DEDUCTION IN SCIENCE 9 



laboratory experiment has come in conflict with 

 reahty and may have to be abandoned. 



The method of deduction takes us away from 

 reahty, and Darwin understood well the danger 

 in its use. In a letter to A. R. Wallace (August 

 28th, 1872), he wrote : " I am not convinced, 

 partly, I think, owing to the deductive cast of much 

 of his reasoning ; and I do not know why, but I 

 never feel convinced by deduction, even in the case 

 of H. Spencer's writings." In subjects that can be 

 treated experimentally, deduction is an invaluable 

 discipline. The experimenter is able to say " that 

 being so, this other must follow," and if, on trial, 

 this other does in fact follow, not only does the 

 major proposition receive confirmation, but a direct 

 addition to knowledge has been made. A pitfall 

 gapes widely when the deduction cannot be subjected 

 to experiment, for we are too readily disposed to say 

 " that being so, this other must foHow, and as it must 

 follow, it is true." The moment we try to divert a 

 proposition from the exact set of facts out of which 

 it was derived, and to apply it to any other set of 

 facts whatsoever, we are in imminent danger of 

 foisting on ourselves and on others an analogy as a 

 truth, and the glamour of apparent certainty that 

 comes from the word science makes our fall only 

 more disastrous. Of all the philosophers, Immanuel 

 Kant (by descent a Scot) put the antinomy between 

 science and reality in the sharpest fashion. All 

 science, all experience, all scientific laws, he referred 

 to what he called the theoretical reason, and 

 regarded as the reaction of the human mind to the 

 external world. All morality, all that relates to the 



