a sufficient condition for making the proposition " It is now snowing ", true. If truth does not consist in the correspondence of statement to fact, it either does not consist in this at all, or else it does not consist only in this. But in either case the proposition that it is now snowing would not be made true just because it is as a matter of fact now snowing. But quite apart from how, if at all, such a view would be defended, the point is that no analytic philosopher would ever dream of defending it. Finally, with respect to c), since he holds that analytic truths are logically necessary truths, our analyst clearly cannot deny that analytic statements are true. While a proposition can be true without being necessary, it cannot be necessary without being true. Further , to deny that analytic statements are true is really to deny that they are statements at all. For by no stretch of the imagination can analytic statements be false, so that, if they are not true either and if, as he holds, all and only statements are true or false, then our analyst must deny that analytic statements are bona fide statements in the first place. But then his own division of statements into analytic and synthetic statements is undermined. If, then, he can abandon neither a), b) nor c) and if his holding these three theses precludes his identifying the bearer of truth either with analytic and synthetic statements both or with synthetic statements alone, then our analyst is in trouble as regards the referent of truth, it being impossible for him to hold a), b), and c) consistently with holding either that analytic and synthetic statements both, or that synthetic statements alone, are true. In the face of this difficulty, what is our analyst to do? How can he continue to affirm that analytic statements are verbal truths, that truth consists in correspondence to fact, and that analytic statements are...