THE ANATOMY OF SCIENCE 81 



(appropriate, he hopes) concepts— e.g., velocity and acceleration— 

 which invest certain selected (significant, he hopes) elements with 

 relatedness. A considerable abstraction is already expressed in the 

 selection, and a much greater one is now to be made. 



(2) With bundles of data bound by his concepts, he withdraws 

 from the complex world of experience toward an "ideal" conceptual 

 realm. Refusing to accept the limits common sense sets itself, the 

 scientist deals quite elaborately with refined abstractions akin to those 

 of the philosopher. He seeks to penetrate behind the appearances: he 

 hypothesizes existent in nature certain simple entities and relations 

 {e.g., "free fall"). In terms of these he asks certain new questions, 

 perhaps comes to grips with the core of his problem. Always he seeks 

 for ever greater abstraction, thinking thereby to penetrate ever more 

 deeply— to the essence of an ever wider body of phenomena. 



(3) How can he gain confidence that his abstract conceptions do 

 hold any such essence? Unlike the ancients, he rejects the expedient 

 appeal to self-evidence. He tries instead to break back out to ob- 

 servables: drawing deductions from his abstract postulates, he ven- 

 tures to predict what, under certain circumstances, should be found 

 in actual experience. Thus always in the end the scientist returns 

 from the world of abstractions, however remote, to the world of ex- 

 perience the man of common sense never fully leaves— and the philos- 

 opher never fully regains. 



(4) The scientist need not return all the way to the raw experience 

 of observation, but only as far as experience provided by laboratory 

 systems purposefully designed to highlight the eflFects of the factors 

 to which the abstract concepts direct attention. With Galileo we 

 come then at last to the classic experiments on the inclined plane. 

 The theoretic predictions being borne out, the theoretic construction 

 may be sound. But were inexplicable predictive failures to be found, 

 these— not to be brushed aside as marks of the nonideality of an 

 intrinsically imperfect world— are to be recognized as calling the ab- 

 stract constructions in question. 



Considering our experience, we create in our thoughts a world of 

 ideal abstractions. Whatever its beauty, this world is nonetheless an 

 artifice of thought. Ultimately the business of science is to talk not 

 of this world but of the "real" world, with all its imperfections and 

 complexities. Experience thus becomes the touchstone by which we 

 prove (in the sense of test) our abstract conceptions. 



