OSHEA — ASPECTS OF MENTAL ECONOMY. 39 



It may occur to some that this subject is as a whole an ex- 

 tremely large and complicated one to be treated according to 

 scientific methods. They may feel conduct consists of such a 

 bewildering complexity of factors operating together that it is 

 for the present impossible to handle it with that precision which 

 is the sine qua noil of scientific procedure. Science in our day is 

 minutely analytic, dissective ; it seeks to single out from habitual 

 associates particular factors or phenomena and investigate their 

 properties. It attaches relatively little value to investigation 

 which cannot control the conditions accompanying a phenome- 

 non to be studied; for when we take a thing in the large, when 

 we cannot separate it from its milieu, so to speak, it is not easy 

 to see what elements are really most active in determining its 

 nature. But now so far as psychological experiment is con- 

 cerned, it is utterly impossible by any known methods of in- 

 vestigation to cut off elementary psychical processes wholly 

 from each other. Experimental psychology may attempt to 

 investigate an isolated sensation, for instance ; but such a thing 

 is a pure abstraction. In the intricacy of the psychic organism 

 one element never can be entrapped apart from others with 

 which it has become inseparably associated through experiences 

 no less in phylogenesis than in ontogenesis. The most that can 

 be done is to limit the run, as it were, to hedge it in on every 

 side; to have, that is to say, as few factors as may be enter- 

 ing into a problem. The results of the most accurate experi- 

 ments then in the mental sciences are only relatively more ex- 

 act than those of observation or experience. It may be permis- 

 sible to point out in passing that in our time there is some 

 danger of experimental psychology conveying the impression 

 that it is more precise than it ever can be. In a considerable 

 part of present-day scientific writings one is led to think that 

 the most primary phenomena have been examined with mathe- 

 matical accuracy, when as a matter of fact there must have 

 been left out of account accompanying conditions which deter- 

 mined to a greater or less extent the behavior of the thing 

 studied. 



