The question is meant this way: do your ontology and category theory, and then ask what true propositions are there that say that some such things are epistemically justified. Two answers have some plausibility: beliefs can be epistemically justified, and propositions can be epistemically justified. What can’t? Well, a good Chisholmian will say: everything else. (Actually, that’s not quite right: Chisholm says that only cognitive attitudes can be justified, but cognitive attitudes include believings, withholdings, disbelievings, and preferences between the first three. More as an appendix about what led Chisholm to this view.)
This is relevant to an argument F/M want to reject for a strong pragmatic conception of belief, one that maintains that you believe a claim iff your credence for that claim is high enough to be your motivating reason for any action whatsoever.
The argument for this claim that they wish to resist uses the following inference rule:
You are justified in X iff you are justified in Y, so you X iff you Y.
They rightly reject this rule of inference (think of: you are justified in torturing one baby iff you are justified in torturing all babies, so you torture one iff you torture them all).
The question with which I began, however, raises a deeper worry. The premise to which a defender of the strong pragmatic conception of belief wishes to apply the rule of inference above is this:
You are justified in believing p iff you are justified in being such that, your credence for p is high enough for p to be your motivating reason for doing anything.
Suppose, then, that we are good Chisholmians about the scope of epistemic justification. The right side of this equivalence amounts to say that the proposition that p is epistemically justified for you (meaning: preferring belief to disbelief and withholding is required of you). The left side says . . . what? Well, it says, “preferring belief over disbelief and withholding is required of you for the following claim: your credence for p is high enough to be your motivating reason for doing anything.”
But if this is the logical form of the biconditional above, the rule of inference can’t apply to it in support of a pragmatic account of belief (you’d end up with a conclusion like: you believe one thing iff you believe another). To get the applicability of the rule of inference, we need a claim that attributes epistemic justification to the cognitive attitude of belief on one side of a biconditional and attributes epistemic justification to some range of actions on the other side.
But then a Chisholmian will cringe: actions aren’t the right kind of thing to be epistemically justified! If one says that driving on the right side of the road in the USA is epistemically justified, one misspeaks. One can be justified in driving on that side of the road, but not epistemically justified. Epistemic justification obtains, for the Chisholmian, only for cognitive attitudes.
If Chisholm is right, then, the inference rule above is invalid for another reason: the two justification operators on the biconditional premise can’t be the same epistemic justification operator where one side is about actions which are justified.
I’m inclined to side with Chisholm here: no actions can be epistemically justified (including assertions), and neither can hopes, wishes, fears, and desires. Though most of these kinds of things can have a content which is itself epistemically justified (I think Chisholm was wrong about the fundamental things that are epistemically justified–I think they are propositions, whereas he thought that they were preference (types); the difference here doesn’t matter for the present issue, however). So I’m inclined to think the argument used on behalf on the strong pragmatic conception of belief is doubly defective: defective for the reasons F/M use as well as for reasons that the epistemological saint of their graduate program endorses!
APPENDIX: On the Chisholmian View
Chisholm’s view is a product of recognizing the failure of the empiricist program that attempted to generate epistemic connections from logical ones. It was long known that our basic empirical evidence doesn’t entail anything about the external world, but the hope was that one could build evidential connections from inverse logical ones. C.I. Lewis, for example, thought that material object statements entailed hypothetical experiential claims, and perhaps something could be made of building evidential connections from this idea. For Lewis, it went this way: material object statements were equivalent to big collections of hypothetical experiential claims, so if one thinks of probability in terms of partial entailment, one could get the needed probabilities from experience to the world from the logical connections between experience and the world.
Chisholm exploded any hope of this working in “The Problem of Empiricism”: there simply is no entailment of any sort to be found here. So what can work? Well, one might continue to pretend that there is some notion of probability that will do the work, but it won’t be anything recognizable in terms of the Kolmogorov axioms. So best to do what Chisholm does: just admit that the connection is intrinsically normative. Experience makes credible certain claims about the external world, and that is all there is to it.
Chisholm moved from this intrinsic credibility move to insisting on intellectual requirements to prefer one of the three doxastic attitudes when confronted by relevant experiences, but one needn’t go that far to appreciate the limits already in place on the notion of epistemic propriety. For if there are intrinsic credibility links between experience and the world, and nothing else available to connect them epistemically, it is obvious that epistemic justification is constrained within the domain of the cognitive. The credibility in question is credibility given the cognitive or intellectual goal or goals, and the plausible ones here are truth, and something factive like knowledge or semi-factive like (objectual) understanding.
The result, then, is a notion of propriety that is epistemic because it is limited in scope to the cognitive realm. It is for that reason that it is a mistake, according to the Chisholmian perspective, to think that actions can be epistemically justified and to think that affective states such as hopes, wishes, fears, and desires can be epistemically justified. On the latter, it is sufficient to note the direction-of-fit difference: whereas cognition has a world-to-mind direction of fit (as encoded in the goal or aim of cognition involving the truth), affective states have a direction of fit in the opposite direction.
Some may worry here that the view overreaches, ruling out assertions from being epistemically justified. Yes, it does, but there’s no reason to worry: assertions can still have epistemic conditions among those needed for an assertion to be justified in the way actions in general are justified. They just can’t be properly said to be epistemically justified.