J.— PSYCHOLOGY. 191 



which we can say that the taste reactions thus experimentally studied 

 are 'simpler' than those of normal everyday life. Now take the analogous 

 case of remembering. A part of the total process is obviously the initial 

 learning. If we use ordinary combinations of meaningful words or forms 

 there may be a mass of associations, in some respects peculiar to each 

 subject who submits to experiment, set up during this initial period of 

 observation. By the use of nonsense syllables we cut off all, or most, 

 of these thronging associations, just as in the other cases we cut off the 

 olfactory sensations, and so we find out what pure and simple remembering 

 can do. 



The argument is feeble. We do not isolate the taste experiences by 

 simplifying the stimulus, but by cutting off, through extirpation or tem- 

 porary paralysis, those other reactions with which they are normally 

 integrated. We can operate on the olfactory nerves in this way if we desire, 

 but there are no specific memory nerve endings or nerve centres, or if 

 there are we do not know them. Moreover, even if we could do this, the 

 experimental psychologist would be untrue to his pretensions if he were 

 satisfied with this alone. He pretends to deal with the intact organism. 

 Suppose we do simplify enormously our stimuli, or our experimental 

 situations and then confront our intact subject with these. All that 

 happens is that he is faced by such an odd and unusual state of affairs that 

 he is forced to mobilise all his resources and make up some novel reaction 

 ad hoc. This is, of course, exactly what happened when the nonsense 

 material was used. The early experimenters, being in other ways exceed- 

 ingly good psychologists, saw this at once. Their way out was curious. 

 They proposed that every subject must have long-continued practice with 

 nonsense syllables before any of his results should be allowed to count. 

 That is to say, having taken immense precautions to simplify experimental 

 material and methods, we must then take equally great precautions to 

 simplify, by rendering familiar, the highly artificial and complex response 

 which we have thereby set up. The way in which this sort of method 

 has run rampant over the whole of the laboratory psychology of the 

 higher mental processes is distressing to anybody who wishes to adopt an 

 experimental approach to this field of research. It is, perhaps, hardly an 

 exaggeration to say that nine-tenths of conventional laboratory psychology 

 - — outside of experiments on the special senses — consists largely in showing 

 how habits of response may be set up towards odd and out-of-the-way 

 situations, and makes little contribution of moment towards the solution 

 of any other problem. 



This is the old difficulty in a new form, the exaggerated respect for the 

 stimulus or the situation. The psychologist is studjdng the complex 

 responses of a highly developed organism and how they are determined. 

 They have been called forth to meet the claims of a very unstable and vary- 

 ing objective environment. They are no doubt in many ways more stable 

 than that environment. But if the environment is violently simplified it 

 it is mere superstition to trust that they also get simplified in a correspond- 

 ing manner. They become different, but are just as likely to become yet 

 more highly complex. Stability of determination, not simplicity of 

 structure in objective determining factors, is what we need to make our 

 experiments convincing. Stability of determination is compatible with 



