Modification by Experience 237 



ments involved in a definite order. They showed an inter- 

 esting tendency to skip at once to the movement that im- 

 mediately preceded the opening of the door (82). 



The question arises, as in the case of the labyrinth ex- 

 periments, whether, when the animal has learned the proper 

 movements to open a box, he opens it by "remembering" 

 the movement ; that is, by having some kind of an idea or 

 image of it in his consciousness, or whether we have to do 

 with the formation of a habit by a process in which ideas 

 are at no time involved. Here, again, the gradual character 

 of the learning process, where it is gradual, points to the 

 absence of ideas; a human being who had once hit by 

 accident upon the right way to open a lock could hardly fail 

 on being confronted with it a second time, at not too great 

 an interval, to recall an idea of the successful movement 

 and perform it at once, without any unnecessary accom- 

 panying movements. We have seen an approach to this 

 state of things in the monkeys; accordingly it is possible 

 that they may learn by means of ideas. On the other hand, 

 rapid learning, where the action is very simple and closely 

 connected with the animal's instincts, does not necessarily 

 mean the presence of ideas; in certain cases there may 

 exist arrangements for the rapid modification of an instinctive 

 mechanism which do not involve the production of images 

 at all. The most we can say is that slow learning, by gradual 

 elimination of the useless movements, indicates, so far as 

 we can judge, the absence of any guiding idea of the action. 

 Other evidence against the idea hypothesis was derived by 

 Thorndike from various facts. In the first place he found 

 in the animals observed by him an entire lack of what has 

 been termed inferential imitation. 



Imitation in animals has by some writers, notably Was- 

 mann, been classed as a special method of learning by 



