approvingly Karl Pearson's characterization of scientific laws as "conceptual shorthand." Or as he himself expresses it, "the conception of mechanism enables us to summarize details that would otherwise bewilder us," but "this cannot possibly nullify our independence." "Such conceptions may furnish an admirable descriptive scheme of 'the motions that occur in nature,' but they explain nothing." "In short, one may take it as definitely conceded by the physicists themselves that descriptive hypothesis takes the place of real theory."

But what can this disparagement of description possibly mean? Is it possible to mention any motive of thought more completely governed by the nature of its subjectmatter than the motive of description? Description means the reporting of things as they are found. The gradual substitution, in the procedure of Science, of description for 'explanation,' means simply that science has grown more rigorously empirical. 'Explanation,' as contrasted with description, suggests a reference to trans-experiential powers, and mysterious essences, or a one-sided version of things in terms of human interests. Science has abandoned explanation in this sense, because such attempts diverted the attention from its proper subject-matter, and engaged it in irrelevant speculation. If we are to believe some of the critics of science, description is a sort of game, and the adoption of this method a sort of senile playfulness that has overtaken science in its degeneracy. It happens, however, that this descriptive period of science is the period of its most brilliant successes. And science is of all branches of human knowledge the one in which caprice is most fatal. For science is engaged at close quarters; dealing as it does with the proximate environment, its findings are promptly verified, or discredited; its day of judgment is always near at hand. It is impossible that science should have succeeded,

1 James Ward: op. cit., Vol. I, pp. 304-305, 83; Vol. II, pp. 251, 88–89, 73.

2 See above, pp. 53, 54.

save by a scrupulous fidelity to fact. This is what the descriptive method properly signifies. It is a discriminating disregard of the irrelevant, and a single-minded renunciation of ulterior motives.

And yet Professor Ward would have us believe that description is somehow arbitrary, that it does not necessarily reflect the nature of things. "To suppose," he says, "that the rigorous determinism deducible from the abstract scheme for the simple reason that it has been put into its fundamental premises must apply also to the real world it has been devised to describe, is just as absurd as to take a very trivial illustration - it would be to say that a man must fit his coat, and not that the coat must fit the man." As though a coat could be fitted to a man without the man's fitting the coat, or a scheme be "devised to describe," the real world without "applying" to it!

The Ideal of
Descriptive
Economy

§ 6. But what, it may be objected, are we to make of the formal criteria of the descriptive method, such, e.g., as simplicity? Is this not, after all, an æsthetic or subjective criterion, a matter of convenience, rather than a revelation of reality? Professor Ward can quote scientists, in their capacity as exponents of naturalism, in support of such a view. But does science justify such a view?

In the first place, it is necessary to distinguish within the system of science itself, between written symbols or signs, and the concepts, ratios, and laws to which they refer. There is evidently a difference between the Greek letter π, or the mark, and what these signs mean. Signs are conventions, arbitrarily chosen and agreed on; and their abbreviation of complexity is a matter of convenience. But this does not in the least affect the status of the things which the signs mean. Because the signs which I use in the equation, 2+2=4, are arabic, lower font, etc., I am not justified in concluding that the numerical equality expressed is similarly contingent on the choice of language and type. 1Op. cit., Vol. II, pp. 67-68.

Yet this confusion, obvious as it is, has played no small part in the notion that descriptive analysis is artificial and unreal.1

If it be admitted that the formulas of scientific description express definite logical and mathematical relationships, whose meaning and truth is independent of the exigencies of discourse, it may yet be contended that the application of these relationships to nature is arbitrary. I can only reply that just these relations are found to subsist in nature; if they were not, the scientist would not account them verified. If it be objected that nature never exactly corresponds to such formula, I may then ask for specific cases. And when the disparity between the case and the formula is pointed out, some new and similar formula will be at the same time exhibited.2

But, it may be asked, does not the formula always leave something out; does it not, for the sake of practical convenience, always over-simplify nature? Of course it leaves something out. In empirical procedure, it is as important to omit the irrelevant as to include that which is germane. And it is further true, as has been stated above, that science is peculiarly, if not exclusively, interested in discovering identities and constants. And these find expression in the formulas of science to the exclusion of individual differences. But it does not follow that this procedure involves over-simplification. For that would mean either that the formulas omit something which they intend to cover; or that the identities and constants they do cover are not actually present in nature. But neither of the charges can be substantiated. Science abstracts, but does so deliberately. And to abstract is not to invent or falsify -but only to discriminate and select.

