Causation, dependence, and the fragility of events
One problem for simple counterfactual accounts of of causation comes from the simplest kind of counter-example. Al throws a rock at a window, but betty throws one at the same time and a bit harder so it gets there first and breaks the window so that Al's rock just sails through an open space.
Betty's rock caused the window to break, but the breaking of the window didn't depend on Betty's rock because if she hadn't thrown it (or hadn't thrown it well) Al's rock would have broken it.
One way out of this is to think of the causal relata as very fragile token events. The homeowner may not care to slice things as thinly, but the metaphysician, I say, does. We want a cause of *this* breaking, not just of the *generic* fact that the window broke (or we might want either given our purposes and some kind of pluralism reigns here, but I want to bracket that for now if I can).
Objection: but then there are all kinds of causes of events, since there are ever so many events influencing the character of effects (taken as pretty fragile event tokens.
Spin: Of course! This is just to *recognize* the plain fact that all kinds of causes have all kinds of effects on events. (Host of good points in Lewis's "Causation as Influence" and Hall's "Aspect Causation."
Objection: but doesn't this threaten to make *causes* essential to their effects!? And wouldn't that *trivialize* an account of causation?
Spin: If it does make causes essential to their effects, it just reflects the fact that what the *metaphysician* is interested in are event tokens for which lots of essentialism is true (and property exemplifications and like event-candidates are pretty good candidates for lots of essentialism if you ask me). Of course! the folk don't care about slicing it so thin because when it comes to the properties of the effect-event *they're* interested in *whom it makes responsible* being paramount--almost all of the nearby alterations have all the same properties. Everything that's essential-to-them is true of all the closest alterations. But they're not trying to do metaphysics. They just want to know who the Hell is going to pay for that damn window!
And it might not imply that causes are essential to their effects, but that functionally equivalent causes are. So if some other event would have caused an event of the same type with all the same essential properties then it would have been the same token. So then what if Al's throw was somehow interfered with by Betty's throw such that if Betty hadn't thrown Al's throw would have caused the breaking of the window to occur in *exactly* the same way that Betty's did (save that it was Al's rock and not Betty's)?
Well, I don't know. I don't have a strong enough intuition on this case, so it could never be a deal-breaker for me. Still, what do I want to say about it? It could be a really strange case between asymmetric and symmetric redundancy. I think it's not too much of a stretch to think of the two rocks as a single system at that point: one with a primary trigger and a backup trigger. Come to think about it, that might be what I want to say in the first place.
Labels: causal relata, causation, David Lewis, events, fragility, Laurie Hall, relata
<< Home