If dreams are the royal road to the unconscious, as Freud claimed, then that route may be a highway full of tortuous twists and turns—leading nowhere. But it affords some spectacular vistas along the way. By turns, dreams have been deemed prophecies of the future, full of meaning—if only someone could figure out what it is—or the effluence of nerve cells randomly unwinding from a busy day. Once considered a hallmark of the periodic surges of brain activity known as rapid-eye-movement sleep, dreaming now seems somewhat less bundled up; at least 25 percent of dreams are scattered through other parts of the night. Dreaming has been seen as critical for learning, or at least important for solving problems—or as nice but unnecessary. It's an emblem of mental illness—or a safety shield deflecting it. The newest switchback on dreams comes from South African neuroscientist Mark Solms. Maybe, says Solms, we've been confusing cause and effect. Dreams, he suggests, are not a by-product of sleep, as has been assumed all along. Dreaming may be what allows us to sleep in the first place. "Dreams protect sleep," Solms says. They furnish an ersatz world to keep the brain temporarily occupied in its unyielding quest for activity. His iconoclastic view of dreams springs from emerging evidence that REM sleep and dreaming are not synonymous, and that the brain mechanisms involved in REM sleep may be entirely different from those involved in dreaming. Dreaming, in fact, is now thought to recruit areas of the brain involved in higher mental functions. In other words, dreaming does for the brain what Saturday-morning cartoons do for the kids: It keeps them sufficiently entertained so that the serious players in the household can get needed recovery time. Without such diversion, the brain would be urging us up and out into the world to keep it fully engaged. "Dreams are a delusional hallucinatory state" driven by activation of the brain's basic motivational system, Solms told a gathering of scientists in New York City. And like delusions, they appear to be stoked by an abundance of the neurotransmitter dopamine. Dopamine, scientists now know, plays a critical role in directing our attention. The neurochemical decrees what is salient in our environment, regardless of whether that environment is inside us or outside. Under dopamine's influence, events or thoughts jump out of the background, grab our attention, move us to act and drive goal-directed behavior. Dreams trick us into thinking we're out striving in the wider world. "The fundamental problem of being alive is that we must get all our needs met in the outside world," says Solms. The brain has an answer to that; it has developed a kind of unified motivational force variously called the "seeking" or "wanting" system, an orchestration of primitive and higher neural structures that orients us to the outside world with an air of anticipation and positive expectancy. As Solms puts it, "It's an all-purpose looking-for-pleasure-in-the-world drive" that sends animals out to satisfy their needs. Pioneering neuroscientist Jaak Panksepp describes the seeking system as a "goad without a goal" (goals, such as gathering food, being dictated by the specific situation). It is a readiness for action, an appetitive arousal, the neurobiological descendant of Freud's idea of libido. Representing a very basic function of the brain, it commands and activates an array of neural circuits. Researchers mapping the functions of the brain have shown that the hallucinations of psychosis involve hyperactivation of the seeking system's structures. They also involve dysregulated dopamine transmission. Increasingly viewed as "the wind of the psychotic fire," dopamine prompts the brain to assign abnormal importance to its own internal representations. Delusions, in other words, are errors of salience attribution. We overvalue our own thoughts, which are mistaken for perceptual experience of the world. Dreams share many qualities with hallucinations. They are the hallucinations we all experience. Both dreams and hallucinations involve intensive activation of the seeking system. And Solms points to accumulating evidence that dreaming, like hallucinating, is driven by dopamine. French neuroscientist Claude Gottesmann reported that dopamine release in the brain's nucleus accumbens, a site long recognized to be involved in the hallucinations of schizophrenia, is maximal during dream sleep. "Dreaming and schizophrenia have the same neurochemical background," Gottesmann says. Other studies show that the dopamine-boosting drug L-dopa, commonly used to treat Parkinson's disease, prompts people to have more dreams, more emotional content to their dreams and more bizarre dreams. Driven by dopamine, dreams fill our minds with myriad stimuli that feel worthy of our attention, says Solms. "That's necessary because the body is withdrawn from the external world." Goaded into seeking but blocked from action by paralyzing neurochemicals released during dream sleep, we feed on our own internal representations of the world. And we wake hungry for new experiences that build our psychic cinema of internal representations. Says Solms, "The dopamine hypothesis is at the core of why we dream."