- I am interested in discussing, if not quite defending, what has come to be called 'anti-realism'.
- The catalyst of my remarks on this occasion is John McDowell's paper ‘On The Reality of the Past’, though I shall also be looking directly at the sources of his paper in the work of Michael Dummett ["Dummett (Michael) - Truth and Other Enigmas"] and noticing the long, luminous shadow of Wittgenstein which falls, I think benignly, across this whole domain of enquiry.
- According to McDowell, philosophers such as Gottlob Frege and Donald Davidson "have found attractions in the idea that a theory of meaning for a language might include a component capable of specifying, for any indicative sentence of the language, a condition under which it is true" (‘On The Reality of the Past’, p. 127).
- Details of formal development and axiomatization aside, the basic idea that to understand a declarative sentence consists in knowing (at least implicitly) the circumstances in which it is (or would be) true is solidly incorporated in philosophically innocent common sense, the common sense of the schoolmaster, e.g., who invites his student to explain what some tortuous sentence means by saying what the world would be like in case the sentence in question is true.
- Dummett concedes the "intuitive obviousness" of the truth-conditions account of meaning which he thinks derives in large measure from what he calls the equivalence principle:
Any sentence A is equivalent in content to the sentence “It is true that A”. "This seems to show [Dummett writes] that truth must be the right notion to use to explain meaning: we could not say, for example, that to know the meaning of a sentence A is to know what has to hold good for A to be known to be true, for 'It is known that A' is much stronger than A itself; nor could we say that it is to know when there are adequate grounds for asserting A, for these could exist even though A was false." ("Dummett (Michael) - What is a Theory of Meaning?", p. 77)
- Dummett's own move away from a truth-conditions account of meaning consists, not in his coming to doubt the equivalence principle, but in his ceasing to take the notion of truth for granted, in his coming to find our possession of such a notion itself problematical and explanation-requiring. (I shall call a semantics 'realist' or 'classical' to the extent that it employs or believes in a truth-conditions ex plication of meaning. This is merely convenient taxonomic nomenclature, though, in another paper, 'Saussure's "significant / signifie"', I try to show that the essence of the various realisms philosophers have actually defended or derided is simply semantic realism.)
- Philosophers have often talked of truth when they desired something guaranteed.
- But if 'truth' is defined as 'undefeated warranted assertibility' (or 'warranted assertibility' is defined as 'defeasible truth'), talk of truth will serve instead to remind us of how little, if anything, is guaranteed against revision, revocation, cognitive reconsideration.
- This is what I mean by saying that truth is 'fugitive'.
Text Colour Conventions (see disclaimer)
- Blue: Text by me; © Theo Todman, 2019
- Mauve: Text by correspondent(s) or other author(s); © the author(s)