THE LOGICAL STRUCTURE OF SCIENCE 45 



of the legitimacy of isolating systems for purposes of study. 1 

 In every experimental situation certain forces, usually of 

 minor intensity, and having their loci in remote points, are 

 assumed to be negligible. For example, in describing the 

 rolling of a ball down an inclined plane, the attractions of 

 the fixed stars may be neglected. This does not involve a 

 denial of the existence of such forces, but simply a conven- 

 ient neglect of them for purposes of simplification. It is 

 similar with reference to the background of thought. Though 

 thinking arises out of problem situations and exists only as 

 an instrument of adjustment, one may conveniently neglect 

 this fact in his analysis of the scientific method. He may 

 consider thinking as a system which is temporarily isolated. 

 This simplifies the problem, and necessitates only that in 

 the final description the possible disturbing influences be 

 reintroduced through some method of successive approx- 

 imation. 



One should also recognize that in speaking of the logic of 

 science as an empirical study he can hardly expect its data 

 to have the same clarity as those, say, of physics. The 

 physicist is confronted with readings of pointers on scales, 

 deflections of needles on galvanometers, clocks, and meter 

 sticks, and other sense-data. These constitute "hard" data, 

 about whose existence there can be little doubt, and as to 

 whose character there can be little disagreement. But the 

 data for the logician of science are elusive and vague. They 

 consist of the "ideas" and "mental processes" of the scien- 

 tist as these are described by the psychologist either in 

 introspective or in behavioristic terms. They consist of 

 symbols — pictures and diagrams drawn by the scientist, 

 words and mathematical formulas uttered and written by 

 him — but these symbols are considered not as mere physical 

 existents such as configurations of chalk, or ink, or series of 

 noises; they must be examined in their "meaning" aspect, 

 and this is their characteristically elusive feature. Further 



1 This point has been well emphasized by H. Levy, The Universe of Science, 

 (New York: Century, 1933), especially Chapter II. More detailed examination of 

 the function of isolation in science will be made in Chapter VI. 



