Realism and Justified Belief

I see no inconsistency in being a realist about what there is, and a negative coherentist about what one is justified in believing about it. In any case, positive coherence and narrow foundationalism are no guarantee of truth either. That is, it is imperfectly consistent with a belief's being justified in any of these ways, that it be false. If we are to follow Descartes in taking the possibility of a belief's being false as incompatible with its being justified, we will end up in skepticism pretty quickly. 

There are in general three ways to meet the skeptical challenge by bridging the gap between knowledge and objective reality. The first is to maintain extremely high standards for what will count as knowledge, and extremely high standards for what will count as objective reality, and nevertheless heroically attempt to bridge the gulf. This was Descartes' approach, and it is generally agreed that it failed. In its wake came two responses: One response, that of Berkeley, Kant, and the recent Putnam (but ultimately going  back to Protagoras) is to lower the standards for what will count as objective reality. The other response, employed by such philosophers as Aristotle and [Thomas] Reid (and to a lesser degree, and in his less skeptical moods, Hume), is to leave reality as it is but lower the standards for what will count as knowledge. And this seems most reasonable. When knowledge and reality go dancing, it had better be reality that leads.

--Roderick T. Long, Reason and Value: Aristotle versus Rand