THE HUMAN SCIENCES 39 



And the result which ensues is that on examining itself it 

 discovers that it is employing, whichever of the contra- 

 dicting truths it adopts, an unjustified and even an 

 unsuspected hypothesis ; and on examining the facts that 

 their opposing aspects are, both alike, phases of a deeper 

 unity. And these hypotheses prove, I believe invariably, 

 to be metaphors. The object which is investigated has 

 been observed through the medium of a familiar concep- 

 tion which was serviceable in the interpretation of other 

 facts, but is not applicable to it. For metaphors cannot 

 give the truth, any more than analogies can prove. Meta- 

 phors are not metaphors except when they omit something 

 relevant and introduce something irrelevant. But facts 

 have their individual rights, and will not submit to vicarious 

 treatment. And no science can ever set forth on a pro- 

 gressive path till its regulative hypothesis and method of 

 inquiry are determined for it by the facts it investigates, 

 and not merely by some other facts with which the 

 investigator happens to be better acquainted. 



Now, modern theories of society have been passing 

 through precisely this process. They began by endeavour- 

 ing to explain man, and his social manifestation of his 

 rational nature, by means of the categories which had 

 proved effective in the explanation of the natural world. 

 // was taken for granted that subjects are related to each 

 other as objects are. The interaction of their wills in 

 society was an instance of stress and strain, of attraction 

 and repulsion and external collision. Society was a 

 mechanism ; its elements were compacted tog-ether 



1 O 



coercively. 



But this view led to an endless controversy in both 

 social theory and social practice. It was impossible if I 



