A long-simmering debate over the purpose of dreaming takes a surprising turn with a Columbia ocular physiologist's hypothesis. Whether or not REM sleep exists to stir the eye, David Maurice has stirred up several disciplines

The biology of dreaming:
a controversy that won't go to sleep

Maury M. Breecher

N o one would normally consider David Maurice, Ph.D., professor of ocular physiology in the Department of Ophthalmology at Columbia-Presbyterian Medical Center, a revolutionary. Nevertheless, he has reignited a decades-long controversy that could spark a revolutionary re-evaluation of an entire field of behavioral research. Dr. Maurice has developed a startling new line of scientific inquiry that, when added to other findings, could change our understanding of rapid eye movement (REM) sleep and the nature of dreams.

What Maurice has done is to suggest an alternative explanation for the phenomenon known as REM sleep, the stage in which the eyes rapidly move and most dreams occur. Dr. Maurice isn't convinced by currently accepted theories about why REM occurs, including the widespread belief that REM exists mainly to process memories of the preceding day's events during dreams. Instead, he hypothesizes that while they sleep humans experience REM to supply much-needed oxygen to the cornea of the eye. In a recent issue of Experimental Eye Research, Maurice suggests that the aqueous humor--the clear watery liquid in the anterior chamber just behind the cornea--needs to be "stirred" to bring oxygen to the cornea.1

"Without REM," Maurice told 21stC, "our corneas would starve and suffocate while we are asleep with our eyes closed."

Maurice's interest in REM began a few years ago. "I wondered why animals born with sealed eyelids needed REM or why fetuses in the womb experience great amounts of REM. The processing-of-cognitive-signals theory doesn't seem to explain those occurrences." He developed his hypothesis after learning of a young man whose eyes had been immobilized by an accident and whose corneas had become laced with blood vessels, presumably to supply the corneas with oxygen. The Columbia eye expert knew that when the eyes are closed during non-REM sleep, oxygen can reach the cornea from the iris only by diffusion across the stagnant aqueous humor. Using a mathematical model, he established that oxygen supplied under those conditions would be insufficient. That realization led to his proposal that REM exists to bring oxygen to the cornea.

Often an idea from one scientific realm has important consequences to another, and Maurice's hypothesis has ramifications for a cross-disciplinary controversy in the fields of cognitive neurobiology and dream research. The debate concerns the relation between physiologic events (such as REM) and dream creation. Neurobiologists and neuropsychiatrists tend to think of dreaming sleep as "physiologically determined" and shaped by the activation of brain neurons, according to J. Allan Hobson, M.D., a Harvard Medical School psychiatrist who directs the Neurophysiology Laboratory of the Massachusetts Mental Health Center. The implications of this activation hypothesis contrast sharply with the psychoanalytic view of the dreaming process.

"What is at stake here is a theory of dreams that is scientifically valid," Dr. Hobson told 21stC. "If psychoanalytic dream theory is not scientifically valid, then psychoanalytic dream interpretation is not scientifically valid. I believe it is not."

Interpretation vs. observation

Dreams, Sigmund Freud famously stated, are "the royal road to the unconscious." Psychoanalysts still believe that dream contents are unconscious manifestations of subconscious mental turmoil and that the examination and understanding of that content can help dreamers rid themselves of inner conflicts. Many of today's psychoanalysts take a more pluralistic approach than the field's founder did: "Modern psychoanalytic thought now holds that dreams are just one of several roads to the unconscious," according to Arnold Richards, M.D., editor of the Journal of the American Psychoanalytic Association. Still, these practitioners separate the psychological from the physiological and view dreams primarily as grist for interpretation. Observations in the biological sciences, on the other hand, have given their discipline a bracing challenge.

In 1953, roughly a half century after Freud posited his theories, sleep researchers Eugene Aserinsky and Nathaniel Kleitman discovered that dreaming was associated with REM sleep. They foun d that sleepers could recall dreams most frequently if they were awakened when their eyes appeared to be moving rapidly beneath their eyelids. This discovery gave researchers a tool with which to monitor dreams. Over the years, findings of those studies caused several scientists to challenge Freudian dream interpretation theory. A landmark 1977 scientific paper by Hobson and his Harvard colleague Robert W. McCarley, M.D., boldly called for "important revisions" in psychoanalytic dream theory. Drs. Hobson and McCarley proposed what they called "the activation-synthesis model of dream production," explaining,

The primary motivating force for dreaming is not psychological but physiological since the time of occurrence and duration of dreaming sleep are quite constant, suggesting a preprogrammed, neurally determined genesis . . . [I]t casts serious doubt upon the exclusively psychological significance attached to both the occurrence and quality of dreams.2

To Hobson, dreams are reactions to random nervous system stimuli, which the brain "interprets" as bizarre images and other sensory hallucinations. "The activation-synthesis hypothesis," he asserts, "assumes that dreams are as meaningful as they ca n be under the adverse working conditions of the brain in REM sleep. The reason that dream content often seems disoriented or bizarre is because the activated brain-mind does its best to attribute meaning to the internally generated signals of the brain."

