I was talking with Matt McGrath yesterday about the following problem. If we consider ordinary epistemic principles such as “if you’re appeared to F-ly, without grounds for doubt, then it is reasonable to believe that something is F,” we can distinguish conferrers of justification (in this case, being appeared to F-ly) from enablers (in this case, lacking grounds for doubt). (The question is important especially regarding memorial beliefs which are dispositional and regarding which one no longer possesses the original evidence for the beliefs. It looks like about all that such beliefs having going for them, if one insists that justification must depend on present factors, is the idea that one has no reason to abandon them.)
Here’s one idea, though I think Matt was ultimately not satisfied with it.
Conferrers obey a principle connecting the state of being justified with the practice of justifying a belief. A conferrer of justification is an item one could legitimately cite in defense of a belief. The lack of grounds for doubt can’t be so cited, so it’s not a conferrer of justification.
One problem that worried Matt, and he’s right about this, is the question of how complete a cited defense might be. So if asked, relative to the above principle, one might say, “I’m appeared to F-ly”. But one might also say, “I’m appeared to F-ly and have no reason to doubt that my appearances are accurate in this case.” Both answers are acceptable, and hence the latter conjunction would have to count as a conferrer as well.
This possibility would be a problem if, whenever A&B is a conferrer of justification on p, then so is A and so is B. But that problem can be resolved just by denying &-Elim in this context.
Here’s a different and deeper worry, however. If asked about continuing to believe the kind of memorial belief mentioned in the first paragraph, one likely response would be that you have no reason to abandon the belief. We often defend our actions this way (“because I always eat oatmeal for breakfast”), and it is hard to see why such a defense isn’t legitimate. It can, of course, be challenged, but so can the defense in terms of appearances (“you say you believe something is F because your appeared to F-ly, but that’s a lousy reason unless you supplement it with something else, such as a principle of self-trust”).