vii ASSIMILATION AND READJUSTMENT in 



itself seems to be the pivot of all inference. It is through 

 this relation that -an object of experience points the mind 

 to something beyond itself. Hence any action resting 

 implicitly or explicitly on an experienced relation between 

 datum and consequence may be said to be of the inferential 

 type. Such a " relation " may be tc experienced," as we 

 shall see, in very different forms and in very different degrees 

 of explicitness, but we ought not, as I think, to exclude 

 from the type any action based on an experience into 

 which, however inarticulately, such a relation enters as an 

 essential element. I should say, then, that when we 

 experience a certain result as flowing from a certain line of 

 conduct, and modify our behaviour accordingly, we are 

 drawing a simple inference. Action of this kind is of the 

 inferential type, and is in line with the general development 

 of intelligence. When, on the other hand, we bear heat 

 or pain better merely because we have borne a good deal 

 already, it is not our conscious intelligence that is directing 

 our conduct in accordance with experienced results, but the 

 heat or pain which have had a physical effect on us that 

 has left a permanent trace. When I speak of experience, 

 then, I mean experience of the kind on which inference is 

 founded, and that is, to speak generally, experience of 

 data in relation. 



3. Primary Retentiveness. 



When an inference is based on a past experience it is 

 clear that the effects of that experience must in some way 

 or other endure, or, in the ordinary phrase, be retained by 

 the organism. Needless to say, it is not the experience 

 itself which is retained. That is a temporary state, an 

 event which happens once and passes. What is retained 

 is some result manifesting itself in the subsequent life of 

 the organism. Nor is this result necessarily a memory of 

 the past experience. Memory is a form of retentiveness, 

 but not all retentiveness is memory. The result may 

 show itself simply in some modification of future be- 

 haviour, and among such modifications are inferences, or 

 modified responses which look like inferences, based on 

 results that have at some previous time been experienced. 

 But there are forms of retentiveness which neither involve 



