Dream Research
What children dream about, and why very young children dream so little
David Foulkes spent decades waking children in a sleep laboratory. What he found contradicted almost everything assumed about childhood imagination — and remains the field's most important longitudinal study.
The intuition is that children, being imaginative, dream vividly and elaborately. David Foulkes tested it properly, and the results were not what anyone expected.
Between the 1960s and 1990s at the University of Wyoming and later Emory, Foulkes ran longitudinal laboratory studies in which children slept in a sleep laboratory across multiple years, were woken from REM and non-REM sleep, and were asked what had been going through their minds. Interviewers were trained to avoid leading questions — a serious issue with children, who are highly suggestible and will readily produce a story if one seems expected.
What he found
Ages 3 to 5. Dream reports were obtained from REM awakenings only about 20 to 30 per cent of the time — far below adult rates. When obtained, they were brief, static and bland: a single image, frequently an animal, with no narrative, no movement and no self. "A chicken." "I was sleeping in a bathtub." Emotion was largely absent, and the dreamer rarely appeared as a character.
Ages 5 to 7. Recall rates rose. Dreams acquired movement and sequence, and characters began to act. The self appeared as an active participant.
Ages 7 to 9. A marked change. Reports became narrative, with plot, and with the dreamer as protagonist. Emotion appeared. Length increased substantially.
Ages 9 to 13. Reports approached adult characteristics in structure, length and content.
Dream development did not track verbal ability or general intelligence. It tracked visuospatial skill — performance on block design tasks specifically. Foulkes concluded that dreaming is a cognitive achievement that develops alongside the capacity for mental imagery and spatial construction, rather than a primitive faculty present from birth.
Why this was controversial
It contradicted psychoanalytic assumptions that children's dreams are simple wish fulfilments present from infancy. It also contradicted parental reports, which describe young children waking distressed and recounting elaborate dreams.
Foulkes's response to the latter was methodological. Home reports are collected by parents who ask leading questions, and young children confabulate readily and cannot reliably distinguish memory from imagination. Much of what is collected at the breakfast table is co-constructed with the adult asking.
The laboratory data has been partly challenged since — some later studies using different methods have found somewhat higher recall rates in young children — but the developmental trajectory, and the association with visuospatial ability, has broadly held.
Nightmares and night terrors in childhood
Both are common and are frequently confused.
Nightmares peak between roughly three and six years, occur in REM in the second half of the night, and the child wakes fully, is frightened, seeks comfort, and can describe what happened. Reassurance works.
Sleep terrors arise from deep slow-wave sleep in the first third of the night. The child may scream and appear terrified, with eyes open and racing heart, but is not awake and cannot be consoled. In the morning there is no memory of it. They peak between about four and twelve years and are usually outgrown.
The advice for terrors is counter-intuitive: do not wake the child, keep them safe, and wait. Attempting to rouse them typically prolongs the episode. Where they occur at a predictable time, scheduled awakening — briefly rousing the child fifteen to thirty minutes before the usual time, for a week or two — has reasonable evidence.
Both are worsened by sleep deprivation, irregular schedules, fever and stress. Frequent terrors in an older child warrant assessment for sleep apnoea, which fragments deep sleep.
What children actually dream about
Once narrative dreaming establishes, content follows the continuity principle. Animals feature heavily in younger children's dreams and decline through childhood. Family members predominate over strangers. School, friendships and — increasingly with age — social evaluation appear.
Threatening content is common, and there is evidence that children in higher-threat environments report more of it.
Talking to a child about dreams
Two practical points from the research.
Ask open questions. "Did you dream about the monster?" will produce a monster. "What was happening?" will produce whatever there was, which may be nothing.
And take a frightened child's dream seriously as an experience without treating it as a message. Young children have limited grasp of the distinction between dreaming and waking reality — the understanding that dreams are internal and private develops gradually and is not secure until around seven. Telling a four-year-old that it was not real does not necessarily land. Comfort does.
