I'm a novice here, but it seems to me that (based on Tarski or Godel or whomever) if we want to decide what the subject matter of logic is, we cannot use the system of logic for the definition. I.e. we need a kind of metalogic--which we might call metaphysics (not in the sense of a religious or spiritual realm, but simply a realm that is above the physical / logical realm). However, then the problem presents itself: "How do we define this metaphysical system of rules?" And so it's tempting to use Alfred Jarry's terminology "pataphysical." But again, how do we define the rules / system of pataphysics? Dot dot dot infinite regress. Hence my conclusion: The subject matter of logic may be intuited (whatever that means) but it cannot be defined.