342 METHODOLOGY OF SCIENCE 



will remain so. It must not be an arbitrary combination; such a 

 thing would have only momentary duration, and the task of science, 

 to learn the unknown from the given, could not find application. 

 Rather must this element have such a nature that it can be charac- 

 terized and recognized again, that is, it must already have a concept- 

 ual nature. Therefore only parts of our experience which can be 

 repeated (which alone can be objects of science) can be characterized 

 as things or objects. But in saying this we have said all that was 

 demanded of them. In other respects they may be just as different 

 as is conceivable. 



If the question is asked, What can be said scientifically about 

 indefinite things of this sort? it is especially the relations of arrange- 

 ment and association which yield an answer. If we call any definite 

 combination of such things a group, we can arrange such a group 

 in different ways, that is, we can determine for each thing the relation 

 in which it is to stand to the neighboring thing. From every such 

 arrangement result not only the relationships indicated, but a great 

 number of new ones, and it appears that when the first relationships 

 are given the others always follow in like manner. This, however, 

 is the type of the scientific proposition or natural law (page 335). 

 From the presence of certain relations of arrangement we can deduce 

 the presence of others which we have not yet demonstrated. 



To illustrate this fact by an example, let us think of the things 

 arranged in a simple row, while we choose one thing as a first member 

 and associate another with it as following it ; with the latter another 

 is associated, etc. Thereby the position of each thing in the row is 

 determined only in relation to the immediately preceding thing. 

 Nevertheless, the position of every member in the whole row, and 

 therefore its relation to every other member, is determined by this. 

 This is seen in a number of special law r s. If we differentiate former 

 and latter members we can formulate the proposition, among others, 

 if B is a later member with reference to A, and C with reference to 

 B, then C is also a later member with reference to A. 



The correctness and validity of this proposition seems to us beyond 

 all doubt. But this is only a result of the fact that we are able to 

 demonstrate it very easily in countless single cases, and have so 

 demonstrated it. We know only cases which correspond to the 

 proposition, and have never experienced a contradictory case. To call 

 such a proposition, however, a necessity of thinking, does not appear 

 to me correct. For the expression necessity of thinking can only rest 

 upon the fact that every time the proposition is thought, that is, every 

 time one remembers its demonstration, its confirmation always arises. 

 But every sort of false proposition is also thinkable. An undeniable 

 proof of this is the fact that so much which is false is actually thought. 

 But to base the proof for the correctness of a proposition upon the 



