Sigmund Freud's views on the meaning of dreams formed the core of his theory of mental functioning. Mark Solms and others assert that modern science is now validating Freud's conception of the mind. But similar scientific investigations show that major aspects of Freud's thinking are probably erroneous.

For Freud, the bizarre nature of dreams resulted from an elaborate effort of the mind to conceal, by symbolic disguise and censorship, the unacceptable instinctual wishes welling up from the unconscious when the ego relaxes its prohibition of the id in sleep. But most neurobiological evidence supports the alternative view that dream bizarreness stems from normal changes in brain state. Chemical mechanisms in the brain stem, which shift the activation of various regions of the cortex, generate these changes. Many studies have indicated that the chemical changes determine the quality and quantity of dream visions, emotions and thoughts. Freud's disguise-and-censorship notion must be discarded; no one believes that the ego-id struggle, if it exists, controls brain chemistry. Most psychoanalysts no longer hold that the disguise-censorship theory is valid.

Without disguise and censorship, what is left of Freud's dream theory? Not much—only that instinctual drives could impel dream formation. Evidence does indicate that activating the parts of the limbic system that produce anxiety, anger and elation shapes dreams. But these influences are not “wishes.” Dream analyses show that the emotions in dreams are as often negative as they are positive, which would mean that half our “wishes” for ourselves are negative. And as all dreamers know, the emotions in dreams are hardly disguised. They enter into dream plots clearly, frequently bringing unpleasant effects such as nightmares. Freud was never able to account for why so many dream emotions are negative.

Another pillar of Freud's model is that because the true meaning of dreams is hidden, the emotions they reflect can be revealed only through his wild-goose-chase method of free association, in which the subject relates anything and everything that comes to mind in hopes of stumbling across a crucial connection. But this effort is unnecessary, because no such concealment occurs. In dreams, what you see is what you get. Dream content is emotionally salient on its face, and the close attention of dreamers and their therapists is all that is needed to see the feelings they represent.

Solms and other Freudians intimate that ascribing dreams to brain chemistry is the same as saying that dreams have no emotional messages. But the statements are not equivalent. The chemical activation-synthesis theory of dreaming, put forth by Robert W. McCarley of Harvard Medical School and me in 1977, maintained only that the psychoanalytic explanation of dream bizarreness as concealed meaning was wrong. We have always argued that dreams are emotionally salient and meaningful. And what about REM sleep? New studies reveal that dreams can occur during non-REM sleep, but nothing in the chemical activation model precludes this case; the frequency of dreams is simply exponentially higher during REM sleep.

Psychoanalysis is in big trouble, and no amount of neurobiological tinkering can fix it. So radical an overhaul is necessary that many neuroscientists would prefer to start over and create a neurocognitive model of the mind. Psychoanalytic theory is indeed comprehensive, but if it is terribly in error, then its comprehensiveness is hardly a virtue. The scientists who share this view stump for more biologically based models of dreams, of mental illness, and of normal conscious experience than those offered by psychoanalysis.