796 



SCIENCE. 



[N. S. Vol. XX. No. 519. 



form that I have described— the form of 

 introspective analysis. I have little sym- 

 pathy or patience with those experimental- 

 ists who would build up an experimental 

 psychology out of psychophysics and logic ; 

 who throw stimuli into the organism, take 

 reactions out, and then, from some change 

 in the nature of the reactions, infer the 

 fact of a change in consciousness. Why in 

 the world should one argue and infer, when 

 consciousness itself is there, always there, 

 waiting to be interrogated? This is but a 

 penny-in-the-slot sort of science. Com- 

 pared with introspective psychology, it is 

 quick, it is easy, it is often showy. We 

 have been a little bit corrupted by the early 

 interest in psychophysics; or perhaps, 

 more truly, we have not all learned in- 

 stinctively to distinguish between psycho- 

 physics and psychology proper; and so we 

 are apt to take the tables and curves of 

 reactions for psychological results, and the 

 inferences from them for psychological 

 laws. Now the results, where they are not 

 purely physiological or anthropometrical, 

 are psychophysical results. As such, they 

 have their usefulness; and the psycholog- 

 ical laboratory is their right place of or- 

 igin. But there is no reason why one 

 should gain psychological credit for them 

 — still less for erecting a speculative psy- 

 chology upon their foundation. This mode 

 of psychologizing is inherently as vicious 

 as any of the constructive modes of the 

 older psychology, the psychology before 

 experiment. Historically, it has proved 

 disastrous ;* it falsifies problems and ob- 

 scures real issues; we must set our faces 

 against it now and for all time. How, 

 indeed, shall one call a man a psychologist 

 who deliberately turns his back upon the 



* Is proof needed ? Think of the early work 

 upon the just noticeable difference, upon the 

 simple reaction, upon the ' time sense ' ; or think 

 of Wundt's current discussion of Weber's and 

 Merkel's laws! 



one psychological method, in the one field 

 to which that method directly applies? 

 There is no excuse, in psychology, for the 

 neglect of ijitrospectiou, save the one— and 

 that must be demonstrated— that intro- 

 spection is impossible. 



Having said this much by way of pref- 

 ace, I may take up the further question. 

 We can hardly open a magazine nowadays 

 without finding applications of the experi- 

 mental method beyond the limits of the 

 normal, adult, human mind. In animal 

 psychology, in child psychology, in various 

 departments of mental pathology, the ex- 

 perimental method is employed. Even the 

 conservative Studien contains articles on 

 the state of sleep and dreaming, and 

 Wundt has looked more favorably upon 

 experiments under hypnosis since they 

 promise to confirm his theory of feeling. 

 Experiments on children and animals have 

 for some years past occupied the attention 

 of leading American psychologists; work 

 on child psychology is characteristic of the 

 Annee psychologique, and is being pub- 

 lished more and more freely by the Zeit- 

 schrift; you all know the avowed purpose 

 of Kraepelin's 'Arbeiten.' I need not 

 multiply references. Wherever psycholog- 

 ical interest has gone, in these fields, the 

 experimental method has gone with it. 

 Sometimes the particular experiment is 

 borrowed forthright from the normal prac- 

 tise of the laboratory, sometimes the pro- 

 cedure has been recast to suit the novel 

 problem ; sometimes the experimental 

 method is taken seriously, employed with 

 care and knowledge, sometimes it is 

 thrown in as a makeweight, without re- 

 sponsibility or understanding; sometimes 

 it is praised, sometimes decried. All this 

 is natural. The important thing for us is, 

 I think, the recognition that the experi- 

 ments are a part of 'experimental psychol- 

 ogy,' in the sense of this paper, and must 

 be taken account of in any general review 



