Rea has an interesting paper arguing that if one assumes bivalence, then presentism is incompatible with free will. I think there are problems in his argument, but I was also thinking that there might be a variant argument to similar effect.
Start with this. Sally intentionally started a forest fire F, but she ceased to exist before the fire started. (E.g., she set fire to a pool of gasoline, intending the fire to spread to the forest, but ceased to exist before the fire spread to the forest.) The forest fire, then, is plainly something that Sally had a choice about, something that Sally is (or was? -- the grammar of English is unclear on this point) responsible for.
Now:
A1. If x had a choice about the occurrence of E, then E depends on x.
A2. If E never depended, does not depend and will never depend on x, then x never had a choice about the occurrence of E.
A3. Necessarily, if E depends on x, then x exists.
A4. If E depended, depends or will depend on x, then there is a t such that E depended, depends or will depend on x at t.
A5. If E depended, depends or will depend on x at t, then it was, is or will be true at t that E depends on x.
A6. If was, is or will be true at t that E and x exist, then E and x existed, exist or will exist at t. (Presentism)
A7. Necessarily, if E depends on x, then E and x exist.
A8. Suppose for a reductio that Sally had a choice about the occurrence of F.
A9. Therefore, F depended, depends or will depend on Sally. (A2, A8)
A10. Therefore, there is a t at which F depended, depends or will depend on Sally. (A4, A9)
A11. Therefore, there is a t at which it was, is or will be true that F depends on Sally. (A5, A10)
A12. Therefore, there is a t at which it was, is or will be true that F and Sally exist. (A7, A11)
A13. Therefore, there is a t at which F and Sally existed, exist or will exist. (A6, A12)
A14. There is no t at which F and Sally existed, exist or will exist. (Sally ceased to exist before F started.) This contradicts A13!
Now, let t0 be the time at which Sally is setting fire to the gasoline. Then, at t0, the forest fire does not exist. But, perhaps, the state of affairs F* of a forest fire being about to happen exists. And F* depends on Sally. This seems to be the answer that those presentists who believe that truth supervenes on being will give.
It's worth noting that this answer will fail if one accepts open future views, because at t0, F* does not yet exist, since it is not yet settled whether a forest fire will occur. After all, God might miraculously intervene, or weird quantum stuff might happen and the heat might all transmute into a dance of trees.
But suppose we reject open future views. Still, what does this answer come to? It comes to this: Sally had a choice about F*. Yes--but it is F* that we attribute to her, because F* does not exist. What exists now is F, and this Sally had no choice about, because F, unlike F*, never depended on Sally, since never was it the case that both F and Sally exist.
Here is a simpler variant.
B1. Being responsible for F is a property.
B2. One only has properties if one exists.
B3. One only exists if one exists presently. (Presentism)
B4. Sally does not exist presently.
B5. Therefore, Sally does not exist.
B6. Therefore, Sally does not have properties.
B7. Therefore, Sally is not responsible for F.
Now, an obvious retort is that Sally is not responsible for F, but she was was responsible for F. But this fails: for at any time at which F exists, Sally does not, and so Sally never was responsible for F, assuming B1-B3. All we can say is that Sally was responsible for F*. But F* is not such a big deal. It is F we're worried about. (A forest fire's being about to happen is only of interest because it is followed by a forest fire.)
Of course, one might say that our grammar misleads. "Sally is (or was) responsible for the forest fire" should be read as short for: "Sally was responsible for its being the case that a forest fire is about to happen." Open futurists can't say that, but other presentists can. Still, there is a cost for all closed futurist presentists. For they all leave it a mystery how what F has to do with Sally, given that Sally was only responsible, "in the proper sense of the word", for F*. Is it a transitivity, maybe? Sally was responsible for F*, and F* caused F? But how could F* cause F, given that if A causes B, then A exists and B exists, whereas at no time did both F* and F exist, and hence never did both F* and F exist (assuming Presentism).
'Rae' should be 'Rea'.
I think your presentation glosses the instrumental causality of the match, lighter, etc. that Sally uses--even intentionally--to bring about F or F*. The energy that the match, etc. has is the proximate cause of the fire and is a necessary condition for it.
Alex,
This argument seems clearer to me that Rea's. The reply I was going to give is one you give: "Sally is (or was) responsible for the forest fire" should be read as short for: "Sally was responsible for its being the case that a forest fire is about to happen." You then point to a cost of this reply. I think your cost ultimately translates into a general question concerning the relationship between an event of the form x will obtain (E*) and an event of the form x is obtaining (E). There is a sense in which I'd want to say that E* causes or explains E. But if we assume that E obtains just once, then given presentism, there is no time at which E* and E both exist (thus, there is no time at which the one is causally related to the other.)
