Neurological tests provide evidence that Freud's theory of dreams may have been right, says Mark Solms
It is not very fashionable now to believe in Freudian theories about how the mind works. Psychological explanations of mental functions are taken seriously only if they can be confirmed by neuroscientific experiments. Research that suggests that Freud's theory that dreams are caused by wishes and desires will probably prove controversial. But that is what my research indicates.
Neurological testing of theories about the origin of dreams began in the 1950s, when rapid eye movement sleep was discovered. Eighty per cent of dreams occur in this state of agitated sleep. In the 1960s and 1970s it was found that REM sleep is switched on by the pons, a part of the brainstem, and that the switch is a chemical one. This seemed proof that dreams are generated automatically by a chemical switch in a lowly part of the brain. It seemed that dreaming had nothing to do with higher mental functions such as "wishing".
But this mechanical conception of dreaming was based on an erroneous assumption. The researchers who isolated the brain mechanisms that control REM sleep assumed that the same mechanisms also control dreaming. This was understandable as the relevant experiments were performed on cats, who cannot tell us if they are dreaming or not. But it overlooked the "almost" in "almost all dreaming occurs in REM sleep".
What generates the 20 per cent of dreams that occur outside REM sleep? Why do most of these non-REM dreams occur soon after we fall asleep or just before we wake? And why are some non-REM dreams caused by complex partial seizures that affect only the higher parts of the brain, which control emotion and memory? These have nothing to do with the pons. Clearly there is something missing in the theory that dreams are generated by lower brainstem mechanisms in REM sleep.
I began to question the REM mechanical theory of dreaming when a patient told me that he had stopped dreaming since his stroke - which damaged only the higher forebrain, which controls spatial imagery. Intrigued, I studied more than 350 cases to determine how damage to different parts of the brain affects dreaming. I learned that people stop dreaming only when two specific parts of the higher forebrain are damaged. Damage to the lower brainstem (including the pons) does not result in dreamlessness.
It is perhaps not surprising that the part of the brain that controls spatial imagery is centrally implicated in dreaming. But a more interesting finding concerns the ventromesial quadrant of the frontal lobes. Why should this region be essential for dreaming?
It contains a massive fibre pathway that transmits the chemical dopamine, and this pathway controls appetitive strivings (what Freud would have called "libidinal wishes"). Neurobiologists call this pathway the "seeking" or "wanting" circuit. It is stimulated by three drugs - cocaine, amphetamine and L-Dopa - which excite unusually vivid and frequent dreams but have no equivalent effects on REM. We also know that these dreams can be suppressed by dopamine blockers.
These findings suggest that many stimuli can trigger the dream process. REM is the most common; but stimulant drugs, complex partial seizures, falling asleep and waking can all trigger the dreaming process. The word "process" is important. Dreaming itself begins only when the stimulus that triggers the dreaming process attracts appetitive interest.
Appetitive interest (basal forebrain dopaminergic activation) is the final common path to dreaming. If this pathway is damaged by a stroke, for example, REM occurs without dreaming. Likewise, if it is excited by a dopamine stimulant, a limbic seizure or anything else, dreaming occurs without REM. The essential mechanism behind dreaming is not brainstem activation, not REM, but rather basal forebrain arousal - "wishing". It seems Freud may have been right about dreams.
Mark Solms is lecturer in psychology at University College London and in neurosurgery at St Bartholomew's and Royal London School of Medicine.