Thursday, June 25, 2009

Causation and presentism


  1. Some events have causes that strictly precede them in time (i.e., the two events are never simultaneous).
  2. If E causes F, then both E and F exist.
It follows that presentism is in trouble. For suppose that presentism is true, and that a present event F has a cause E that strictly precedes it in time. Since E strictly precedes F in time, it follows that if F is present, E is not present. And if E is not present then, according to presentism, E does not exist. But if E does not exist, then it does not cause F by (2). Nor can the presentist say that even though E doesn't cause F, E caused F. For at any time at which E might have been said to cause F, by (2) and presentism, both E and F would need to exist, while there is in fact no time at which they both exist.

The presentist might make the following move. Let E* be the state of affairs of its being the case that E occurred. Then, the presentist can deny that E causes F, but affirm that E* causes F, and say that that is close enough to do justice to our intuitions.

But it's not close enough as, surely, the cause of F was an event earlier than F, namely E, not an event like E* simultaneous with F. Granted, E might have acted through an intermediate cause, E1, that is simultaneous with F, but E* is not an intermediate cause.

Moreover, if the cause of F is not E but E*, then what did E do? Let's suppose that E is present and F is going to happen. Then E is causing something. What is it causing? The presentist cannot say it's causing F. The presentist can only say it's causing F*, where F* is the state of affairs of its being the case that F will happen. So, the presentist will need to say that E* causes F and E causes F*.

But now consider this: What connection is there between these two complex states of affairs: (A) E* causes F and (B) E causes F*. They clearly are not independent. There must be some connection between them. This connection cannot be causal if presentism holds, because A is strictly earlier than B. This is puzzling. Is there maybe an explanatory relation between them? Is it, perhaps, the case that E* is causing F because E had caused F*, or maybe the other way around? Neither option seems right. But there should be some explanatory relation between them, or else both should be explained by some third thing. I can't think of what that third thing could be. So let's think whether maybe E* is causing F because E had caused F*, at the time of F. Observe that unless F is an intermediate cause between E and F*, which seems absurd, F had better be prior in the order of explanation to the state of affairs of E having caused F* for the reason that F is prior in the order of explanation to the state of affairs of F* having occurred. So it does not seem possible to make E* be causing F because E had caused F*, because F is prior to F*. Could we say that E was causing F* because E* is causing F? But now the past occurrence of a causal relation ends up happening because of a present occurrence of a causal relation. That does not seem right, either.

In any case, where on eternalist views we had one instance of causation, E causing F, the presentist has cut it up into two, an earlier and a later instance of causation. It is now (let's say) true that E* is causing F, but it was earlier true that E was causing F*. But this is weird: we have two instances of causation in place of one. And instead of the relation relating E with F as it should, what we have is one past causal relation from E to F* and a present one from E* to F.


Jonathan D. Jacobs said...

I offer a first try at a solution here. In short, the presentist shouldn't think of causation as a relation at all.

Of course, the eternalist has trouble accommodating the intuition that causes *produce* their effects in a robust sense, since on that view the effects are there from eternity, so to speak. One will need a weaker notion of production or responsibility than the robust notion that the presentist can offer.

Alexander R Pruss said...

Thanks for the link. That's a clever solution. I particularly like it in the case of creaturely dependence on God.

So, at t0, when the cause is present, the effect does not exist. Then at t1, when the cause is no longer present, the effect does exist. That is not yet production: that is simply a succession of non-existence followed by existence. To get production, you need the claim that the later existence is explained by the exercise of the causal powers of the cause. But of course the eternalist will say that, too. The only difference is that the eternalist makes the cause be responsible for its always having been the case that the effect exists simpliciter. So the eternalist makes the cause be responsible for more than the presentist does. That does not sound like a less robust notion of responsibility.

There is, of course, a difference in that according to our presentist (assuming backwards causation doesn't happen) the effect doesn't already exist simpliciter before the cause acts. But that non-existence isn't something the cause is responsible for, so this non-existence fact does not seem to add at all to the cause's responsibility.

I see nothing less robust about producing an effect in the present than about producing an effect in the future.

Jonathan D. Jacobs said...

The eternalist will no doubt say that the existence of the effect is explained by the exercise of a power of the cause. The question is whether her claim is as plausible as the analogous claim by the presentist. I admit I don't have an argument here, but the presentists claim seems more plausible—probably for reasons not independent of those that make presentism itself more plausible to me. For example, it won't be true, on the eternalist view, that the *present* existence of the effect is explained by just moments ago occurrence of the cause

Whether the eternalist makes the cause responsible for more is, to me, doubtful. It seems it just makes it responsible for the same amount, just for a whole lot longer.

True enough, the cause is not responsible for the effect's not having existed, but it is responsible for its coming to exist, something the eternalist can't offer. (God's continual sustenance of the universe complicates this a bit, but not, I think, in a way that makes much of a difference.)

There are, it seems to me, genuine issues here about production that, thanks to this conversation, I'm just starting to get my head around. (I made this claim to a bunch of philosophers recently—that presentists have a more robust sense of production—and I was met with some skepticism along similar lines.)

Alexander R Pruss said...

It seems that to have come to exist is either (a) to exist and for there to be a lower bound to the set { t : one existed at t }, or (b) to exist and to have once never existed.

In the first sense, the eternalist can admit that objects come to exist. In the second sense, the eternalist can do so as well, as long as "once never existed" is understood temporally rather than simpliciter.

However, let us bracket the question whether the eternalist has objects come to exist. Observe that in both senses, to have come to exist is a conjunctive claim: it is a conjunction of an existence claim together with a past non-existence claim.

Now, when we produce something and thereby are responsible for its coming into existence, we are actually causally responsible only for one of the conjuncts: the existence conjunct. The non-existence conjunct is not an effect of our causal powers.