ON GENERIC EPISTEMOLOGY