THE question, "Is history a science or can it become a science?" has long both fascinated and irritated historical students. A few think that they have already discovered a science of history, but in reality have made only a premature and primarily speculative attempt at a philosophy of history. Some would limit their ideal to indiscriminate verification of the greatest possible number of "facts" and to indiscriminate exposure of the greatest possible number of historical Actions—sometimes adding the hard requirement of setting forth their unsystematic results in attractive literary form. Some are pessimistic because the historian can not perform experiments in the past and because even his observation is seriously limited by the scantiness of the material at his disposal. This is true, but of course does not exempt the historian from dealing with the data he does have in a scientific spirit. It only renders it the more imperative. Others more hopefully point to the progress made in modern times in the publication and criticism of sources as a sign of history's conversion to scientific method. But of scientific method in the process by which the sources are transformed into history and presented to the public one hears little. On the contrary, one finds prominent writers on historical method not merely admitting but almost rejoicing in, the impossibility of fixed principles of research, of scientific exposition of results.
Bernheim, in his "Lehrbuch der historischen Methode" after designating history as "Wissenschaft" and describing its purpose as not esthetic but informing, in proceeding to speak of the process of converting the sources into history draws all his similes from the fine arts. History is like a pianoforte rendition of an orchestral performance; it should, like a painting, make use of perspective; in it, as in a drama, the characters should lie silent part of the time; gaps to fill in as between the scenes of a play should be left to the reader's imagination. In short, the turning out of the finished product is a fine art according to Bernheim, despite his denial. He says that it ought not to be poetry, but implies that it must be prose. His model historian's aim is to present the past vividly, not necessarily to prove anything. He should give specific bits from the sources occasionally, but more in order to make a story realistic than to make an exposition scientific. This story should follow either chronology or geography or "the logic of events."