I finally got around to reading the Hall and Paul paper on causation and preemption that ends with the invective against cheap counterexamples. And…I have a cheap counterexample. One of the theses that does a lot of work in the paper is .

Intrinsicness: Let S’ be a structure of events consisting of event E’, together with all of its causes back to some earlier time t. Let S be a structure of events that intrinsically matches S’ in relevant respects, and that exists in a world with the same laws. Let E be the event in S that corresponds to E’ in S’. Let C’ be some event in S’ distinct from E’, and let C be the event in S that corresponds to C’. Then C is a cause of E.

Well, here’s the counterexample. Let Brian- be me minus a small part of the end of the nail on my left big toe. In this world @, I throw a bottle at the wall, and it shatters. The cause of the bottle’s shattering is my throwing it at the wall, not Brian-’s ‘throwing’ it at the wall. More precisely, his movements are not a throwing, though they seem to constitute an event. In world w´, I trimmed that toenail last night, so I am now a duplicate of Brian-. I throw the same bottle at the same wall with the same force and direction at the same time. In this world my throwing is a cause of the bottle’s shattering. But it is a duplicate event of Brian-’s ‘throwing’ in this world, which is not a cause.

Perhaps it is not so cheap though. There is a deepish philosophical point here. Whether an event counts as a cause seems to depend on which naturalish properties it instantiates. And, as Ted Sider has pointed out from time to time, naturalish properties are almost always extrinsic because they are almost always maximal. So these cases can be multiplied with ease – the only problem is finding one that is sufficiently amusing.

Note also that the ‘relevant respects’ qualifier won’t help here. The two events, one them a cause, the other a non-cause, are perfect duplicates, so they match in all intrinsic features.

Ned and Laurie note that there is one well-known class of counterexamples to Intrinsicness, cases of double prevention. But they hold out some hope that this can be dealt with by positing two kinds of causation. If maximality related problems generalise, however, Intrinsicness will turn out not to be true of any concept of causation.