676 SOCIAL SCIENCE 



Nature is a pedant. Nature does not alter her habits. Nature al- 

 ways repeats herself. Not only is it true that, the same conditions 

 being given, the same effects will always follow; not only is there 

 this constancy of relation, but it is likewise true that the same con- 

 ditions will repeat themselves, or conditions sufficiently similar to ex- 

 hibit the constancy of relation. Indeed, when we speak of the phys- 

 ical order, we are speaking of an abstraction, of one side of things, 

 to which our attention is restricted for the time, namely, of that 

 side of things which is characterized by just this constancy both of 

 relation and of recurrence. But in the case of the social uniformities, 

 while there is an approximate stability of relations between sets 

 of phenomena, there is no such constancy in the repetition of the 

 phenomena. Physical nature is a pedant, human nature is a 

 Proteus. In the case of physical nature the object studied may be 

 compared, let us say, to a building, which for some reason we are 

 not permitted to approach closely, the various features of which are 

 foreshortened now in one way, now in another, according to our point 

 of view. The object itself does not change its shape. Could we come 

 near enough, we should be able to report it with perfect exactitude. 

 But in the case of society, the object is not only remote from our 

 apprehension, despite its apparent nearness, but in addition it changes 

 its shape while we are engaged in the act of contemplating it. It is 

 in this way that I should reply to the argument of those who hold that 

 the law of gravitation also is but an incomplete transcript of Nature's 

 process, and that the inexactitude which is put forward as an objec- 

 tion to the use of the word " law " for social uniformity would equally 

 apply to natural law. In the one case the inexactitude is in the appre- 

 hension of that which is fixed and stable, of that which, by reason 

 of its stability, permits of a considerable approximation to exacti- 

 tude of description. In the other case the inexactitude is due both 

 to the complexity of the object studied and to the unpredictable 

 changes which take place in it. These changes, let me now again 

 remind you, arise from the fact that uniformities of conduct are 

 adaptations of means to ends, and that these ends are ideas, and, 

 therefore, that the uniformity lasts only so long as the idea lasts or is 

 dominant. But one dominant idea may be displaced by a different 

 one, as happened at the time when the Germanic peoples were con- 

 verted to Roman Christianity. And when a new idea becomes domi- 

 nant, the order of the motives that govern conduct is revolutionized, 

 the old ideas, to the extent that they still remain operative, entering 

 into new combinations, and the whole complexion of conduct, in con- 

 sequence, being altered. If, indeed, it could be shown, as has been 

 attempted, that there is not only uniformity as between ideas and 

 the mode of realizing them in a given society, but also a uniform pro- 



