192 SECTIONAL ADDRESSES. 



complexity, and even with considerable variation of objective determinants. 

 Somehow an experimental method has to be developed which recognises 

 this fact. 



I turn for a moment to experimental work on ' recognition ' merely to 

 bring out one further point of method. Since very early days an enormous 

 amount of work has been done on this topic. It has issued in five or six 

 different theories no one of which can claim finality. The diversity of this 

 result is due to different causes, but one is perhaps particularly important. 

 There is a strong tendency, when any complex response like recognition 

 is being studied, to attempt to draw a ring round it and to seek its explana- 

 tion within these imposed limits. Thus the explanation of recognition is 

 sought in something that happens at the moment of recognition. This is 

 surely wrong. An object or event may be recognised or not largely on a 

 basis' of how it was reacted to in the prior perception. A sound, for 

 example, may be heard : it will not be recognised unless it is so listened 

 to that it possesses qualities, characteristics, a setting and a significance. 

 To the persistent study of a complex mental response as if its psychological 

 explanation must be found inside an imaginary circle that encloses it, 

 much of the disrepute into which experimental psychology tends to fall 

 may be traced. It is perhaps the last and subtlest form of the outworn 

 ' faculty ' psychology. 



I said at the beginning that the early experimentalists in psychology 

 were physicists and physiologists with a strong bent towards philosophy. 

 If in the history of the subject experimental psychologists have shown them- 

 selves too submissive to physical and physiological methods, it is even 

 more true that they have often pursued philosophical ideals. This 

 pursuit is in full cry still. A very brief consideration of current movements 

 which originate in the laboratory will illustrate this point. 



There is probably no contemporary movement in psychology which Has 

 more profoundly influenced psychological thought in English-speaking 

 countries than the so-called Gestalt psychologie. I have, as every experi- 

 mental psychologist must have, a very great admiration for the brilliant 

 work of Wertheimer, Kohler and Koffka. It has shed much new light on 

 old problems, as well as a good deal of old light on new problems. It 

 starts specifically from experiments upon visual perception and its primary 

 method is that of phenomenological description. When we are presented 

 with a perceptual situation what is it that we experience ? The answer is 

 inevitable and is one which, from this point of view, has, I think, always 

 been given. We cannot describe our experience in this sort of situation 

 as a mosaic of tiny bits each corresponding with its isolable part of the 

 stimulus or situation. The blue sky which is seen is not, as Kohler says, 

 made up of an infinite number of blue sensation units, but is seen as a 

 continuous blue expanse. The moving dots and lines, in Wertheimer's 

 experiments, are seen, not as stationary points in temporal relations, but 

 as a unitary and indivisible movement. Sometimes the experiments are 

 more behaviouristic, but still it is the attitude of phenomenalistic de- 

 scription that determines their interpretation. The animal that has been 

 trained to react positively to a and negatively to b at once reacts positively 

 to 6 if a is removed and c is introduced bearing that relation to b which b 

 had to a. This must be because the initial reaction was not to a, or to b, 