1 See below, pp. 232-234. See below, pp. 236–237. See pp. 54-55. Were science to assert that nature is only what is expressed in the formula, it would be guilty of what James calls "vicious intellectualism.” As a matter of fact science makes no such assertion. On the contrary it specifically provides for individual differences by its use of 'variables.' See below, pp. 234-235.

It will appear, in short, that the ideal of 'descriptive economy' is not a fantastic hobby, but a canon of knowledge. The discovery of this ideal has not debased science, but has enriched logic and methodology. Through adopting it, science has not departed from reality, but has acquired a closer and more sure grasp of reality.

§ 7. There is one further charge against the descriptive method, that is held to involve not only physical science, The Option of but logic and mathematics as well. It is said Hypotheses that the choice of hypotheses is optional.1 Now as respects physical science, it is clear that this option has to do with the preliminary stages of investigation, and not with the conclusion finally adopted. The trial of a hypothesis is optional; but its success, or verification, is determined. Furthermore, the internal relations of the hypothesis itself are determined. The hypothesis selected for trial must be logically and mathematically correct.

But it may now be urged that logical and mathematical correctness is optional. And this consideration assumes a growing importance in the light of recent developments in the philosophy of mathematics. It is often said that logical and mathematical truths depend on the arbitrary selection of postulates. Time will show, I believe, that such expressions are one-sided, and, when taken unqualifiedly, misleading. There are evidently compensating considerations. In the first place, no logician and mathematician, however modern he may be, invents postulates in order to build systems on them; like the physical scientist, he searches for the postulates that will determine certain facts. As a recent writer expresses it, while postulates are not necessary from," they are necessary for; namely, for

1 Cf. e.g. F. C. S. Schiller: "Axioms as Postulates," in Personal Idealism. Cf. on the other hand, T. P. Nunn: The Aims of Scientific Method, Ch. V. Cf. e.g. E. V. Huntington: "Sets of Independent Postulates for the Algebra of Logic," Transactions of the Amer. Math. Soc., Vol. V, 1904. "These postulates are simply conditions arbitrarily imposed on the fundamental concepts," etc. (p. 290). Cf. also Poincaré: Science and Hypothesis,

pp. 37-39.

the solution of the problem." So postulates are in the end verified, and not merely chosen. In the second place, there are well recognized canons or criteria, by which postulates may be judged, such as 'purity,' 'consistency,' 'independence,' etc. And finally, all systems, whether the postulates be chosen or not, are made up of terms, relations, propositions and implications, which, whatever is done with them, are certainly not chosen to be what they are. In short, here, as elsewhere, thought accommodates itself to things, and its option is confined to selection from among them.

The 'Real' Cause, and 'Mere' Description

§ 8. In the background of every mind that hesitates to accept the descriptive method as valid and adequate, will be found one or both of the notions of explanation which science has gradually abandoned, the notion of 'power' or the notion of 'good.'' More commonly the two will be fused in the notion of 'activity.' This is regarded as the real cause, by which 'mere' description is judged and found wanting. It becomes a question as to whether the development of scientific method has thrown light on the meaning of 'cause'; or has simply abandoned it. The answer depends, evidently, on the validity of this extrascientific notion of cause, which science once employed, and which is now defended by the critics of science.

The notion depends entirely upon the inner experience of activity. Fortunately this issue cannot be argued at length. A man must look for himself, as Hume did, and see whether he finds in the depths of his own nature, a power to do, which is clear, simple, and self-sufficient. He who makes the experiment, and resolutely declines to accept the confusion and vagueness of familiar immediacy as profound insight, will, I believe, conclude as Hume did.

'Karl Schmidt: "Critique of Cognition and its Principles." Jour. of Phil., Psych., and Scientific Methods, Vol. VI (1909), pp. 281–282. 'Cf. Schmidt, op. cit., passim. See above, pp. 53-54. Cf. e.g. James Ward: op. cit., Vol. I, p. 64; Vol. II, pp. 79, 237, 247.