Tuesday, January 8, 2008

Promises and the future

Suppose I find out that tomorrow I will make you a certain promise (maybe God tells me; or maybe I know this by induction, having made you that promise on the second Wednesday of every year for as long as I remember). Moreover, when I think about it, I realize that I can only keep the promise if I do some preparatory work today. What sort of a reason do I have to do the preparatory work today? If I had to do some preparatory work to fulfill a past promise, I would have a reason grounded in the promise--in promising to do something, I become obliged (at least prima facie) to do what it takes to do what was promised. But does this obligation extend backwards in time?

I clearly have some reason to do the preparatory work. I do, after all, have reason to be a keeper of promises. This is, however, a self-interested reason (though the self-interest here is of the unsordid, Aristotelian virtue ethics sort), and is a different kind of reason from the kind of reason I have to keep a promise. I keep a promise to you because I owe it to you, rather than to be a certain kind of person.

Suppose now one of my children makes you a promise, and I know that she will be unable to keep it unless I do something. I then have some reason to do that, and my reason now does concern you. I owe it to you insofar as I am responsible for my children's actions. It seems to me that my reason for doing the preparatory work for keeping my future promise to you is similar to my reason for doing what makes it possible for my child to keep her promise to you. In both cases, the promise is made by someone for whose actions I am responsible. But in both cases, the reason that I have for acting seems different from the reason I have when I have made you a promise in the past.

Does analysis of promises thus show that there is some metaphysical asymmetry of time, with the past metaphysically different from the future? Maybe not. For it may be that promises bind us over the time period over which they are intended to bind us. Promises create consented-to obligations. Typically, when we make a promise, we are not consenting to bind ourselves in the past, because we typically have no way of communicating the fact of the promise to the past. If backwards causation were possible to us, however, then maybe it would not be so absurd to suppose that a promise could pastly bind. Suppose I have a transtemporal communicator. In the morning I come across a note from the future: "Alex: Send George a check for $100 per the promise of February 16, 2043. Best wishes, Alex". Maybe I really would be bound?

4 comments:

Martin Cooke said...

You wrote the note knowing that you did get such a note, and did pay George? If so, then suppose that now you don't pay; the writer of the note (in 2043) would not then be you (since you are not then going to have such a past), so you would not be obliged to pay. Or if not, then why (in 2043) would you have thought that your past self could have paid? You may have thought that you could have forgotten; or you may have thought that you could have a different past and still be sufficiently the same person to have made the promise. That latter is odd because prima facie the making of a promise is also the kind of thing that could itself turn out to have not happened (e.g. are we obliged by promises that were not made?!); while the former is that you make a promise because you hope that (for some reason that you've also forgotten) you've forgotten that you've already kept it, and then you try (by writing the note) to become part of that reason...

Martin Cooke said...

...and in that last case, you were not really in a position to promise to do it, since all you really had was the hope that you'd already done it!

Jeremy Pierce said...

Given a B-theory of time, there's no logical inconsistency here, just a psychological one if you know whether you wrote the note. It does seem hard to imagine the point of writing the note. If you wrote it, what's the point of writing the note? Possibly you remember being motivated to do it by receiving the note, so you write the note to fulfill the past. If you didn't write the note, however, there's also a reason you might write it. You might not be a B-theorist and might think (per impossibile) that you can change the past.

But there's one other scenario where it would make sense to write the note. Perhaps future you had undergone memory loss, though, and don't remember if you wrote the check. You then make the promise and write the note. When you in the present receive the note, you can then make your decision and thus determine whether it's true at the time of the note-writing that you kept the promise.

Alexander R Pruss said...

Yes, memory loss does need to be posited for the note to make sense.