PROBLEMS OF EXPERIMENTAL PSYCHOLOGY 683 



record at the end of each experiment. We worked at the problem 

 for a year, only to learn that we had been too ambitious; we had, 

 as even with experience one is apt to do, underestimated the com- 

 plexity of consciousness. At the same time, we decided that the 

 problem was soluble; we gathered in a good store of introspective 

 results, even if they were too individual, and too discrete, to be 

 employed for generalization; with more time and more observers, 

 or with a simpler set of voluntary movements for study, we should 

 have accomplished something for psychology. I regard such studies 

 as those recently made on the control of the retrahens of the ear, 

 or on the control of the winking reflex, as extremely promising in 

 this field. At any rate, whether we work from the classical reaction 

 experiment, or whether we take voluntary movement under more 

 natural conditions, the problem is quite definite: we must submit 

 action to an introspective analysis as detailed and as searching as 

 that to which we have subjected perception. 



I have put off (7) imagination, because I am a little afraid of the 

 term. It is a word which, like perception, I should be glad to see 

 discarded from the vocabulary of experimental psychology. I think 

 that we employ it more vaguely even than we employ perception; 

 and I think that the future will substitute for it a number of de- 

 scriptive terms. If we begin with the elementary process, the image 

 itself, we must plead ignorance on two fundamental points: whether 

 image-quality is coextensive with sensation-quality, and whether 

 image-difference is adequate to sense-discrimination. If we go to 

 the other extreme, and regard imagination as the general name for 

 a group of typical formations, as a concept coordinate with 

 memory, we must surely say that experimental psychology is, as 

 yet, hardly over the threshold of the subject. We know, perhaps, 

 how to set to work: some investigations have been made, and some 

 hints toward method have been given; but, in the large, this chapter 

 of experimental psychology remains to be written. 



(8) Of the more complex affective formations we can say but 

 little until we have a better psychology of feeling. No doubt, there 

 are certain problems in the psychology of sentiment, and more 

 especially in that of the esthetic sentiments, that can, within limits, 

 be handled without regard to the ultimate categories of feeling. I 

 should, however, consider these limits as very strictly drawn. 



(9) For the higher intellectual processes we have, I think, three 

 sources of knowledge: direct experiment, that, as you know, has 

 been well begun, the indirect results of experiment upon sensa- 

 tion, and Volker psychologic. I am inclined to lay great stress upon 

 the second of these sources. Experimental psychology has often 

 been reproached, on the one hand, because it devotes most of its time 

 to sensation, and on the other because the results of its dealings with 



