[144] can test the experiment by which his predecessor came to a special result and can find out the mistakes or the truth of them. Nothing like this can be done in history^a, not even in biology, not even in the psychology of an individual person, because even the conditions in which an individual lives are beyond repetition: his life process has gone beyond the moment in which you have first observed him. He is not the same tomorrow that he was today. Therefore you cannot subject him to a test of repetition. Or you can do it only in some realms which can be abstracted from his life process, and all test methods are based on such abstractions. But if you want to know the total life of a person, then this total life cannot be repeated in any moment of its whole process. And so it is with history^b as a whole. Historical events are not subject to experiment. But they are subject to an experience which is, to a certain degree verifiable, or verifying. Here we have an analogy to the physical experiment in historical research, namely the dealing with documents. If, concerning the same event, you have two independent documents which point to the same fact, then you have a high degree of probability that the assertion is verified. But I would say "a high degree" of probability---you never have more than this, in ANY historical research, even if it refers to the last year of your own life, and your own personal experiences. You can ask others who were with you, you can ask your memory, you can do a lot of things, but you can never trespass probability, you never can reach that amount of verifiable hypotheses which you can reach in science. What does history^c do, in its work? What does the historian do? He does three things---