Human Dream Processes as Analogous to Computer Programme Clearance

Abstract

RECENT experiments¹ have led to a renewed interest in the very long-standing problem of dreams. With perhaps the single exception of the contribution of the nineteenth-century psychoanalysts, dream theories have been based on negligible experimental evidence and have never been sufficiently rigorous or explicit to attract serious scientific interest. The work of Dement, however, demonstrates what appears to be a convincing behavioural measure of dream periods and, though criticized on several counts, has had an undoubted effect from the point of view of stimulating further experimental investigations and theoretical considerations. We have ourselves recently proposed elsewhere² that the dream process might be likened in function to the systematic programme clearance which is absolutely necessary where computer programmes are being continuously evolved to meet changing circumstances. The greater the change in circumstances being programmed for, the greater must be the amount of programme evolution and the more urgent the programme clear-out. In our view, the primary function of sleep is probably to allow such a clearing process to get under way without interference from external information; ‘dreams’ occur when the level of consciousness shifts for one reason or another and the clearing process is interrupted. Prolonged deprivation of the opportunity to dream would inevitably produce a breakdown in human efficiency, most probably in the region where novel situations must be handled. More recently yet, Kales et al.³ have attempted an experimental reappraisal of Dement's original findings; but, while depriving subjects of the opportunity to dream for a number of nights, they did not notice any significant deterioration in their performance in certain psychometric investigations. In substance they agree that dream-prevention leads to increased attempts on the part of subjects to dream, but noted no ‘psychic changes’ as the result of this deprivation. However, the apparent discrepancy between the results of Kales and those of Dement is not, we believe, serious for present theoretical interpretations of the dream-deprivation investigations, for a number of reasons which we shall consider here.