In this thread, Kent Bach finds himself having trouble answering a question from me, because I hadn't actually succeeded in asking one. This reminds me of what happened when John MacFarlane gave a talk at the University of Utah on epistemic modals. (I shouldn't cite his paper, but surely I can link his papers page.) A couple of us were chatting in the break before the discussion period; I raised some point (which I've forgotten) and John asked me if I was planning to ask is as a question. I said, "No, I'm planning to say some long thing with no question marks in it at all."
When I was called on in the question period, I said:
Take this example: Batman and Robin are chasing the Joker around an abandoned warehouse. The Joker fires two shots, which miss. Then Spiderman shows up and starts chasing the Joker too. The Joker fires four more shots, which also miss. Spiderman then advances on the Joker and subdues him. Batman says to Robin, "That was very foolhardy, because the Joker might have had more bullets left." That seems to me true, and the theory on which "might" depends on the knowledge of some group G can account for it, as long as it's OK to have the group exclude the speaker and hearer. And I think your proposal has trouble with that.
After I finished speaking, Matt Shockey whispered to me, "You should have had them chase the Riddler. That way what you said would've had question marks in it."
(My example was inspired by Brian's second example here.)
Posted by Matt Weiner at April 15, 2004 03:57 PM