The second difficulty, closely related to the first, is the inherent complexity of Industry, the continual and close interaction of a number of phenomena whose exact size and relative importance is continually shifting and baffles the keenest observer.

These difficulties, common to all sciences, are enhanced in sociological sciences by the impossibility of adequate experiment in specially prepared environments.

The Evolution of Modern Capitalism — Part 7

The degree of exactitude attainable in industrial sciences may thus appear to be limited by the development of statistical inquiry. Since the collection of accurate statistics, even on those matters which are most important, and which lend themselves most easily to statistical description, is a modern acquirement which has not yet widely spread over the whole world, while the capacity for classifying and making right use of statistics is still rarer, it is held by some that in a study where so much depends upon accurate statements of quantity little advance is at present possible.

The Evolution of Modern Capitalism — Part 7

And it is, of course, true that until the advance of organised curiosity has provided us with a complete measurement of industrial phenomena over a wide area of commerce and over a considerable period of time, the inductive science of Economics cannot approach exactitude.

The Evolution of Modern Capitalism — Part 7

But a study which cannot claim this exactness may yet be a science, and may have its value. A hypothesis which best explains the generally apparent relation between certain known phenomena is not the less science because it is liable to be succeeded by other hypotheses which with equal relative accuracy explain a wider range of similar phenomena. It is true that in studies where we know that there exists a number of unascertained factors we shall expect a more fundamental displacement of earlier and more speculative hypotheses than in studies where we know, or think we know, that most of the phenomena with which we are concerned are equally within our ken: but the earlier scientific treatment, so far as it goes, is equally necessary and equally scientific.

In modern industrial changes many different factors, material and moral, are discernibly related to one another in many complex ways. According as one or other of the leading factors is taken for a scientific objective the study assumes a widely different character.