i 5 6 



MIND IN EVOLUTION CHAP. 



for twenty-four hours, while as many as twelve readings 

 fixed less than four syllables out of ten in Mr. G. W. 

 Smith's experiments. 1 



1 The effective presentation of the problem of the rapidity of learning 

 has been the subject of several experimental studies, but the results are 

 not as yet as clear as might be desired. From what we know of the 

 definiteness of memory in the case of some particular event, we naturally 

 infer that a single instance should be decisive where there is a true 

 memory judgment. Conversely, from our common experience of the 

 gradualness in the formation of habit and the acquisition of skill, we 

 naturally take slowness of acquisition as evidence of the absence of the 

 true memory judgment. How far do these views stand the test of experi- 

 ment? With regard to skill, Mr. E. J. Swift, in a valuable article in the 

 American Journal of Psychology, 1903, on " Studies in the Psychology 

 and Physiology of Learning," shows that in the practice of passing and 

 catching balls, progress is at first slow, and then more rapid. Where 

 improvement is rapid in early stages it is in relation to things which have 

 symbols or other devices for handling and presenting ideas. This is 

 precisely the contrast which we should expect, and the convex curves 

 which Mr. Swift presents have exactly the anticipated contrast to the con- 

 cave curves, with steep descents at the beginning and very gradual sub- 

 sequent change, which belong to the memorising of something in which 

 we understand the connections. On the other hand, Mr. Hicks and 

 Professor Carr (" Human Reactions in a Maze," J. A. B. 1912, pp. 98-125) 

 have shown that the form of time curves, /.*., the rate of the diminution of 

 the time in which an animal performs a trick, is not to be taken as a 

 criterion of the intelligence applied, under penalty of making rats more 

 intelligent than human beings. Further, Shepard and Brede (J. A. B., 

 1913, pp. 274-285) have shown that the development of the pecking instinct 

 in chicks improves with great rapidity in the early stages, and then tails off 

 into a slow advance towards perfection. This is the opposite result to 

 those of Swift for the acquisition of skill. Again, Schaeffer, as noted 

 above, found inhibitions learnt by a frog in one or two instances alone, 

 and it would be very hard to attribute to a frog the higher order of 

 intelligence which we are now considering. Some other cases of inhibi- 

 tion by very few experiences, if not by one alone, have been recorded 

 among quite low animals. 



I think we may find the key to these difficulties if we contrast the 

 rapidity with which the frog in Schaeffer's experiments learnt to reject 

 nauseous food with the slowness of its maze reactions as described 

 by Yerkes. We may suppose that the direct confirmation or inhibi- 

 tion of an instinctive response operates rapidly, while the building up 

 of habits that have no basis in instinct requires a large number of 

 experiments, accident probably playing a considerable part in the 

 process. The maturation of the pecking instinct would then follow the 

 law for the direct modification of response. It would advance rapidly 

 with the success of one type of movement and the failure of another, 

 and it is only the subtler adjustments which would require time. On 

 the other hand, the acquisition of something new, such as an acquired 

 art in the case of man, would be rapid only in those portions which 

 demand intelligence and definite memory, and would be slow on the 

 side of the sensori-motor adjustments. In the case of an animal learn- 

 ing a trick the rapid initial fall so frequently seen would probably be due 

 to the more intelligent elements in the process, and the slow later fall to 



