Page:Popular Science Monthly Volume 78.djvu/117

This page has been proofread, but needs to be validated.

he shall gain a full explanation of all the observed facts; that is, an understanding of the manner in which they have been produced. While the invented counterpart remains of uncertain value, because untested, it is often called an hypothesis; and its uncertainty may be further indicated by calling it a provisional, or a working hypothesis. If later on, it survives all the tests that can be applied to it, it is then usually called a theory; or in order to emphasize its proved value, an established theory. But it is never, so far as the unseen facts of the past are concerned, anything more than the mental counterpart of those facts. Indeed, inasmuch as an hypothesis, when first invented, is usually based on only a few of the observed facts, it will then be only the counterpart of a few of the facts of the past, or of some general principle that suggests the genetic relation of partial groups of facts, past and present. Much more than the mere invention of such an hypothesis must be done before a complete explanation of all the facts is reached; and it is through the additional work, by which supplementary facts and fancies are correlated, that an invented hypothesis ia tested.

As soon as the tentative nature of an hypothesis is understood and its possible failure is recognized, the investigator should realize that he must not stop inventing when his first hypothesis is brought forth; he must urge his subconscious mind to continue bringing forth inventions as actively and ingeniously as possible. He must thus equip himself with several rival working hypotheses,[1] to each of which he must give warm welcome and impartial friendship, but to none of which must he offer special protection or advocacy. The defence for a hypothesis is provided chiefly from new details that are added to it after its invention, or by new facts which are brought to light by its aid. If no defense of this kind is found, the hypothesis must be regarded as only a tentative speculation.

Deduction of the Consequences of an Hypothesis.—Before any decision as to the truth of a hypothesis is attempted, the question must be asked: What consequences must it have in addition to those facts which it was made to explain? An altogether new faculty is now called into play, the faculty of deduction, by which the consequences of a hypothesis are logically worked out. Here again experimentation is extremely useful in physics and chemistry, and it is coming to be more useful than it has been in biology and geology; but in the study of land forms experiment is at present rather imitative than demonstrative, and it will not be further considered here. What we have to examine now is a logical faculty that can be more consciously used than invention, but one which, unlike observation, can be carried

  1. T. C. Chamberlin, "The Method of Multiple Working Hypotheses," Journ. Geol., v., 1897, 837-848.