[up] [home]



Presented at the New York Divisional Meeting of the American Psychiatric Association, November 16, 1957.

The notion that dreams exert a behavioral effect is generally taken for granted. It is implicit, for example, in the concept of the dream as the guardian of sleep. The dream is thought of as the attempt by the sleeping individual to perpetuate a certain form of behavior, in this instance, sleep. As important as this idea is to the classical interpretive theory of dreams and much as it accords with commonly shared beliefs about dreams, it is one which, to the best of my knowledge, has not as yet been experimentally validated, nor have there been any studies until quite recently which focus directly on this specific problem. This is a point of no little interest when one considers how much of Freud's thinking about dreams, including the role of wish-fulfillment and the importance of dream work, is so intimately related to the sleep-preserving function of the dream. Freud postulated that the gratification of wishes by direct representation of fulfillment, or as a result of the dream work, allowed the individual to continue in a state of undisturbed sleep by disposing of potentially bothersome stimuli arising during the night.

The recent reports of Aserinsky and Kleitman (1, 2), Dement (3) and Dement and Kleitman (4, 5, 6, 7) shed considerable light on this area and provide us with some of the first definitive observations on the relationship of dreaming to the sleep-waking cycle. I have reference to their work on the correlation between the occurrence of rapid eye movements during sleep and dreaming. The most complete statement of their results is contained in two reports now in press. This account is based on these as well as on the earlier ones. Simultaneous oculographic and electroencephalographic tracings were made on subjects over the course of many nights of sleep. At times other data were included in the record, e.g., body movements, respiratory and cardiac rates. In the later studies the tracings were made continuously throughout the night. The leads taken from the eyes recorded action potentials when the eyes moved in any direction. The rapid conjugate eye movements (REM) noted were easily distinguishable from the slow, dysynchronous movements previously known to occur during sleep. Without going into details concerning the many varied techniques they used to control their observations, the significant findings may be summarized as follows:

(1) Bursts of rapid eye movements occurred in discrete periods during the night in each subject on every night they slept. The number of such episodes varied from two to five depending on the total duration of the sleep period. The mean time of onset of the first such period was one hour after the onset of sleep. Intervals of movements lasted from one to 72 minutes. The time between intervals tended to be shorter as sleep progressed, whereas the duration of the movements tended to be longer with each successive period. The pattern of occurrence of these movements revealed stable individual variations. The movements were binocularly synchronous and occurred in all directions. The average total time for these movements was 64 minutes or 17.7 per cent of the sleeping time. No rapid eye movements were ever noted during the initial onset of sleep.

(2) A strikingly consistent relationship was noted between the periods of REM and the EEG. They were always associated with a stage 1 pattern, the latter defined as any EEG pattern between full wakefulness and the appearance of spindles, and characterized in the main by low voltage fast activity. According to their findings "the eye movement periods and sleep onset accounted for nearly all the stage 1 patterns" (6).

(3) An 80 per cent incidence of dream recall is reported after close to 200 awakenings during periods of REM over many nights of sleep. There was only 7 per cent recall with random awakenings when the eyes were quiescent, and most of these were associated with awakenings occurring soon after the end of a period of REM. The character of the eye movements, e.g., whether predominantly in the vertical or horizontal plane, appeared to be specifically related to the imagery occurring in the dream.

(4) External stimuli do not initiate dreaming. When applied during periods of REM they may to some extent modify dream content. When given during periods of ocular quiescence, however, they do not initiate eye movements nor is there any subsequent dream recall. Their observations further lead the authors to the belief that dreams are initiated by intrinsic physiological mechanisms related to the sleep-wakefulness cycle rather than by psychological ones.

If we were to summarize these findings, it would appear that there are rhythmical variations in the sleeping phase of the sleep-waking cycle characterized by the occurrence of repetitive bouts of arousal activity, initially falling short of complete behavioral arousal, but culminating finally at the end of the sleep cycle in complete awakening. These bouts occur progressively closer together and each succeeding bout lasts somewhat longer than the preceding one. It is as if the sleeping individual had a built-in mechanism designed to insure arousal independent of prevailing environmental stimuli. This mechanism is fired more often and for longer periods as the need for sleep diminishes. After a number of such dress rehearsals in the service of awakening, the arousal process is finally successful and awakening occurs. These arousals are initiated in the absence of external stimuli. When the latter occur in periods of ocular quiescence they are either subliminal and without effect or they bring about a sudden transition to the waking state. These episodic states of partial arousal are characterized by three constant features:

(1) The occurrence of an activated EEG pattern of low voltage, fast activity, and the absence of spindling.

(2) The occurrence of bursts of rapid eye movements.

(3) The subjective experience of dreaming or of having dreamed during this period.

