Jtjnb 17, 1898.] 



SGIENGE. 



821 



througli them by the experimenter. They 

 were thus unable to infer that since another 

 by pulling a string obtained fish, they 

 might, or that since fish were gained when 

 I pushed round a bar with their paws it 

 would be gained if they pushed it round 

 themselves. 



Experiments were made on imitation by 

 giving the animals a chance to see one of 

 their fellows escape by clawing down a 

 string stretched across the box, and then 

 putting them in the same box alone. It 

 was found that, no matter how many times 

 they saw the act done, they could not 

 thereby learn anything which their. own 

 impulsive activity had failed to teach them, 

 and did not learn any more quickly what 

 they would have sooner or later learned by 

 themselves. One important consequence of 

 these results is the resulting differentiation 

 of the Primates from the other orders of 

 mammals. If the Primates do imitate and 

 the rest do not, we have located a definite 

 step in the evolution of mind and given a 

 new meaning to the line of human ancestry. 

 I do not, however, hold that these results 

 eliminate the possibility of an incipient 

 faculty of imitation among mammals in 

 general. They do deny the advisability of 

 presupposing it without proof, and emphat- 

 ically deny its presence in anything equiva- 

 lent to the human form. Finally many 

 actions which seem due to imitation may 

 be modifications of some single instinct, 

 such as that of following. 



Perhaps the most valuable of the experi- 

 ments were those which differentiate the 

 process of association in animals from the 

 ordinary ' association by contiguity ' of hu- 

 man psychology. A man, if in a room from 

 which he wishes to get out, may think of 

 being outside, think of how he once opened 

 the door, and accordingly go turn the knob 

 and pull the door open. The thought of 

 opening the door is sufficient to arouse the 

 act of opening the door, and in most human 



association- series the thoughts are the es- 

 sential and sufficient factors. It has been 

 supposed that the same held true of ani- 

 mals, that if the thought of doing a thing 

 were present an impulse to do it would be 

 readily supplied from a general stock. Such 

 is not the case. None of these animals could 

 form an association leading to an act unless there 

 was included in the association an impulse of its 

 own which led to the act. Thus cats who had 

 been induced to crawl into a box as the 

 first element in a pleasurable association- 

 series soon acquired the habit of crawling 

 in of their own accord, while cats who had 

 been dropped in did not. In the second 

 case the idea of being in would be present as 

 strongly as in the first, but the particular 

 impulse to go in was not. So also cats who 

 failed of themselves to learn certain acts 

 could not be taught to do them by being 

 put through them, while cats who were 

 thus put through acts which accident would 

 of itself alone have taught them, learned 

 them no more quickly and often made the 

 movement in a way quite different from 

 that which they were shown. Their as- 

 sociations are not primarily associations of 

 ideas with ideas, but associations of sense- 

 impressions and ideas \oith impulses to acts, 

 muscular inner'vations. The impulse, the in- 

 nervation, is the essential. 



This does not mean that the animals can 

 have no representations or images at all. 

 Another set of experiments show that they 

 probably can. It means that they have no 

 stock of free-floating impulses which can 

 be called on at will ; that the elements of 

 their associations occur chiefly just in their 

 particular connections; that their ideational 

 life consists not of a multitude of separate 

 ideas, but of a number of specific connec- 

 tions between ideas and impulses. 



Having thus denied that animal associa- 

 tion is homologous with human association, 

 as the latter is ordinarily conceived, we 

 find the true homologue of animal associa- ^ 



