A brief essay on the conditions for the possibility that philosophy is not complete bullshit:
Epistemology may be practiced as a subdivision of metaphysics, inquiring after the ontological dimensions of knowledge, or it may be motivated by a practical concern for actually having knowledge. In the former capacity, it is dependent on the fundamental findings of metaphysics. In the latter, it is prerequisite to the sure success of metaphysics itself.
While this order may seem circular, it is not viciously so: epistemology as a pursuit of method does not necessarily require a comprehensive metaphysical self-interpretation, just as one does not need to understand the physical principles of buoyancy, muscle flexion, and action and reaction in order to learn how to swim. On the other side, metaphysics can make positive progress even before its methods are perfectly established, in something of the way the earliest astronomers began to gather observations about the celestial sphere.
To insist on a comprehensive interrelation of metaphysics and epistemology would indeed create a vicious circle and force the whole enterprise of philosophy to collapse. Thus, an understanding of the independence of epistemology from metaphysics is essential for the justification of epistemology and ultimately of an epistemologically funded metaphysics (if that is desirable), which would include the metaphysics of knowledge itself.
Should this epistemologically funded metaphysics of knowledge turn out to undermine or problematize the independence of epistemology by producing methodologically significant conclusions, this cannot be allowed to destroy the legitimacy of the starting point. Thus, the independence of methodologically oriented epistemology must be defined in such a way that its viability does not depend on its comprehensiveness or accuracy.
The point of all this is…well…something to do with Hegel, I think.