- Inconceivable Indeterminacy.
There are two particular applications of indeterminacy that many of us have trouble getting our heads around. These are indeterminacy about the most fundamental aspects of the world—-what exists, what’s identical to what, where something is located; and what most basic characteristics it has—-and indeterminacy in one’s own survival. Interestingly, in the latter case, there is a first-person/third-person asymmetry: from the outside, it seems obvious there can be indeterminate survival. From the inside, it’s hard to know what to make of it.
What explains the intuitive repugnance of these two potential loci of indeterminacy? Might there be a common source? I’m interested in exploring the idea that there are interestingly strong constraints on conceivability—that to conceive p, we need to form a complete conception of a scenario that is such that p. Indeterminacy in the concepts that frame such scenarios is prima facie incompatible with the kind of “completeness” required. Indeterminacy in the fundamentals may be outlawed if our basic conceptions of scenarios are framed in terms of what exists, what is identical to what, what is where, and what basic features things have. Indeterminacy in personal identity may be outlawed if scenarios are irreducibly “centred” or “de se”.
I think the case that such indeterminacy is inconceivable is only prima facie, though—-but we learn a lot, particularly about the cognitive role of indeterminacy—by exploring the argument.
- Uncertainty about logic.
If logic sets the limits for intelligible doubt, how is it that we can rationally doubt logic? I think that the frameworks for a unified treatment of nonclassicality I’ve been developing can help with this question—for they give us a way to deny that logic’s role is to set the limits of intelligibility in this way. And they allow us to expand the space of reasons in this way without losing all content—since there is still a distinctive set of rational demands on the relation between what one regards as doxastic possibilities, and one’s degrees of belief. (There is another question, then, about whether it’s possible to doubt these new limits of rationality—but however we answer that, we have a picture that makes sense of doubt and deliberation about logic, specifically).
- Accuracy domination and conditionalization.
Alethic arguments for probabilism appeal to dominance: if you violate the norms, then there’s a specific possible belief state such that, no matter what the world is like, it does better than you do on grounds of accuracy. Alethic arguments for conditionalization, however, have taken a less satisfying form. They typically argue that conditionalization minimizes expected inaccuracy. But why should your old expectations be trusted when you have new information available? Shouldn’t we rewrite the old textbooks, rather than take their advice about how to react? I think the stronger dominance idea can be extended. The key is to look at the normative significance of accepting an update strategy—I think it can be understood as accepting certain new norms on one’s first-order beliefs—essentially, treating the state one is committed to move to, on receipt of information E, as a suitable “proxy” for any E-world. This picture allows a derivation of generalized reflection and conditionalization.