[145] and if he doesn’t do these three things, he is not a good historian, and I will give you immediately the consequences of this... The first and basic is that he tries to state those verifiable facts which are important for his enterprise. Now I bring in here the word "important," which I will explain much more. But the first thing is: verifiable facts, verifiable in the very narrow limits of historical probability, which can never be trespassed and which can never reach the state of scientific hypothetical probability. That's the one thing. The second thing he does is: explaining. Explaining means to give an account about the origin of the facts, of the way they came into existence, of their relations. The third: he must UNDERSTAND them. This is the most discussed point in historical work, and the only really worthwhile point in all historical enterprise, namely to understand the MEANING of the explained facts. So we have three things: describing---as it actually has been; explaining---how it came to this; and understanding---what it means. In doing all this, the historian^a behaves differently from the scientist^b. The scientist must remain in a complete detachment from the content of his research; he never can go beyond it. The historian tries to keep the same attitude--and rightly so---with respect to the facts: "in the year so-and-so, Caesar crossed the Rubicon." That is a fact, and this fact must be assured, and it can be assured only by the detached analysis of all the documents which point to this fact. Then he has to explain, and there, a little bit more of participation is involved. It is