Regarding the topic of deontological and non-deontological justification, one way of conceiving of reasons for beliefs is as either (1) not being under obligation not to believe it and so being entitled to believe something or (2) something that arises from some kind of minimally natural process, perhaps a cognitive one. In my view, both could be true. For example, the first could be a kind of explanation for ordinary beliefs. But you could also credit the second as being a kind of scientific explanation for why people believe what they believe.
This next topic concerns what makes the belief reasonable, and two competitors are evidence and reliability. Even a summary examination of these two positions shows that it's not entirely clear what 'evidence' would be or what 'reliability' would be. But here's a shot. Does the belief become reasonable as being arrived at in the right way, according to some kind of cognitive process (reliability), or does a belief become reasonable in light of evidence providing verification or falsification?
This kind of argument seems strange to me since, as with the last distinction, both could be true. Whatever we count as evidence for a belief could provide a reason for a belief and whatever correct cognitive processes we used, consciously or non-consciously (the latter because it's just a natural process), could provide reasons for a belief. The problem is the senses in which we're using these terms are imprecise: 'reasons for a belief,' 'evidence,' 'reliability,' and the like could be made precise or defined in terms of an explanatory theory of knowledge, but to argue at cross purposes without realizing how we're going to use these concepts and make them part of a science of knowing something is pointless and just so much armchair-theorizing.