ON THE THEORY OF SCIENCE 351 



All history has primarily the task of fixing past events through the 

 effects which have remained from them. Where such are not access- 

 ible, only analogy is left, a very doubtful means for gaining a concep- 

 tion of those events. But it must be kept in mind that an event which 

 has left no evident traces has no sort of interest for us, for our interest 

 is directly proportional to the amount of change which that event has 

 caused in what we have before us. The task of historical science is 

 just as little exhausted, however, with the fixing of former events 

 as, for instance, the task of physics with the establishment of a single 

 fact, as the temperature of a given place at a given time. Rather the 

 individual facts must serve to bring out the general characteristics of 

 the collective mind, and the much-discussed historical laws are laws 

 of collective psychology. Just as physical and chemical laws are 

 deduced in order with their help to predict the course of future phys- 

 ical events (to be called forth either experimentally or technically), so 

 should the historical laws contribute to the formation and control of 

 social and political development. We see that the great statesmen of 

 all time have eagerly studied history for this purpose, and from that 

 we derive the assurance that there are historical laws in spite of the 

 objections of numerous scholars. 



After this brief survey, if we look back over the road we have come, 

 we observe the following general facts. In every case the development 

 of a science consists in the formation of concepts by certain abstrac- 

 tions from experience, and setting of these concepts in relation with 

 each other so that a systematical control of certain sides of our 

 experience is made possible. These relations, according to their gener- 

 ality and reliability, are called rules or laws. A law is the more 

 important the more it definitely expresses concerning the greatest 

 possible number of things, and the more accurately, therefore, it en- 

 ables us to predict the future. Every law rests upon an incomplete in- 

 duction, and is therefore subject to modification by experience. From 

 this there results a double process in the development of science. 



First, the actual conditions are investigated to find out whether, be- 

 sides those already known, new rules or laws, that is, constant relations 

 between individual peculiarities, cannot be discovered between them. 

 This is the inductive process, and the induction is always an incom- 

 plete one on account of the limitlessness of all possible experience. 



Immediately the relationship found inductively is applied to cases 

 which have not yet been investigated. Especially such cases are 

 investigated as result from a combination of several inductive laws. 

 If these are perfectly certain, and the combination is also properly 

 made, the result has claim to unconditional validity. This is the 

 limit which all sciences are striving to reach. It has almost been 

 reached in the simpler sciences: in mathematics and in certain parts 

 of mechanics. This is called the deductive process. 



