372 THE REAL WORLD 



be expected— become the basic laws [i.e., theoretic premises], and the 

 more completely and accurately do they explain the factual course of 

 events. 



The attainability of such simplicity is again a discovery, a fact to be 

 explained. And, like heuristic power, this is again a fact explicable, I 

 think, only on the basis of the postulates I have proposed: though 

 humanly invented and appraised, scientific theories do, in increasing 

 degree, represent the structure of a real world not too complex to be 

 humanly comprehensible. 



A choice between geocentric and heliocentric theories of planetary 

 astronomy cannot be made reliably a priori. However, with the long- 

 term accumulation of obsers^ations, we find that the geocentric sys- 

 tem can be "saved" only by ceaseless expansion of its postulational 

 foundation. We find no necessity whatever for this kind of compli- 

 cation of the heliocentric theory. We readily understand that any 

 theory can always be saved by addition of further postulates : such a 

 theory may be handsome testimony to human inventive talent. But a 

 theory which involves little or no such salvage operation seems to 

 testify to something more than human inventiveness, and so to be 

 singled out by something more than human predilection. The phe- 

 nomenon of calcination offers a parallel example. A priori we have no 

 firm basis for judging whether, in this process, something is added to 

 or subtracted from the metal. But we find that with the long-term 

 accumulation of data phlogiston-type theories become hopelessly 

 complex, while the oxygen theory does not. To dismiss our choice 

 of the oxygen theory as nothing but a matter of human subjectivity 

 is then nothing but nonsense. 



Our subjective preference is for simplicity. Our at least partially 

 objective disco\'ery is that simplicity is attainable in some ways and 

 not in others. Weyl says: 



It often happens that for some partial domain an explanation A is 

 simpler than B; but while A becomes increasingly complicated as the 

 circle of experience widens, the same does not apply to B, with the 

 result that eventually B emerges as the superior theory. Further- 

 more the required simplicity is not necessarily the obvious one, but 

 we must let nature train us to recognize the true inner simplicity. 



Nature's training is too energetic to be ignored. Thus, after noting 

 how quantum mechanics grows out of a supremely confident en- 



