Rob Stainton’s talk at the INPC inspired me to come up with a little example designed to show that if we accept the speaker-context-relativity of knowledge talk, we should accept audience-context-relativity of knowledge talk. In other words, whatever support there is for contextualism is really support for relativism in the sense Andy, John and I discuss in the epistemic modals paper, which uses many concepts from John MacFarlane’s paper about future contingents.

Here’s the case. It’s basically a real life case, though I’ve tinkered with some of the words.

ESPN, as I’m sure you all know, runs an all day broadcast from the NFL draft. They set up a small broadcast desk where the panelists sit around and talk football etc all day. At one stage, Chris Berman says the following thing.

(1) Mel Kiper Jr knows how fast each of the top prospects can run the 40.

It’s a little more natural to say he knows the 40 times for all the top prospects, but I don’t want to get caught up in ‘knowledge the’ debates here. So let’s say that he says (1). And to be sure, Kiper does have true beliefs based on good evidence about how fast each of the top prospects can run the 40, though of course he hasn’t ruled out that one of them can now run twice as fast because of some drugs he got from visiting aliens last night.

I’m watching the draft at home, and I say “That’s true – Kiper knows all that trivia.”

Cecil, a sceptic, watching the draft at his home says “That’s not true – Kiper can’t rule out the speed doubling aliens hypothesis.”

My (second-order) intuition is that if we are to respect the kind of sceptical/anti-sceptical intuitions that lead to Lewisian contextualism, we should respect both mine and Cecil’s intuitions here. We should, that is, say that both of us speak truly in the contexts we are in. But ordinary contextualism can’t do this, because we are both talking about what Berman said in the very same context – the ESPN desk at Madison Square Garden. The best option seems to be to say that the truth of Berman’s utterance is audience-relative, so it is true relative to me (and my standards) and false relative to Cecil (and his standards).

Well, not everyone likes Lewisian motivation for contextualism. Others think that airport cases or bank cases are better motivation. I find my intuitions in those cases go with Stanley and Hawthorne’s invariantism not contextualism, but let that pass. I claim that if these cases support contextualism they support relativism, and I can stay silent on the antecedent here.

So same case as above with Berman uttering (1).

It doesn’t matter much to Andy whether or not Kiper has this knowledge, so he agrees with me. He says “That’s so true – Kiper knows an insane amount about the prospects.”

It does matter a lot to Tyler however. I’m not sure just which dimension of mattering is meant to be important – whether it’s important to Tyler what the 40 times are (if he’s got some large bets on those times say) or whether it’s important whether Kiper has sufficient grounds for his beliefs (if he’s considering hiring Kiper say). But whatever your contextualism says is important, let it be important to Tyler. So he says “That’s not true – Kiper doesn’t _know_ this unless he has rechecked all of this _this morning_ and I bet he hasn’t done that.” (Tyler’s right by the way – often Kiper’s last check of 40 times is as much as 24 hours before the draft. The slacker.)

I don’t have the contextualist intuitions in airport and bank cases, but I can’t imagine that I would have them in those cases and not have the intuition that both Andy and Tyler spoke correctly. Nor can I imagine a theoretical ground for saying that we should accept the varying intuitions in the airport or bank cases, and not accept them in cases like this. So again I think the best option for one moved by the contextualists examples is not contextualism at all, but rather relativism.