Yes, the issue of the epistemic/epistemology vs. the semantic/semantics is complex. One way of framing it is to say that the epistemic/epistemology is about beliefs & their justification, or reasons for believing, including knowledge & certainty, especially in the face of skeptical challenges, whereas the semantic/semantics is about meaning, which includes reference, sense, & truth. Put that way, one could apparently focus on the former without paying attention to the latter, & many philosophers, especially the neo-Kantians & classical epistemologists in the Analytic tradition, have done that. But in fact it’s much more complicated, since beliefs must meet various logical & semantic conditions (consistency, well-formedness, referentiality, conceptual coherence, etc.) in order to be fully meaningful & count as truth-bearers; & of course knowledge is usually defined as sufficiently justified true belief, which presupposes truth. So the epistemic/epistemology presupposes the semantic/semantics. In a logical context, after Goedel & Tarski had shown that proof-theoretic definitions of truth fail, then formal semantics focused heavily on reference (objects) & models (configurations of objects constituting actual facts = truth-makers). Looked at that way, it might seem possible to do semantics without any connection to the mind. But of course the Kantian & phenomenological traditions focus on mental representation & intentionality, so they hold that cognitive semantics is more basic than formal semantics. And even Frege’s notion of sense as “the mode of presentation of the reference/referent,” is explicitly cognitive. Also, from a cognitive-semantic point of view, the fact that cognitive semantics is presupposed by epistemology is obvious, since belief is a mode of human cognition….