The experimental data are in harmony with the notion of arousal from below, but go further by linking dreams to arousal processes and the waking phase of the sleep-wakefulness cycle. The studies on the reticular system suggest the possible brain mechanisms involved (8, 9, 10, 11, 12). The existence of reverberating connections between the brain stem reticular substance and the cortex allows the cortex to participate in its own activation (13). Once cortical activation comes about, the subjective accompaniments experienced as the dream may play a role in: determining the end point of the process, e.g., awakening or a return to deep sleep. In this role, however, they are only one among the many factors impinging upon the controlling subcortical centers. If dreaming does in fact occur over an extended period of time, it is likely that only the terminal phase of the period of dreaming would be con-concerned with a behavioral effect.

The fact that repeated partial awakenings are so intrinsically a part of sleeping and the fact that these awakenings are associated with dreaming suggests the possible relationship of dreaming to the vigilance needs of the sleeping organism. These needs exist for the human as well as for lower animals when faced with the task of reestablishing connection with the external world. Whereas lower animals may handle vigilance needs on awakening by first alerting themselves to information they can gather from their distance receptors, the human organism functions differently, but toward the same goal, that of safety. Safety, however, for the human is now a matter of the nature of his relatedness to his fellow humans, rather than primarily a question of physical protection. It now has to do with the relative intactness of patiently erected and carefully maintained symbolic systems. Elsewhere (14, 15) I have attempted an analysis of the form and content of dreams in connection with problems of vigilance. It was suggested that the characteristic form of consciousness while dreaming (the use of imagery, the hallucinatory quality, the reduction of the abstract to the concrete) was not a regression phenomenon, but was simply an appropriate form of expression considering the sensory function subserved by consciousness at the time. Another way of expressing the same idea would be to consider that, during the time that the cortex becomes engaged in the process of self-arousal, thought becomes afferent to the subcortical systems. The concrete sensational form is consistent with this function. In a like manner, dream content can be viewed from the perspective of vigilance. The dream reflects any currently acting alerting stimuli or any recalled disturbing psychological stimuli from the immediate past. The sleeping organism is faced with the need to effect a transformation from a state of absent consciousness to the state of waking consciousness. There is a broader need for vigilance under these circumstances as compared with the waking state where there are habitual sources of security and where no such drastic change occurs. In line with this the dream telescopes relevant material from the longitudinal life experience of the individual into the current scene. In so doing, the dream symbolically but accurately depicts the resultant stresses and strains occurring in the wake of disturbing current situations.

Viewed from this point of view, the principle of wish fulfillment loses its motivational connotations but remains as an empirical fact associated with many, though not all, dreams. In the so-called dreams of expediency, where the wish motive appears so strikingly, the point would be that it is not the wish which has preserved sleep, but rather that the preservation of sleep is reflected subjectively in what appears to be a wish-fulfilling fantasy. This may be illustrated by the following sequence of events:

At some point in sleep I find myself beginning to awaken and I become aware of the urge to empty my bladder. For some reason I do not awaken, but find myself returning to a deeper sleep. As this is occurring I dream of going to the bathroom and urinating. I am no longer uncomfortable and I continue to sleep peacefully.

The emphasis here is not that the dream in any direct fashion preserved sleep, but rather that the stimulus, namely, the urge to micturate, was not great enough to bring about full arousal. In the course of the return to deeper sleep the stimulus loses its effectiveness. The individual is somewhat of an innocent bystander to this interplay and all he can do is to reflect the disappearance of the disturbing stimulus by creating a series of visual images depicting himself in the act of disposing of the stimulus in a way familiar to him from his past experience in the waking state.

If we were to condense our remarks about dreams into a statement of their essential features, this would include:

(1) The form the dream takes is that of a hallucinatory episode whereby the dreamer confronts himself with aspects of his past experience expressed at a concrete, sensational level.

(2) Dreaming is an involuntary experience in the sense that regardless of the part we may play in the dream, the total effect is that of something intruding itself into awareness over and above any desires we may have in the matter.

(3) The content is totally experiential in the sense that it is derived from internal sources and is not structured by currently existing referential cues from the environment.

(4) Content becomes telescoped under these circumstances and cuts across temporal and spatial boundaries to link related fragments of experience together.

I wish to consider one further point and that is the relevance of these interpretive remarks on dreaming to phenomena noted in states of sensory deficit. If we regard dreaming as a state of relative de-afferentation, as Hebb (16) has suggested, the above formulation may be reduced to two general postulates applicable to other such states.

(1) In response to vigilance needs operative during states of relative de-afferentation, an inner source of afferent stimuli replaces the absent external source. Thought becomes afferent to subcortical structures in the interest of further cortical activation. The prime requisite under these circumstances is to bring about or maintain a state of behavioral arousal as a first or basic requirement for effective action.

