I will write up a longer post on the SWM (from which I just returned) but for now I just wanted to record one neat argument for the ambiguity of ’caused’ that turned up over one lunch at the conference. (It turned up in a conversation b/w John Hawthorne, Carolina Sartorio and me, but I honestly can’t remember exactly who was responsible for what, except that (4) was Carolina’s example. I’m not sure whether John or Carolina were convinced by the argument, so don’t hold them responsible for any wild conclusions.)

Consider a fire that has a lot of consequences. For one thing, the building that is on fire explodes. For another, I’m saddened by all the web servers that are burning down losing valuable websites. So both (1) and (2) are true.

(1) The fire caused the explosion.
(2) The fire caused Brian to be sad.

We can say both those things, but we can’t say them at once.

(3) *The fire caused the explosion and Brian to be sad.

You might think that’s problematic for pragmatic reasons, but we really can say what we want to say in (3) using other words.

(4) The fire caused the explosion to happen and Brian to be sad.

The natural conclusion to be drawn from this is that ’caused’ is syntactically ambiguous. On one disambiguation it takes an event as its second relata, on another it takes whatever is denoted by “Brian to be sad”. (I guess that’s a state of affairs, but I don’t really know the literature around here. I suspect I should go back and read Vendler on the issues in this neighbourhood.)

If that’s right, all the different arguments about what the relata of the causal relation are need to be at least restated. There is no fact about what the relata of the term ’caused’ are, because there is no single such term, and different terms spelled that way take different arguments. There might still be an argument about what the deep metaphysical relation causation relates, and if there is syntactic facts probably won’t be taken to have much bearing on it.