274 THE EVOLUTION OF SCIENTIFIC THEORIES 



should scarcely be possible to think on chemical subjects in a way dif- 

 ferent from their theories. 



The heuristically potent theory shows itself viable: it proliferates 

 itself in the minds of scientists. After a time its heuristic power may 

 wane, but by then its position is secured. It is enshrined in textbooks, 

 linked in scientific tradition with the names of famous men, woven 

 into the fabric of scientific language and thought. Most important of 

 all, its position is secured because the theory with which it had once 

 to compete has been displaced, crowded out. The once-novel theory 

 is now orthodoxy, maintained (if by nothing else) by sheer intellec- 

 tual inertia. Natural selection is thus decisive where "persuasion" 

 fails and, within the space of a single generation, a viable scientific 

 theory can so attain practically universal accreditation. 



Viability and validity. Natural selection in nature need not select 

 "the good," and natural selection in science need not favor "the true." 

 A slightly deeper exploration of the metaphor will be helpful. In the 

 world of nature, any "part" of an organism reacts with its other parts 

 through an internal environment. Any mutation producing a "part" 

 fundamentally incompatible with the others is then a lethal muta- 

 tion: those who carry it will not survive. Any surviving mutants then 

 face a further screening, in the external environment with which the 

 organism interacts as a whole, and natural selection will favor just 

 those mutations that give the organism some advantage in its habitat. 

 Thus biological natural selection always functions in terms of two 

 distinct sets of environmental interactions, and all such selection is 

 made relative to two kinds of prevailing conditions. Clearly then the 

 "judgments" of natural selection may well vary with time and situa- 

 tion: once notably successful, the dinosaur is now extinct. 



In the world of scientific ideas natural selection also proceeds by 

 way of two sets of interactions— with an inner (conceptual) and an 

 outer (empirical) environment. A new theoretic idea is stillborn, at 

 once rejected even by him who first conceives it, unless it has some 

 minimum degree of compatibility with the intellectual atmosphere 

 of the age. Some still larger degree of compatibility will be required 

 if the new idea is to function efiPectively. Aristarchus' conception of a 

 heliocentric system was nonviable simply because it was so deeply 

 incompatible with much the ancient world took for granted. The 

 very early (and quite imperfect) anticipations of an oxygen theory, 

 by Mayow and others, similarly failed to "catch on" in a scientific 



