Early behavionsts, troubled about introspection as an appropnate method for psychology, tried to get around the problem of awareness by dealmg only with discnmmation. Thus the "scientific" problem of perception rested for them upon the correlation between stimulus and response mdicated m the act of discnmmatmg, 1 e, respondmg differently to the presence or absence of stimulation or m the presence of two unlike stimuh Awareness, -hewever, has another facet the subject's awareness that he is discnmmatmg This is a puzzhng problem m any case, no matter what one's relation to behaviorism A sophisticated objectivism, m the form of operatiomsm, is free to deal with this problem through convergmg operations (Gamer, Hake, & Enksen, 1956), but this^freedom to work on the problem does not make its solution any easierThe facts of perception force a few simple conclusions upon the mvestigator (a) that much perception is highly correlated with the external stimulus, le, corresponds to reahty, (b) that some perceptions which appear to be realistic are m fact distorted m one way or another, based on stimulus charactenstics or on the subject's expectations, needs, wishes, or cogmtive styles, and (c) that there are some perception-hke cogmtive expenences^ that bear very httle relation to external stimuh, and seem to reflect largely some kind of "state" of the person, as m dreams, hallucmauons, and depersonalization. Expenmental psychologists have dealt very largely with the fijst of these three sets of facts, somewhat with the 'The expenments on tachistoscopic perceptton referred to here were conducted m coUaboratioa with Judah Landes, and supported by a grant from the Ford Foundation to R R. Sears and E R Hilgard, which led to the establishment of the Laboratory of Human Development at Stanford * The word perception is usually confined to responses closely correlated with a presented stimulus, while cognition is a more general term which includes perception, judgment, imagery, and thought The boundaries of what should be called perceptual are not clear m marginal cases