Five short-term memory experiments are reported which examined whether the instructional set to forget affects trace formation or trace retrieval. The experimental paradigms included a manipulation of covert rehearsal efforts, variation in the temporal point in the memory sequence at which a remember or forget instructional cue is introduced, a "latent memory" design in which 5s perceived identical stimuli under shifting instructional sets, and a procedure in which identical stimuli were repeatedly perceived under constant instructional sets. The results strongly suggest that the motivational set to forget hinders trace utilization, rather than original learning. In addition, the retrieval deficit caused by the instruction was not necessarily mediated by conscious cognitive processes. The relevance of this work to the clinical conception of repression was emphasized.Recent investigations (Bjork, LaBerge, & Legrand, 1968;Weiner, 1968) have demonstrated that instructions to forget perceived stimuli greatly reduce the retention of those stimuli, and lessen the interfering effects of that material on the retention of other stimuli. Why a direction to forget influences retention remains unclear. It may be that the instruction minimizes covert rehearsal, hence retarding the degree of original learning of the stimuli. On the other hand, perhaps the addition of the set to forget results in the distinct storage or differential retrieval of the information. The following five experiments investigated which process in the memory sequence is affected by the instructional set to forget. In addition, the relationship between forgetting instructions and the clinical conception of repression is discussed. EXPERIMENT I MethodExperiment I consisted of two collateral studies; one investigation employed a within-5s design, the

HENRY REED received his PhD from UCLA in 1970 and is a shirtmaker. For the past three years he has also served as a psychology professor at Princeton University, where he has worked with others in a program of education and reality testing to construct interfacings between awakening and dreaming consciousness. Recently, Henry has been consulting with the Clinic and Research Center for Jungian Psychology in Zumch, on the mcubation of healing and inspirational dreams; and with the Association for Research and Enlightenment in Virginia Beach, on a transpersonalized ritual for parapsychological research. Besides talking to plants and animals, his greatest joy is to fly on his bicycle, and whenever possible, on foot. This paper is dedicated to Gayle, whose loving reminders have helped Henry recognize what he knows. ABSTRACTThe possibility of learning to remember dreams was explored with the cooperation of a group of participant-observers who maintained dream diaries for 12 weeks and were selected on the basis of their desire to experience improved dream recall. A self-administered dream recall rating scale was devised which provided measures of several aspects of memory for dreams. The participants apparently increased their ability to retrieve dream memories, but their actual level of dream recall performance was subject to motivational fluctuations. On the other hand, although the participants claimed to have developed no control over the vividness of their memory for dreams, this aspect of their dream recall nevertheless evidenced substantial improvement. The possibility of improving memory for dreams being confirmed, future research should provide ( a ) more information concerning the control processes governing dream recall, and ( b ) information concerning how to increase volitional access to these processes.

An interference effect within short-term memory (STM) was produced by including an extra item for retention. This interference was manipulated by instructions to 5 concerning whether to retain or recall the interfering item. It was found that instructions to forget this item retroactively eliminated its interfering effect. The earlier within the retention interval that the instruction was given, the greater was the reduction in interference. At recall, instructions not to recall the interfering item also reduced interference. Finally, retesling the items individually showed that the initial variation in recall accuracy obtained by manipulating interference had no effect on rctcsl accuracy. It was therefore concluded that although interference in STM may affect the rehearsal process, the result is upon the accessibility for retrieval from STM rather than upon stored availability within longterm memory (T/fM).Recent studies have suggested that interference among memories can he reduced by instructions to forget the interfering material (Bjork, LaBerge, & Legrand, 1968; Rimes, 1969;Turvey & Wittlinger, 1969). Although such a possibility does have potential significance for an understanding of the interference process, these studies have been somewhat ambiguous. To test whether an instruction to forget can reverse an interference effect within retention itself, it is necessary that the instruction not also affect the conditions of initial registration (Keppel, 1965). By their placement of the instruction to forget, these studies (Bjork et al.

Communication accuracy is defined, and various measures of accuracy which have been employed in the research literature are discussed. Considered as the dependent variable, accuracy in communication is conceptualized as being determined by 5 sets of independent variables: attributes of the communicator, of the addressee, of the channel, of the communication, or of the referent. With these as independent variables, a series of hypotheses is suggested for predicting variations in communication accuracy. For each proposed hypothesis, available research findings are reviewed.A survey of the literature on verbal and nonverbal communication reveals a surprising neglect of the direct study of the problems of communication accuracy. Studies which have some bearing on communication accuracy emerge from diverse areas and are interpreted within unrelated frameworks. For example, Festinger (19S7) hypothesized that the probability of erroneous inferences made from a communication is greater when accurate inferences from that communication can induce dissonance in the decoder. He then interpreted related findings from Cooper and Jahoda's (1947) study as providing support for his dissonance formulation. It is interesting to note that Cooper and Jahoda (1947) considered their study as being relevant to the reaction of prejudiced persons to antiprejudice communications. In these instances, then, the relevance of dissonance or prejudice to distortions in the decoding process was bypassed in the interest of relating hypotheses and findings to the conceptual frameworks involved -that is, whereas Festinger's (1957) theory and Cooper and Jahoda's (1947) study do have some bearing on the phenomena of communication accuracy, this aspect of their investigations was not made explicit.The present paper is an attempt to conceptualize variations in communication accuracy within a framework and to relate relevant findings to that framework.

The particular problems of this experiment center about (i) the relation of associative aids to the rate of learning and of forgetting; (2) their relation to other associations; (3) their relation to the transfer of learning; (4) their relation to repeated learning or practice; (5) their relation to thinking and the theory of thought; and (6) their relation to methodology in psychology.