418 ANNALS NEW YORK ACADEilT OF SCIENCES 



in a primary stimulus-idea through a specific apperception-mass into a 

 derived system of secondary images which form the manifest dream con- 

 tent. ■ The derivation of the latter must be concretely demonstrated in 

 the "settings of ideas^' — ^not assumed. 



The reconstitution of this particular dream illustrates the reductio ad 

 absurdum of the two previous psychoanalytic "solutions."' The fact that 

 either of them would apparently have satisfied the demands of the prob- 

 lem is an artifact evolved by the interpreter's confabulations and forcing 

 of analogy. It is a matter of "will to interpret.''' 



The Freudian technique is unsound in so far as it fails to consider the 

 meaning of dream-items as determined by "unconscious settings of ideas.'' 



The study of individual differences in dreams indicates that the sup- 

 posed "language of dreams" is an artifact ; that the psychic "censorship" 

 is only an occasional phenomenon. The reconstitutive method brings into 

 relief the trial-and-error character of the dreaming process, depicting the 

 organism as attempting a physiological resolution of persisting and unad- 

 justed stimulus-ideas. The images evoked in the dream have the psycho- 

 logical character of "trial percepts" or tentative apperceptions. 



Sleep favors apperceptive errors ; hence the inconsistency and bizarrerie 

 of the dream. The significance of a dream can be found only by recon- 

 stituting it from the above standpoint. 



Dr. Myers said in abstract : Two experiments in progress were re- 

 ported — Reconstructive Eecall and Confusion in Recall. In the first 

 experiment the subjects daily tried to recall as much as possible of certain 

 selections which they once knew very well, but which they had forgotten 

 wholly or in part. The subjects wrote introspections. Several times as 

 much was reconstructed at the end of a few weeks as the amount repro- 

 duced in the first recall. Interesting chains of associations were obvious. 

 Paragraphs and sentences mutilated or entire were recalled and improved 

 upon or linked with others in subsequent recalls. 



The results thus far endorse the common statement of psychologists, 

 "We never wholly forget," and they emphasize the importance of the 

 most favorable situations to elicit the learner's, reproduction. They sug- 

 gest a prominent existence of subliminal association, and, most of all, a 

 serious neglect in almost all memory experiments to consider the time 

 for recall as a factor in measuring memory, and, in case of group experi- 

 ments, to provide any time limit for recall. 



In the second study the purpose is to determine the increase of confu- 

 sion with the increase of retention interval. The following test is used : 