(2) In response to orienting needs during the states of relative de-afferentation, the individual attempts to cope with the questions "who am I and what is happening to me?" by expressing his situation in such a manner that relevant past experience is brought to bear upon a currently disturbing life situation and both are reduced to concrete sensational terms.

These postulates would hold regardless of how the de-afferenting process comes about, whether it be as a result of pathological changes in a sensory end organ as in the case of visual hallucinations associated with cataract formation (17), experimental deprivation of afferent stimuli (18, 19), or failure to achieve effective sensory integration as a result of brain damage (20). In each of these situations the individual appears to confront himself as well as others with an autistically created environmental response at precisely those points where sensory integration with the environment is no longer possible.


(1) The guardianship of sleep function assigned to dreams is not compatible with recent physiological findings concerning the nature and extent of dreaming.

(2) What is referred to as a dream is actually an extended period of consciousness occurring during the lightest stage of sleep. It is associated with a terminal behavioral, effect which may be either that of awakening or returning to a deeper sleep.

(3) The evidence cited indicates a close relationship between dreaming and the arousal processes.

(4) This relationship was further explored in connection with the problems of vigilance and orientation involved in effecting the transition from sleep to wakefulness.

(5) The qualitative similarity between dreaming and other states of sensory deficit is brought into sharper focus by a consideration of the altered nature of vigilance needs and the difficulties in orientation that characterize these states.


1. Eugene Aserinsky and Nathaniel Kleitman: "Regularly Occurring Periods of Eye Motility, and Concomitant Phenomena, During Sleep." Science, Vol. 118, No. 3062, September, 1953.

2. -----: "Two Types of Ocular Motility Occurring in Sleep." J. Applied Physiol., Vol. 8, No. 1, July, 1955.

3. -----: William Dement: "Dream Recall and Eye Movement during Sleep in Schizophrenics and Normals." J. Nerv. and Ment. Dis., Vol. 122, No. 3, September, 1955.

4. ----- and Nathaniel Kleitman: "Incidence of Eye Motility during sleep in Relation to Varying EEG Pattern." Federation Proceedings, Vol. 14, No. 1, March, 1955.

5. -----: "The Relation of Eye Movements during Sleep to Dream Activity: An Objective Method for the Study of Dreaming." J. Exp. Psychol., Vol. 53, No. 5, May, 1957.

6. -----: "Cyclic Variations in EEG during Sleep and their Relation to Eye Movements, Body Motility, and Dreaming." EEG Journal (in press)

7. -----: "The Relation of Eye Movements, Body Motility, and External Stimuli to Dream Content" J. Exp. Psychol. (in press)

8. W. R. Hess: "The Diencephalic Sleep Center." In Brain Mechanisms and Consciousness. Charles C. Thomas, Springfield, Illinois, 1954.

9. Herbert Jasper, Cosimo Ajmone-Marson and Julius Stoll: "Corticofugal Projections to the Brain Stem." Arch. Neurol. and Psychiat., Vol. 67, No. 6, 1952.

10. H. W. Magoun: "An Ascending Reticular Activating System in the Brain Stem." Arch. Neurol. and Psychiat., Vol. 67, No. 6, 1952.

11. Montague Ullman: "Physiological Determinants of the Dream Process." J. Nerv. and Ment. Dis., Vol. 124, No. 1, July, 1956.

12. -----: "Hypotheses on the Biological Roots of the Dream." J. of Clin. and Exp. Psychopathology. (in press)

13. Frederic Bremer:. "The Neurophysiological Problem of Sleep," In Brain Mechanisms and Consciousness. Charles C. Thomas, Springfield, Illinois, 1954.

14. Montague Ullman: "The Dream Process." Psychotherapy, Vol. 1, No. 1, Fall, 1955.

15. -----: "Dreams and the Therapeutic Process." Psychiatry. (in press)

16. D. O. Hebb: The Organization of Behavior. John Wiley and Sons, New York, 1949, pp. 217-218.

17. J. E. A. Bartlett: "A Case of Organized Visual Hallucinations in an Old Man with Cataract, and their Relation to the Phenomena of the Phantom Limb." Brain, Vol. 74, Part III, 1951.

18. W. Heron, W. H. Bexton and D. O. Hebb: "Cognitive Effects of a Decreased Variation in the Sensory Environment" American Psychologist, Vol. 8, No. 8, 1953.

19. W. H. Bexton, W. Heron and T. H. Scott: "Effects of Decreased Variation in the Sensory Environment" Canad. J. Psychol., Vol. 8, No. 2, 1954.

20. Edwin A. Weinstein and Robert L. Kahn: Denial of Illness. Charles O. Thomas, Springfield, Illinois, 1955.