NA TURE 



507 



THURSDAY, FEBRUARY 15, 1912. 



THE PHILOSOPHY OF SCIENCE. 

 (i) Natural Philosophy. By W. Ostwald. Translated 



by T. Seltzer. Pp. ix+193. (New York: Henry 



Holt and Co. ; London : Williams and Norgate, 



1910.) Price I dollar net. 

 (2) Prolegomena ziir Naturphilosophie. By Hermann 



Graf Keyserling. Pp. xii+159. (Miinchen : J. F. 



Lehmann's Verlag, igio.) Price 5 marks, 

 (i) "D EADERS of German will remember that 

 1- V Prof. Ostwald published as the first volume 

 of Reclam's series, " Biicher der Naturwissenschaft," 

 a book which is practically a popular summary of his 

 larger "Vorlesungen liber Naturphilosophie." This 

 little work, after special revision by the author, now 

 appears in an American edition. Mr. Seltzer's trans- 

 lation is generally satisfactory, though he has left, 

 here and there, an obscurity of diction which it is 

 difficult to charge to the account of so clear a writer 

 as his original. 



A book in which an investigator of Ostwald's 

 eminence gives in systematic outline his views on the 

 aims, nature, and general methods of science is bound 

 to exercise considerable authority over the opinions 

 of young students and inquiring laymen. It is, there- 

 fore, a matter of considerable importance to deter- 

 mine whether his guidance in these subjects may be 

 recommended with confidence. In spite of great ad- 

 miration for the genius of the distinguished chemist, 

 the reviewer is bound to record on this point an un- 

 favourable opinion. It would be absurd to deny that 

 the book has value. The practical wisdom and in- 

 spiration of a successful man of science inform many 

 of its pages, and may gladly be accepted in lieu of 

 a good deal of correctitude in logic and philosophy. 

 Moreover, Prof. Ostwald's characteristic doctrine of 

 "energetics" leads him in a very direct way to results 

 — such as the limited vafidity of the concept of 

 mechanism — which, though unpopular among physic- 

 ists and chemists, are regarded as of great importance 

 by all students of the philosophy of science. But out- 

 side the thirty-six pages given to the general principles 

 of the physical sciences, the treatment of his topics 

 seems frequently unsatisfactory. His logical and 

 psychological analyses lack precision and thorough- 

 ness, and often suggest that on some questions of 

 fundamental importance he accepts the views of the 

 Mill and Spencer epoch as finally authoritative. 



The pragmatism which made Prof. Ostwald's work 

 so interesting to the late William James appears very 

 early in the book. Strictly speaking, science is, he 

 holds, concerned only with the prediction of future 

 events. The " retrospective prophecy " to which Hux- 

 ley attached equal importance " must take its place 

 with other aimless activities called play " (p. 13). 

 Moreover (as the last phrase indicates) even predic- 

 tions of the future are not properly to be called science 

 unless they bear directly or remotely upon the practical 

 management of human life. At first sight this dictum 

 would seem to confine the legitimate development of 

 some sciences within severely narrow limits. For 

 NO. 2207, VOL. 88] 



example, must the palaeontologist prove that his 

 studies have a bearing (say) upon eugenics before he 

 can be admitted into the company of men of science? 

 In anticipation of such questions, the author is 

 obliged to argue that since we can never know com- 

 pletely "what kind of knowledge we shall next need 

 . . . therefore it is one of the most important 

 functions of science to achieve as perfect an elabora- 

 tion as possible of all the relations conceivable." 

 Thus, judged by the pragmatic test of " alikeness " 

 upon which Prof. Ostwald lays so much stress (p. 52), 

 his definition of the aim and scope of scientific inquiry 

 turns out after all to be identical with the intel- 

 lectualist view upon which he is so severe (p. 13). 



Prof. Ostwald's preoccupation with the practical 

 value of science^ — though a fault which leans to virtue's 

 side— makes him cling to an empiricism which is, to 

 say the least, ddmodi. In spite of an (apparently) 

 wider definition on p. 62, the word "experience" is 

 constantly used as synonymous with " perceive " or 

 "perception." Concepts are simply the "coinciding 

 or repeated parts of similar experiences " — a view 

 which seems identical with Huxley's " composite 

 photograph " theory of general ideas, and is open 

 to the same criticisms. It follows that the degree of 

 certainty reached in reasoning depends upon the 

 number of perceptual experiences upon which the con- 

 cepts are based. Thus the security of the conclusion 

 2 + 3 = 5 is extremely high because the number-concepts 

 are so extremely " general." 



Starting in this way, it is not surprising that Prof. 

 Ostwald repeats Mill's misunderstanding of the 

 syllogism (p. 65), and restricts deduction to the com- 

 paratively small r6le of applying "principles . . . 

 acquired through the ordinary incomplete induction 

 ... to special instances which, at the proposition of 

 the principle, had not been taken into consideration '' 

 (p. 41). It is difficult to regard this as an adequate 

 statement of the aim of Newton's "Principia," or of 

 Maxwell's "Electricity and Magnetism"; yet both 

 these treatises must be held to give a "deductive" 

 treatment of their subject-matter. So anxious is the 

 author to reduce all knowledge to repeated "experi- 

 ences " that he holds mathematics to have been 

 proved to be an empirical science by the fact that 

 certain laws in the theory of numbers have been 

 found empirically and have not yet been proved deduc- 

 tively (p. 56). The same ultra-empiricism seems re- 

 sponsible for the remarkable statement (p. 76) that the 

 power of "a few highly developed individuals," such 

 as Julius Caesar, to "keep up several lines of 

 thought " simultaneously proves that time is not neces- 

 sarily to be conceived as unidimensional. 



It will be seen that, in the reviewer's opinion, 

 Prof. Ostwald's work suffers from the capital defect 

 of misrepresenting seriously the relations between 

 ideas and perceptual experience. It is true that no 

 one has yet formulated an adequate account of these 

 relations. Nevertheless no theory of science can be 

 satisfactory unless it takes account, on one hand, of 

 the criticisms of the inductive process which wo owe 

 to such logicians as Bradley and Bosanquet, and, on 

 the other, of the patient, unbiassed, and pene- 

 trating researches concerning the objects of cognitive 



R 



