RELATION OF ETHICS TO SOCIAL SCIENCE 675 



ity of connotations will tend to confusion. If, on the other hand, 

 we assimilate the social uniformities to physical laws, then we shall 

 create the impression of the inevitable dominance of these uniform- 

 ities over the human will, an impression which is in the highest degree 

 prejudicial to social betterment. That such an impression has 

 already been created by the use of the term " law " in the field of 

 economics is matter of common knowledge. Shall we extend the 

 mischief to other departments of conduct, or can we hope to avert 

 it by surrounding the term " law " as applied in social science, with 

 qualifications which are sure to be ignored? 



The capital question is this: Whether the social uniformities are 

 of the same kind as those uniformities which we designate as physical 

 laws? If they are, let the same term be applied to both. If they 

 are not, let the difference in the designations used indicate and 

 emphasize the difference in the things designated. Let us in that 

 case speak of social uniformities and not of social laws. 



Is there such a difference in kind, and what is it? I can here only 

 shortly outline my position, without attempting to elaborate it in 

 detail or to defend it against adverse criticisms that may be brought 

 to bear against it. Physical law is the expression of a fixed relation 

 between antecedent phenomena as cause and sequent phenomena as 

 effect. A social uniformity, on the other hand, is the expression of 

 a relation between ends and means. All human conduct is directed 

 toward the attainment of ends. All the uniformities of human 

 conduct depend on the average fitness of certain means to make for 

 the achievement of certain ends. The point of view from which 

 Nature operates, if a metaphor be allowed, is causal. The point of 

 view from which man acts is teleological. Nature is governed by 

 forces. Man is determined by ideas. The difference is vital. 



Again, physical law implies not only a fixed relation between phe- 

 nomena, but also involves that these phenomena, together with the 

 fixed relation between them, are of constant occurrence. A physical 

 law which should be true only under transient conditions, that is, 

 under conditions which fail to repeat themselves, is inconceivable. 

 Physical truth is true semper et ubique. Undoubtedly this account 

 holds good only of ideal physical law and does not perfectly apply 

 to any of the physical laws which are formulated in the text-books. 

 As Professor Sidgwick has it: " We assume the existence of natural 

 laws, and the narrowing down of these into exactitude is the endless 

 problem of discovery." An endless problem, indeed, a goal to 

 which we only approximate asymptotically. A remnant of inex- 

 actitude remains unexpunged even in the most certain of natural 

 laws. But the cause of the inexactitude is not any wavering or 

 flickering in Nature's process, but is due to our subjective uncertainty 

 whether we adequately apprehend the process as it actually goes on. 