Adding fuel to the controversy, in 1983, Nobel laureate Francis Crick and his Salk Institute colleague Graeme Mitchison argued in a Nature article3 that the brain's neural memory systems are easily overloaded and that humans experience dream-laden REM to eliminate cognitive debris. In other words, dreams are nothing more than a mechanism for the nervous system to clear the brain of unnecessary, even harmful memories.

Drs. Crick and Mitchison called their theory "reverse learning" and quipped in their 1983 Nature article that "We dream to forget." In essence, they described dreams as garbage to be discarded from memory. In a later article in Behavioural Brain Research, Crick and Mitchison stated, "There is no evidence to suggest that remembered dreams are anything more than an accidental by-product of this (REM) function"; furthermore, they directly attacked psychoanalytic theory by writing, "To a modern neuroscientist Freud's theories, in spite of their appeal to the contemporary imagination, seem little better than the common belief in earlier times that dreams foretold the future, a belief which also held strong intuitive appeal."4 Their views left little room for the idea that it is psychologically valuable to analyze dreams.

Hobson adds, "Psychoanalysts want people to believe that they can interpret dreams and discover deep-seated meanings that are at the root of the dream process. I just don't think there is any scientific reason to believe that."

No final answer in sight

Freudians and their allies have responded to such attacks robustly. For instance, discussing Maurice's ophthalmologic hypothesis, Dr. Richards retorts that "A scientist can develop an understanding of the physiological function of dreaming and still not know anything about the meaning of dreams, because one is a physiological phenomenon and the other is a psychological phenomenon. It may be that REM dreaming does something to the cornea, but that doesn't say anything about the nature of dreams and their m eaning and use in psychoanalysis. One hundred years of psychoanalytic research and experience show that there is much that can be learned about the mental and emotional lives of people by dream interpretation and other psychoanalytic methods."

Neil B. Kavey, M.D., director of Columbia-Presbyterian Medical Center's Sleep Disorders Center, is one of the few psychiatrists in the world who is board-certified in both psychoanalysis and sleep medicine. He points out that proponents on each side of the controversy draw on support from research in their respective fields.

"We know that REM is associated with many physiological processes and effects such as changes in breathing, changes in blood flow to the brain, and changes in brain activity," explains Dr. Kavey. "If Dr. Maurice's hypothesis is proven, it would add to our knowledge about another function of REM sleep, but it wouldn't invalidate the principles of psychoanalysis. "Although Dr. Hobson and others correctly observe that REM is strongly associated with physiological causes and effects," Kavey continues, "it doesn't follow that it is useless for psychoanalysts to treat patients by helping them interpret their dreams."

Experts on both sides of the issue agree that more research is needed to better understand REM and the functions of dreams. Some commentators, indeed, come close to contending that "function" is not always what dreaming is about, liberating some dreams from goals and causality altogether. "Neither the classical psychoanalytic approach nor the physiologic attacks on it have been, in our opinion, been able to explain fully the purpose or function of dreaming," wrote Harvard Medical School psychiatrist Ramon Greenberg and three of his colleagues in a recent issue of the Journal of the American Psychoanalytic Association.5

So the controversy about what dreams are for--if they are "for" anything--continues. David Maurice, however, is opting out of this debate. "My interests are in the physiological mechanisms, the plumbing of the eye. I am happy to leave the subject of dreams to others."


1. Maurice DM. An ophthalmological explanation of REM sleep. Exp Eye Res 66 (1998): 139-145.

2. Hobson JA, McCarley RW. The brain as a dream state generator: an activation-synthesis hypothesis of the dream process. Am J Psychiatry 134 (1977):1335-1348.

3. Crick F, Mitchison G. The function of dream sleep. Nature 304 (1983): 111-114.

4. Crick F, Mitchison G. REM sleep and neural nets. Behav Brain Res 69 (1995): 147-155.

5. Greenberg R, Katz H, Schwartz W, Pearlman C. A research-based reconsideration of the psychoanalytic theory of dreaming. J Amer Psychoanal Assn 40 (1992):531-550.


Related links...

  • Columbia Record article on David Maurice's studies

  • Sleep Home Pages, UCLA Medical School

  • Association for the Study of Dreams

  • The Stages of Sleep and Shift Work, Shiftworker Online

  • Stephen LaBerge, "Lucid Dreaming: Psychophysiological Studies of Consciousness during REM Sleep," in Bootzen, R. R., Kihlstrom, J.F. & Schacter, D.L. (eds.), Sleep and Cognition (Washington, D.C.: American Psychological Association, 1990), pp. 109-126.

  • Centre for Sleep Research, University of South Australia

  • J. Allan Hobson's sleep-studies laboratory, Massachusetts Mental Health Center

  • Neuroscience Information Resources, Society for the Neural Control of Movement, University of Massachusetts

  • Milestones in Neuroscience Research, Eric H. Chudler, University of Washington

  • Neurologist William Calvin's meditations on brains, evolution, and the future

  • "The second state of consciousness: Sweet dreams," from New Scientist symposium "Toward a science of consciousness"


  • MAURY M. BREECHER, M.P.H., Ph.D., author of Healthy Homes in a Toxic World (NY: John Wiley & Sons, 1992), is a member of the American Society of Journalists and Authors and the National Association of Science Writers.


    Art Credits:
    Illustrations: Howard R. Roberts