But doesn't Clark's statement beg the question regarding the ontological nature of knowledge? What is knowledge? Of course, I could then ask how one "knows" what knowledge is, but this question too entails a certain view of knowledge on the part of the questioner.
I have written posts on preconditions for knowledge: logic, language, an omniscient and self-authenticating source, etc. But they are preconditions for a certain kind of knowledge, knowledge as propositional belief in which the possibility of error is precluded. But suppose someone says they reject that anyone can know in that sense. In fact, most people I discuss with say something like, "In the sense of your extreme, radical, Cartesian notion of knowledge, we can't know anything." And then they go on to provide a different definition of knowledge, not realizing that their denial of and substitute for knowledge as I have defined it both presuppose knowledge as I have defined it.
So you could say that I think philosophical knowledge is a precondition both for philosophical knowledge and a more colloquial understanding of knowledge. That is, in part, how I know what knowledge is. But again, I can explain how I know what knowledge is if and only if I also know what knowledge is. Knowing what knowledge is, though, requires that I have a metaphysical as well as epistemological position. It doesn't appear to be meaningful to say that one discipline is logically prior to another when at least some metaphysical questions and assertions regarding knowledge are as much preconditions for knowledge as are [some] epistemic questions and assertions.
This doesn't take away from the primacy of epistemology in philosophy, it simply adds metaphysics. Both are necessary; we must still explain how we know, but we also must explain what it is to know. One can't have an epistemic position without a metaphysical one nor a metaphysical position without an epistemic one.