III. Having now defined what sort of history is under discussion, we may now turn to ask what we are to understand by science. This term is generally considered to be synonymous with classified or organized knowledge. But if we confine ourselves to this meaning of the word science, and if we think that we are studying the history of science when we study the gradual accretion of classified knowledge, we shall not be able to get from our labors much illumination on the subject of culture epochs; for in the early stages of civilization, in the ruder culture epochs, we find no classified knowledge that would now be recognized as science—no laws of nature, no great abstract principles. Yet there must have been, in those barbaric and primitive times, something that bore the earmarks of science—something which could serve as a means of identifying the nature of the culture epoch from the point of view of science. What was this, and how discover it? Is there any characteristic of scientific work—any typical factor which always appears in a scientific investigation, and whose rudiments may be discovered even in so-called uncultivated epochs and in apparently scienceless eras?
Recently it has been suggested that the scientific status of a nation at any epoch may be determined from a study of the kind of problems over which the people puzzled and the way in which they solved them, i. e., problem-solving furnishes a criterion of culture from the point of view of science. This criterion is evidently capable of universal application, since every nation and every individual of every nation has had to meet and to solve problems. Furthermore, problem-solving always involves, to a greater or less extent, the use of the creative imagination; hence this criterion justifies itself in the light of the definition of history just given, since the kind of history that is needed has to be studied through the expressions of that imagination.
One thing more is necessary in order completely to define our criterion of scientific culture, and that is a statement of the conditions under which problem-solving may be classed as scientific. It is probably not necessary here to more than state those conditions, since their meaning is now so well understood. A problem has been solved scientifically when its solution has stood the test of the most unprejudiced and relentless criticism both from the side of reason and from that of experiment; and also, when the limits within which its solution is valid