The assessor-relativity of my view is meant to be able to (1) account for the fact that ‘Margaret ought to pick three’ is true in Margaret’s context of assessment and false in the insider’s context of assessment and (2) account for the sense in which Margaret and the insider’s judgments are in conflict. Since truth-values are relativized to, inter alia, contexts of assessment, my view can explain (1). But since my view is non-indexical, it can still hold that the insiders are denying the proposition that Margaret is asserting.

That’s the sort of view I had in mind. It’s relativized to justifiers, as it were, rather than the justified.

I’ll wait and see what the assessor relative account looks like, I’m not sure I see how the view is supposed to work. There’s two observations that I take it we want an account to account for. First, that there’s a perfectly good sense in which we think/say ‘Margaret ought to pick 3′. Second, that those with insider’s information can say/think ‘Margaret ought to pick 1′. On the assessor relative view, how do you account for this? Do you try to explain away one or the other observations? On the sort of view I favor (but haven’t really developed in any sort of detail), there is one use of ‘ought’ that is supposed to convey the idea that something is the best we could have hoped for given someone’s information, but it’s not really what the person ought to do (not necessarily, at any rate).

I don’t know of anyone who relativizes only to justified beliefs. Parfit argues against this view at length. And I think for good reason, since relativizing to justified beliefs seems to undercut the initial rationale for relativizing to beliefs in the first place. The standard rationale is that what matters rationally is forming new attitudes that cohere with the attitudes one already has. That’s why Parfit argues that what is important with respect to rationality isn’t the justificatory status of one’s beliefs, it’s the contents of those beliefs. Obviously, relativizing to justified beliefs undermines this rationale. Moreover, I still don’t think it will work. For suppose there is propositional justification for believing that there is $1000 in envelope three, but Margaret simply doesn’t believe it. I still think that she ought to choose envelope three. That said, I think the view that holds that rationality supervenes on the propositions that are propositionally justified given one’s epistemic position is one of two views that might be correct. It’s not, however, the view I prefer.

Anyway, suppose you and I are watching Margaret and we know where the most money is. Margaret picks an envelope. I say, “You should have picked envelope two, but since you couldn’t have known that, you should have picked envelope three”.

Part of the positive things that I want to argue for w.r.t. ought is that ought is assessor-sensitive. So, my view holds that if we are watching Margaret and we know where the money is, the proposition Margaret ought to choose envelope three is false. But the view also can say that from Margaret’s context of assessment that proposition is true. I don’t feel the pull of your paraphrase much at all. I have the strong intuition that really and truly, what Margaret ought to do is choose envelope three. Since ought is assessment-sensitive on my view, I can still say that from our perspective what Margaret really and truly ought to do is choose envelope one.

I like the case, the three envelope example seems a lot like the mineshaft case as both seem to show that subjective-01 doesn’t really characterize any interesting notion of ‘ought’. I think that the point you make about the relation between the subjective ‘ought’ and the information available, but isn’t it typical for people who relativize an ‘ought’ to the subject’s beliefs typically relativize them to reasonable beliefs? It seems this might cover reasonable suspensions of judgment, too.

Anyway, suppose you and I are watching Margaret and we know where the most money is. Margaret picks an envelope. I say, “You should have picked envelope two, but since you couldn’t have known that, you should have picked envelope three”. What this says (roughly) is that while she should have picked envelope two, she could not have known how to do that, so the best we could hope for is that she’d cut her losses and pick three.