INTRODUCTION XIX 



son had already perceived some of the weaknesses of Newton- 

 ian axioms and some of the ambiguities in Newtonian concepts. 

 But they were not wilHng to reject the basic theory of New- 

 tonian science. Even after Planck's paper of 1900 and Einstein's 

 theory of 1905, theoreticians of science, such as Henri Poincare 

 and Pierre Duhem, were unwilling to reject Newtonian prin- 

 ciples as erroneous. Instead they conceived all scientific theories 

 as conventional constructs and approximations of the truth. A 

 scientific theory may be induced from experimental data; its 

 predictions may be verified in every detail. But, for Poincare 

 and Duhem, the theory was only one way out of many for 

 interpreting the data; it was an hypothetical approximation. 

 The same data could be interpreted with equal verification by 

 other hypotheses. The irreconcilability of relativity theory and 

 quantum mechanics, as well as the wave and particle theories 

 of light, gave much weight to this interpretation of scientific 

 theory. 



Later authors, it is true, have considered Poincare's interpre- 

 tation of science and hypothesis to be somewhat naive and over- 

 simplified, and they have rejected certain details of his conven- 

 tionalism {commodisme) . Nevertheless, the fundamental ele- 

 ments of his view have been incorporated into the generally 

 accepted theory of science today. His insistence on the hypo- 

 thetical character of scientific theory has, in fact, been extended 

 by modem theoreticians beyond the limits intended by Poin- 

 care himself. He was willing to grant certainty at least to the 

 first principles of scientific investigation and to other types of 

 knowledge. Obviously, he did not reduce his own philosophical 

 speculations to the status of mere convention and hypothesis. 

 In the currently accepted view of scientific knowledge, ex- 

 pounded in philosophies of science, there are three fundamental 

 points which ought to give us pause. (1) It insists that no 

 scientific knowledge can be taken as absolutely certain, that 

 is, without an intrinsic doubt concerning its alterability. The 

 hypothetical character of all scientific knowledge, it is said, 

 requires that we accept current scientific knowledge on a tenta- 



