Here's a little problem about knowledge and reasons.
It seems on its face that these don't come to the same thing:
(1) If there's reason to A and no reason not to, you ought to A.
(2) If there's reason to A and no reason of which you're aware not to, you ought to A.
I'd accept (1) (perhaps with qualifications), but not (2). Even if you're non-culpably ignorant of the reasons against A-ing, I think they can bear on whether to A and make it all things considered wrong to A. But, I meet plenty of folks who disagree and endorse (2). (I meet some folks who endorse (2) and seem to endorse (1) on the grounds that reasons are the sorts of things you cannot be aware of if you try to attend to them.) I suppose one argument for (2) and one argument against (1) on a reading according to which (1) isn't a consequence of (2) rests on the observation that if (1) were true it would allow for the possibility of a reason bearing on whether to A even when the subject did not know about it and so could not know how to do what she ought and respond to its demands.
The problem with this argument for preferring (2) over (1) is that similar problems arise for (2). Think about cases where a reason, R1, speaks in favor of A-ing and defeats the reason, R2, a reason that favor B-ing rather than A-ing. In this case the subject cannot know how to do what she ought and meet the demands of B-ing for the simple reason that there is no action of hers that would be her doing what she ought by B-ing. Nevertheless, the reason that favors B-ing over A-ing bears on whether she should A.
Cheap, I know, but consider a second case. This time the reasons that favor A-ing conflict with the reasons that favor B-ing rather than A-ing. The subject is fully aware of both sets of considerations but the clash between these reasons is the sort of thing that reasonable folks can disagree about. In fact, this is a case where most reasonable folks could not knowingly judge that one set of reasons determined what ought to be done. Nevertheless, by hypothesis, one of these sets of reasons really does make it the case that the subject ought to do something. However, it does so in spite of the fact that the subject cannot know what ought to be done.
Now, you might grant that this is so, but note that it differs from the first case insofar as the subject in the first case is ignorant of the existence of a reason whereas in the second the subject is ignorant of the significance of that reason when balanced against another. But, I can't see that this could matter to the sort of person that prefers (2) to (1). It's not as if the subjects in these two cases differ in terms of rationality, reasonableness, or responsibility in judging that they ought to do something rightly or wrongly. But, as the fact that the subject is not less than fully rational or reasonable in A-ing in the first case is typically offered as evidence against the possibility of reasons that make it wrong to A, I can't see how this is an available response. So, I don't see what someone should give up to defend (2) over (1).