[239] NOT answered its OWN question, "How do l know?" I think the way is always the other way around. First we encounter reality, then we know something about it----this something is put under scrutiny and verification; and then we abstract from these processes the methods of scrutiny and verification. But if we go then to the way around and first discuss the question "How it is possible to know" BEFORE we have started to know, then we remain in a vicious circle. And l would like to [apply] the same criticism to a logic which is non-ontological, because a logic which ...[?] inquires [about] the meanings of words, etc. is justified---here too [?] we have an object [?] which is given and can be discussed. But when logic goes beyond this and asks the question "What is the relationship of the logical forms which we USE, and the new ones which we INVENT, to the reality to which they are applied?", then we are in an epistemological question and implicitly in an ontological one. You cannot escape it. And l would like to have the time, once upon a time--- but l am afraid l never will have---to show the metaphysical^a (if you want this word, instead of "ontology")^b implications and presuppositions which are silently present in a seemingly formal epistemology^c and logic. Now that would be my basic answer to this. And of course l cannot go into this. Now I should have the answer of the answer, and we could [continue]. Of course it is one of the shortcomings of these lectures that we do not have more time for discussion. Is there perhaps another question in your mind before we go ahead? I want at least to try to show you that l HAVE answers, even if they are not accepted, or acceptable. It is not MERE POETRY ^d! [smiling -- laughter].