Cost granted, though I'm not sure how serious it is. I'll have to think about this... My initial reaction is to suggest that the abstract time at which E* occurs bears an explanatory relation to the abstract time at which E occurs. It does so, I suggest, by virtue of the fact that the proposition that x will obtain bears an explanatory relation to the proposition that x obtains. Is there a cost with that reply?
I just noticed that A1 is redundant given A2. And somehow my spelling fix hasn't gone through yet.
RD:
Even when there is an instrumental cause, responsibility and causal dependence bridges across instrumental causes. Plus, we can make up a case where there is no instrumental cause. Maybe Sally has a superpower of causing fires. The way this superpower works is this. She concentrates on intending the fire for five seconds, and then right after the end of the five second interval, the fire happens. Well, suppose that in this case she concentrates on intending the fire for five seconds, and as soon as she is done, not only does the fire happen, but Sally doesn't exist any more.
JR:
It sounds weird to make the relation between F* and F be causal. One thinks of causal relations between spatiotemporal states of affairs as dependent on laws or God's will or the like, while this relation seems to be a matter of logic. It's logically necessary that if F* obtains, then F will obtain.
On the other hand, I find plausible the following thesis: All relations of necessitation between really distinct (no, I don't know how to define "really distinct") states of affairs have a causal basis.
What about the following move? Suppose now it's t1, with the fire raging. Let F1 be the state of affairs of there having been, being or about to be a forest fire at t1. Then F1 obtains presently and F1 obtained back when Sally was intending, and Sally was responsible for F1.
There are two problems with this move, though. (1) On presentist views, F1 is now (at t1) occurring because F is occurring. But Sally is responsible for F derivatively from her responsibility for F1. But these two statements seem to conflict. (2) F1 is a disjunctive state of affairs given presentism, and to bring those in in consideration of responsibility is a cheat. (One can expand on why it's a cheat.)
So, back to F and F*. We can't say that F is occurring because F* is occurring, because F* isn't occurring. So, the question is: Can we say that F is occurring because F* occurred? I think not.
For in fact, on presentist views, the present has a priority. What makes it true that it was the case that there will be a forest fire is the present forest fire (together with whatever makes it be the case that there was time before the forest fire--I will omit this caveat): in other words, F* is (partly) grounded in F. Therefore, F (partly) explains F*. (I will omit the "partly" for simplicity.)
But if so, then F* does not even partly explain F, or we have circularity. (Even if the explanations are of different sorts, this is likely to be problematic.)
Here is another way to see that F* does not explain F but F explaisn F*. There had better be an isomorphism between explanatory and grounding relations at different times. For instance, if right now X is grounded in Y, then ten minutes ago it can't be that Y's being about to happen in 10 minutes is grounded in X's being about to happen in 10 minutes. It would be absurd if right now the knife is hot because its molecules have high kinetic energy, but in five minutes the knife will have had high kinetic energy because they will have been hot.
By the isomorphism, if F* now explains F, then back when Sally was acting, its being the case that F* will occur would explain its being the case that F will occur. Let's expand this out. At the time when Sally is acting, this is true: That a forest fire is about to happen is explained by the fact that it is about to be the case that it was the case that a forest fire is about to happen. But this is really weird. Surely the really complex statement here is instead explained in terms of the much simpler one. So at that time, its being the case that F* will occur is explained by its being the case that F will occur. And so by the isomorphism, right now, F* is explained by F, rather than F being explained by F*.
(In general, presentists need to think more about the tense of "caused". When A is prior to B, and A "causes" B, when is the causing happening: when A is happening or when B is happening?)
I woke up with the thought (realization) that my last comment was misguided. Alex, I'll think about your last note...
See also this post.
Alex,
Previously, I said "My initial reaction is to suggest that the abstract time at which E* occurs bears an explanatory relation to the abstract time at which E occurs." That is still my initial reaction (though what I had said about how the one time explains the other seems to have been misguided). I'll elaborate on it.
If presentism is true, then if X is causally related to Y at some time, then (of course) there is a time at which both X and Y exist. So, suppose that there is no time at which both X and Y exists. There may still be a sense in which Y can be said to have resulted from X. The sense might be something like this: there was a time t at which X causes Z was true; Z's existence is a truth-maker for proposition Px; Py is the proposition that Y exists; were Px not true at some time t0, then Py would not be nearly as likely to be true at some time later time as if Px were true at t0.
This account is supposed to explain how it is (or in what sense it is) that the abstract time at which X exists bears an explanatory relation to the abstract time at which Y exists. The key is to make use of two relations, where one is a causal relation that holds between simultaneous events, and the other is an explanatory relation that holds between abstract propositions describing those events.
Is this a reasonable reply outline for a presentist to make?