178 DOGMATISM AND EVOLUTION 



that of ideas, is an illusion to be explained psychologically; but 

 such explanation in no wise constitutes a justification of such an 

 interpretation. 



Some of the causes to which Mill attributes the growth of this 

 illusion, we have already mentioned. That is to say, experience, 

 through the operation of the laws of association, once having 

 given rise to the idea of groups of possibilities of sensation, these 

 groups come to be thought of as permanent, and as persisting 

 relatively unchanged while our actual experiences are constantly 

 changing. They are naturally regarded, therefore, as, in a sense, 

 independent of ourselves. In the second place, the groups of 

 possible sensations are not realized by ourselves alone; but they 

 are objects of common experience in a way in which the sensations 

 themselves are not. Under the same conditions, others have 

 experiences similar to ours. It is not so much that the sensations 

 compared one to one are precisely similar, but that they exhibit 

 the same uniformities of antecedence and sequence. It is then 

 the groups of possibilities the conditional certainties that are 

 constantly verified by our intercourse with others. Moreover, 

 as we must not forget to recall, the object is given a place in its 

 system independent of our subjective sensations and feelings, not 

 simply because it is accessible to other men, but because to it 

 the universal laws of nature apply. Finally, what serves to 

 complete the emancipation of the object is the inevitable tendency 

 to regard it as bearing a causal relation to our sensations. For 

 the actual sensations we feel are indubitable evidence to our 

 minds of the presence of some group of possibilities of sensation; 

 and these possibilities are held to have been equally possibilities, 

 whether the conditions of their realization in actual experience 

 were fulfilled or not that is, the object is regarded as necessarily 

 existing prior to our perception of it, which is precisely the con 

 dition for the ascription of the causal relation. 1 There is but one 



1 Mill s treatment involves likewise an important modification of Hume s theory 

 of causality. For just as the uniformities of nature hold, not of the antecedence 

 and sequence of sensations, but of the antecedence and sequence of possibilities of 

 sensation, so the causal connection applies, not to the sensations in their relation 

 to each other, but to objects. 



