Does the following sound funny to you?
(1) When Paul says "The CD costs 20 dollars," what he says is true, but the CD doesn't cost 20 dollars.
[My answer below.]
It sounds funny to me--which is possibly too bad, because it certainly can be true. The thing is, Paul is a Canadian, so when he says "The CD costs 20 dollars" he means that it costs 20 Canadian dollars, which is less than 16 US dollars. When the USian who utters (1) [call her Laura] says "The CD costs 20 dollars," she means 20 US dollars. So she can acknowledge that Paul speaks truly while denying what (it seems) Paul said, in Paul's words.
Is it that "dollar" is ambiguous? I don't think so, because this sentence is acceptable:
(2) The CD costs 20 Canadian or 16 US dollars.
If "dollar" were ambiguous between Canadian and US dollars, (2) would at least sound funny, like this:
(3) ??The bench is located between the savings and river banks.
Perhaps "dollar" is akin to a multiply referring proper name, and (2) is akin to the following:
(4) In this course, we will read books by Kingsley and Martin Amis.
There are probably other ways to account for "dollar"--perhaps it behaves like "enemy" (he says, maintaining a prudent silence concerning how "enemy" behaves)--but I'd like to highlight a hypothesis that maybe doesn't settle the semantics. Namely, when we discuss "dollars" we're usually doing so in a context in which only one nation's currency could be in question. So when we have a conversation in which there are two kinds of dollars involved, it sounds kind of funny, as in (1). But there are devices (as in (2)) that let us clear up what we mean, at the cost of a little extra breath.
What I have in mind here, of course, is the word "know." Cappelen and LePore use examples like (1) to argue against the context-sensitivity of "know":
(5)*When Paul says "I know that the Flyer stops in Chicago," what he says is true, but he doesn't know that the Flyer stops in Chicago"
sounds very bad, whereas truly context-sensitive terms should admit of similar negated disquotations:
(6) When Paul says "I am forty-three years old," what he says is true, but I am not forty-three years old.
((6) may be a bit funny because of the use of "but," but that's not what at's issue.)
My thought is that perhaps "know" behaves a little like "dollar." Take for granted what many contextualists and subject-sensitive invariantists agree on: that the amount of evidence required for knowledge can vary with the stakes of the question at hand. Perhaps our use of the word "know" is generally governed by the assumption that the stakes that set the standards for knowledge will be the same for everyone in the conversation. (This may sound implausible, but take Edward Craig's view that the point of knowledge ascriptions is to identify good informants. Then, even if the point at issue doesn't matter to a person, whether we think she has enough evidence for knowledge can depend on how much it matters to us--since we're thinking about asking her about it.)
If that's the case, then we shouldn't be too surprised when "know" behaves oddly in cases in which the stakes are not the same for the people participating in the conversation and the subjects of the knowledge ascriptions. To somewhat awkwardly extend the dollar metaphor*, knowledge has a different value for the ascriber and the ascribee, though it is the same kind of currency for both. If we really want to be perfectly clear about whose standards are in play, we have to use devices to make the standards explcit, of the sort that Peter Ludlow has discussed. (Analogous to specifying whose dollar you're discussing.)
But in most cases, everyone participating in and discussed in the conversation is working with the same currency of knowledge. So it doesn't cause much harm that knowledge ascriptions typically use a word that doesn't sound as though it can pick out more than one epistemic standard, but really does; just as "dollar" sounds as though it picks out one kind of currency, but really picks out several, depending on which country you're in.
At least, that's one position. Have at it, if you've reached this far.
*This is not the worst split infinitive ever**, but it's got to be close; still, I kind of like its self-reflexive quality, so I'm leaving it. Plus I'm too lazy to rewrite the sentence.
**Scroll down to the comments.
Posted by Matt Weiner at August 24, 2004 03:58 PM