vi THE EMPIRICAL ORDER 99 



can express their meaning only by speaking of the dead as 

 continuing to live, so that the practice emerges from a senti- 

 ment, and in turn gives rise to the belief that would justify 

 it.' If this explanation is correct neither magic nor animism 

 is primitive. What is primitive here, as all through the 

 earlier stages of psychology, is impulse-feeling, and here as 

 in those earlier stages the idea formulates, directs, extends 

 and in a way explains the act to which feeling prompts. 

 This is at least one root of primitive belief. On the other 

 hand the extensions of idea involved in magic and animism 

 the tendency to clothe one object with the attributes of 

 another, not through conscious generalisation but through 

 failure in distinction seem equally * natural ' to the dawn- 

 ing intelligence. Idea may suggest impulse, or impulse 

 may lead up to idea. Provided there is fundamental har- 

 mony with feeling, the ideas will be suited to their environ- 

 ment, and will survive. In general we may suppose that 

 the magic ceremony, the animist's spell-prayer, the witches' 

 charm all have an efficacy of their own bringing relief to 

 the tension of suffering or anxiety, or arousing terror and 

 dismay in the mind of enemies at whom they are aimed. But 

 this emotional suitability considered as evidence for truth 

 stands materially below the rough logic of common sense. 



We may then formulate the advance made in passing 

 through the first two stages of human thought much as 

 follows. With the origin of language there arises the 

 germ of the power to group experiential data in accord- 

 ance with their affinities, and so to build up conceptions of 

 individuals, groups and classes as the subjects of rough and 

 ready generalisations. With regard to matters standing 

 out very plainly in experience or very close to practical 

 interests there is not room for much divergence in method. 

 People are not to be persuaded that thirst can be satisfied 

 without drinking, or that a flint stone is soft to lie on. 

 But outside the limited area of readily tested belief lies a 

 mass of more doubtful ideas of great significance in human 

 life. In this region we find in the first stage that the move- 

 ments of fancy under the sway of feeling take the lead in 

 forming belief, and that the ideas formed are so obscure and 

 inconsistent as to blur the deepest lines of distinction drawn 



