Michael Huemer’s paper on Epistemic Possibility — 3 Comments

  1. Mike–Thanks for sending the paper; it’s very nice and I enjoyed it very much. I especially enjoyed the Skeezix example!

    A couple of minor issues to consider:
    1. The last paragraph on page 1: I think of metaphysical necessity as broader than logical, in the sense that logical necessity implies metaphysical, but not vice-versa. There is some controversy about this, and you may think it goes the other way?

    2. On D8: I wonder if you don’t want another variable in the account, to make explicit the relativity to context of the notion of dismissal. If you take the same sentence or proposition and use it in the Airport case and in the Epistemology Class case, dismissal will be appropriate in one case but not in the other, on your view. So appropriateness of dismissal is a relation between the quality of justification and the context in question.

    Viewed in this way, your proposal replaces DeRose’s notion of a relevant way to come to know with a practical rationality condition in terms of what it is practically reasonable to ignore in a given context. Does that sound right?

    2. On p. 13, you distinguish the reliance on counterfactuals in the Skeezix case from the unconscious Sam case. There may be an easier way around the problem, and that would be useful if true, since whenever counterfactuals appear, there’s the Shope problem. In the Skeezix case, the issue is what inferential connections are immediate, direct, or obvious to Sam. One option is the one you take: what would be obvious to him if he considered the question of the existence of Rigel 7. But then there’s the usual problem with counterfactuals: Sam might be weird enough that the thought of outer space causes mental apoplexy. I think this problem may be able to be avoided in the following way. If we ask what it is about Rigel 7 that will make it obvious to Sam that Skeezix is not Rigel 7, the answer is obvious: Rigel 7 is a star and it’s obvious to Sam that cats are not stars. So there is something already obvious to Sam, and an inference pattern that is also already obvious to Sam (No S is P, x is an S, so x is not a P) that implies an underlying epistemic condition that makes it appropriate to dismiss the idea that Skeezix is Rigel 7, where the latter is a name of a star. It is this underlying epistemic condition that explains why, in the normal case, if Sam were to consider the claim in question, he’d dismiss it, and given the underlying epistemic condition, you should be able to appeal to it rather than the counterfactual to explain the difference.

    In fact, in the very next paragraph, you say as much when you point out that in the unconscious Sam case, there is nothing Sam now knows that explains the counterfactual in question. So I think you should be able to appeal to the underlying epistemic condition of Sam to explain the Rigel 7 case without relying on the counterfactual. At the very least, one could rely on the counterfactual-as-explained-by-the-underlying-epistemic-condition to avoid some of the Shope problems.

  2. Jon,
    Thanks very much for your comments.

    1. On page 1: Well, a couple of points are relevant. First, I was supposing someone might say that logical possibility is broader than metaphysical possibility; by that I meant that there are more logical possibilities than metaphysical possibilities (so this agrees with what you suggested, except in different words).
    Second, however, I actually don’t believe in this distinction. I believe in metaphysical possibility; I don’t think there is anything broader than that that is an interesting sense of “possibility”.

    2. Yes, I think you’re right.

    3. (Or the second #2): You’re right here too. Even when I was writing it, I suspected something like that would happen–that there would be a problem with any use of counterfactuals.
    Actually, I am not sure your solution fixes it. How should we understand the notion of an inference pattern being obvious to S? Should that mean that S believes that the inference pattern is valid? Or would believe that if he were to consider it? Suppose S frequently makes correct inferences of the Barbara syllogism type, but, having had no training in Aristotelian logic, does not have the concept of this inference pattern (S has only considered particular instances of it, each of which he finds compelling). So S doesn’t actually have any beliefs about this inference pattern. And suppose S is so strange that, if someone were to instruct him in Aristotelian logic, he’d have a stroke and die. Etc.
    I suspect that there may be no way to solve this sort of problem, without fudging. Perhaps one could use fudge words like “normal”. Or perhaps we should abstract from S’s peculiarities and say something like “a person of S’s knowledge, intelligence, etc. would find it obvious that …”

  3. I hadn’t notice my math deficiency! Anyway, yes, I think the worry about an inference pattern being obvious is a good one, and maybe there’s no hope avoiding recourse to counterfactuals. I think I’d still try to resist, though. I think I’d say that I can test whether you find an inference pattern obvious by seeing what you would assent to if queried, but that what I’m testing is something in the realm of mental commitments, attitudes, and dispositions (though perhaps not a belief). And if it is an actually existing something-or-other in the mental realm, then one can cite the counterfactual as the standard or distinctive testing method for such, without defining it in terms of the counterfactual.

Leave a Reply

Your email address will not be published. Required fields are marked *