The epistemological disjunctivist could defend the view that we have better evidence, reasons, or justifiers in the good case than the bad. The disjunctivist could defend the view that we have justified beliefs only in the good case. I’ve primarily been interested in defending that first disjunctivist thesis, but let me sketch an argument for the second disjunctivist thesis.
Here is an argument that we can have justified beliefs in the good case:
(1) In at least some such cases of veridical perception where it looks to S as if p, there is no reason not to believe p.
(2) If there is no reason not to believe p, there is no undefeated reason not to believe p.
(3) If there is no undefeated reason not to believe p, it is not the case that S ought not believe p.
(4) If it is not the case that S ought not believe p, S is permitted to believe p.
(5) It cannot be that S is permitted to believe p if S’s belief that p cannot be justified.
(C) S’s belief that p is justified if it looks to S as if p and S is in the good case.
Here is an argument that we cannot have justified beliefs in the bad case (assuming that the only reasons to believe are provided by experience):
(6) In every case of hallucination where it looks to S as if p, there is a reason that has among its demands that S refrain from believing p on the basis of that hallucination.*
(7) S has no other reason to believe p.
(8) There is always a reason to refrain from believing what you do not have reason to believe.
(9) S can permissibly believe in the face of reasons not to believe only if there is some conflicting reason that has among its demands that S believe p where this reason is at least as strong as the reasons not to believe.
(10) There are no such reasons.
(11) In the bad case, there is always an undefeated reason for S not to believe.
(12) If there is an undefeated reason for S not to believe, S is obliged not to believe.
(13) S’s belief that p cannot be justified if S is obliged to refrain from believing p.
(14) In the bad case, S cannot be justified in believing p even if it looks to S as if p.
One objection to the argument that we can have justified beliefs in the good case I’ve just sketched is that it seems I have shown that it could be that someone justifiably believes p with no reason at all to believe p. Since it seems that justified beliefs have something positive going for them, it might seem there is a mistake contained somewhere in the argument. I think the way to deal with this objection is to show that there is always reason to refrain from believing in the absence of reasons to believe. That would suggest that (1) could be true only if S has something like evidence or a reason to believe p. Williamson suggests that if, say, someone ought not to believe p unless some condition, C, obtains, then they also ought not believe p unless they have some reason to think that C does in fact obtain. At the very least, to do otherwise would be irresponsible and you might think that whenever there is some norm that we are answerable to, we are under some sort of obligation to refrain from being irresponsible in the way someone would who knew that they ought not believe unless C obtains and then believes with no reason to think that this is not one of the cases where C does not obtain.
I think it is the second argument that most will resist. (7) is just true by stipulation. I am arguing that in the absence of any reasons apart from those provided by experience, hallucinatory experience cannot justify belief. I have just explained why I think (8) is plausible in addressing the objection to (1). (9) simply states a connection between reasons and the obligations they give rise to. (10) might be controversial, but if you think it is false you have to show two things. First, you have to show that there are reasons that apply to the subject in the case of hallucination that demand that the subject believe p that are at least as strong as the reasons not to believe p. Such reasons, however, would make it prima facie wrong not to believe p in the absence of reasons not to and I think there are no such reasons.
What about (6)?
Assuming that epistemological disjunctivism is true, hallucinatory experiences will always pass off a non-reason as if it is a reason because the contents of these experiences do not match reality. As such, the contents of these experiences will not be true and so anyone who treats the propositions that is the contents of these experiences as a reason to form a belief about the external world will violate the (alleged) norm that tells us not to treat a non-reason as if it were a reason.
Epistemological disjunctivists could only answer our first question negatively if they were to say that there is no norm of the sort described here. It is hard to see how someone who bought into the disjunctivist picture could seriously maintain that there is no such norm for then facts about what counts as reasons for what would not automatically have normative significance. What then would make these reasons normative reasons?
There is the issue as to whether reasons inaccessible to the subject have an effect on the normative standing of the agent's attitudes. As I've said before, when the reasons are reasons-against, they need not be accessible or play any role in deliberation to have an effect on normative standing. To stick with an example from earlier, upon discovering that an innocent person has been wrongly incarcerated and forced to suffer for crimes they did not commit, everyone thinks we should set this person free and it seems everyone agrees that there is a duty of reparation that needs to be discharged. This is in spite of the fact that there is no one we can necessarily point to who we can say was culpable for the wrong that needs to be righted. If someone said that reasons against affect normative standing only when someone is cognizant of them, it is hard to see how such a person could say that this is a case of wrongdoing that generates a reparative duty as opposed to a case of right action with unfortunate side effects.