Andy Egan, in "Epistemic Modals, Relativism and Assertion" says:
it can be appropriate for listeners at different times to attribute different truth values to one and the same utterance. (Such as when Blofeld says, at noon, “it’s lunchtime”. Number 2, in the room at the time of utterance, should agree that Blofeld’s utterance is true. Bond, listening to the recorded conversation later that night, should say that Blofeld’s utterance is false.)
I think Andy is probably right about the truth-value that Blofeld's utterance has when Bond listens to it. But it struck me that it's very unnatural for Bond to say that Blofeld's utterance is false in any of the following ways:
Blofeld (on tape): It's lunchtime. Bond: That's false/what he said is false/his utterance is false.
More natural are the following:
Blofeld (on tape): It's lunchtime. Bond: That was true/what he said was true.
Blofeld (on tape): It's lunchtime. Bond: That's not true any more/what he said isn't true any more.
("That's not true anymore" implies "That's not true" simpliciter; that's why I think Andy is right about the truth values.)
And what's interesting here is that it's natural to say that "what is said" was true and isn't any more. What is denoted by "what is said" is vague, of course, but there's at least a bit of reason to think that "what is said" is what plays the "proposition role" (see Andy's p. 5). Anyway, here what is said seems to have an argument for times, rather than just being simply a set of worlds. (This is in accord with Andy's main point, I think.)
This might cause a little bit of trouble for John Hawthorne's final argument against contextualism, in Ch. 2 of Knowledge and Lotteries. Hawthorne discusses the idea that we put propositions in a belief box, and points out that if contextualism about knowledge is true, then many things in our belief box may go from true to false when our context changes. But if the things in the belief box are akin to "What is said" here, then it shouldn't surprise us when the truth-values of those things change as context shifts; since they can change as time goes on. Part of a healthy cognitive economy will be keeping track of what's in your belief-box, and making sure that you don't continue to believe "what Blofeld said" when it's not true any more.
I think Hawthorne's point can survive this, though. It's relatively easy to keep track of the present-tense sentences in your belief box; that just involves keeping track of time. But keeping track of contextually determined "know" sentences in our belief box would be hard, because most of us aren't even aware (even supposing it's true) that "know" is contextually determined.
(Incidentally, I find myself inclined to say that Bond should say "Blofeld's utterance is true," but it's a very weak inclination. Egan's paper via Certain Doubts.)Posted by Matt Weiner at October 8, 2004 12:14 PM