Matt Weiner Opiniatrety makes some very good points about how the temporal stability of a belief plays into questions concerning practical reason; roughly, temporal stability is important to the extent that it underwrites our capacity to carry out plans of action. I agree.
In the previous post, I spent some time defining a notion, not of temporal, but of modal stability. Roughly, x's belief that p is modally unstable at t iff in at least some (contextually determined) nearby worlds, it is false that x believes that p at t. Now modally unstable beliefs don't ipso facto generate the same sorts of problems for practical reasoning as temporally unstable beliefs, since modally unstable beliefs can be quite temporally stable. (It is possible, however, that all temporally unstable beliefs are also modally unstable.)
Question: Is there any similar sort of practical infelicity associated with modally unstable beliefs?
Here is a possible case.
X has a justified true belief that his fiance has been faithful and, as a result, he decides to go through with the marriage. Suppose that the fiance's fidelity had been reasonably, though mistakenly, called into question. Moreover, X's belief (and, hence, decision) is based on the failure of a certain anonymous individual Mr. Y to show up with some purportedly incriminating pictures. Suppose that Mr. Y failed to show up only because, unbeknowst to X, he was struck by a car and killed on his way to a meeting with X to hand over the pictures.I suppose that X's belief that his fiance is faithful is modally unstable because there is a relevant nearby world W in which Mr. Y made it to the meeting and X came to believe (understandably, but irrationally) that his fiance had been unfaithful. In Continuation 1, moreover, X's justification is also modally unstable--there are relevant nearby worlds in which X is not justified in believing that his fiance has been faithful; but in Continuation 2, X's justification is modally stable.
Continuation 1: The pictures are such that they fairly decisively tip the evidental balance in favor of the conclusion that the fiance was unfaithful. Thus, all the evidence that could plausible have been available to X point to calling off the marriage. The marriage is saved only because a fluke accident keeps X from making a fully informed decision.
Continuation 2: The pictures are such that, though incriminating, they do not in fact tip the evidential balance to the conclusion that the fiance was unfaithful; that is, though incriminating, the pictures do not in fact constitute defeaters for X's belief. But, finally, suppose that X is so disposed that, had he seen the pictures, he would have (understandably, but irrationally) come to believe that his fiance was unfaithful and called off the wedding.
Now (sappines about true love and destiny aside!) it seems to me that in both cases X's decision to marry may well be infelecitous. In Continuation 1, X's decision is in some sense inappropriately firm because it is grounded on a quirky contraction of the evidence that would reasonably have been available to him; in Continuation 2, X's decision is inappropriately firm because it is grounded on a quirky masking of X's reaction to the evidence that would reasonably have been available to him. Indeed, to see that this take on C2 might be reasonalbe, consider the case from the perspective of the fiance. It seems to me very plausible to say that, if she were informed of how X would have reacted to the pictures, then she might well regard the marriage as based on a farce! Better, she might say, for the pictures to have come out and to let the cards fall as they will.
Well, I'm not convinced by these cases. Perhaps somebody else can come up with better ones. Incidentally, does X know that his fiance is faithful in either case?