Children’s Dreams

Did you know that before birth, humans spend almost all of their time asleep? During the second trimester of pregnancy, foetuses spend about six hours a day in REM sleep, six hours in non-REM sleep, and the rest of the time in a state of sleep somewhere between the two. During the final trimester, the baby finally ‘wakes up’, and spends about two to three hours per day awake. About two weeks before birth, REM sleep suddenly increases to about nine hours per day, and then in the final week before birth this increases again to about twelve hours. This huge increase in REM is crucial for a process called ‘synaptogenesis’, i.e. neurons (brain cells) making synaptic connections with one another, and this helps the infant prepare for birth. Once they are out here in the world, there will be a million new things to learn, and their brains have to be ready for this huge onslaught of confusing information!

Sleep continues to be very different for children than it is for adults post-birth. Newborns require a great deal of sleep and are polyphasic (they sleep on and off throughout the day and night), which is rather unfortunate for their caregiver(s), who are monophasic (sleep only once in 24 hours). This mismatch of sleep schedules makes for very sleep deprived parents. But as infants grow, sleep becomes more and more associated with the night-time, and sleep need decreases with age all the way up until adulthood. 

Just as with sleep, dreaming in children is different to dreaming in adults, and the ways in which it is different depend on the age of the child. Until the 1960s, psychologists tended to assume that children’s dreams were vivid, colourful, and emotional, especially unpleasant emotions e.g. nightmares about being chased. However, this notion was challenged by one of the largest studies into children’s dreams, conducted over several decades by David Foulkes, who was working at the University of Wyoming in the USA. He spent decades researching dream content in children of various ages, eventually publishing his book Children’s Dreaming and the Development of Consciousness in 2002.

The youngest children he researched were between three and five years old. At this age, he found the children’s dreams were very rare, and the dreams they did have were very brief and lacked many of the aspects of dreams that we consider typical in adults dreams, e.g. the self being a character in the dream and emotions. They didn’t even contain things you might expect from young children’s dreams, such as monsters or other scary characters. Dreams were instead often just static images, often of animals, such as a bird singing, a calf standing in a barn, or a chicken eating some corn.

At ages five to nine, the dreams changed. They became longer and contained more social interaction (people interacting with each other), and more movement. They also became more frequent, and the older the child, the longer, more frequent, and more complex the dreams. Animals appeared less often, and the self appeared more often as a dream character.

Between ages nine and fifteen, dreams became more and more like adult dreams. Going into puberty years, dream content became more reflective of the dreamer’s personal style, such as the dreams of particularly assertive children having more active selves in their dreams, and dreams of children who were feeling particularly hostile before sleep having more angry dream characters.

From his research, Foulkes concluded that dreaming develops in predictable stages over the course of our childhood. First, when very young, we tend to dream only brief static images, often of animals, and those dreams are relatively very rare. Then, getting a little older, these static images become kinematic, and then eventually we dream of ourselves going around in a virtual world – much like we do as adults. Foulkes concluded that dreaming, in the sense we know it as adults, is a high-level cognitive process that requires cognitive skills we just don’t have as young children, such as being able to imagine a virtual world, and imagining ourselves within that virtual world. It’s not until we get to about 12 or 13 years old that children’s dreams before more like adult dreams. In this sense, Foulkes’ model of dream development, based on this research, suggests that dreaming is a cognitive skill that develops alongside other cognitive skills, and becomes more complex and closer and closer to adult level as the child ages.

These findings may seem jarring to parents with young children who report detailed dreams that sound much like adult dreams. Indeed, in the decades that have passed since Foulkes’ studies were conducted, his conclusions have been questioned. For example, in a study conducted in 1994 by a team of researchers at Harvard Medical School led by Jody Resnick, children from ages four to ten years old slept at home and told their dreams to their parents. By having parents interview their children in this way, the researchers found that initially shorter reports increased in length enormously, especially in younger children, ages four and five (dream reports increased to up to 250% from initial report using this method). Not only were these dreams reports much longer than those reported by David Foulkes, they were also different in terms of content, too. In this study, the children dreamt of themselves being active in the dream in about 80% of dreams of both older and younger children, and dreamt of many familiar characters from their waking lives, especially their families. They concluded that even young children experience long dreams with themselves and others as characters in a similar way to adults, and that it is crucial therefore to collect dream reports in a supportive home environment to study dreams.

Other research has since supported the idea that young children have complex dreams. Piroska Sándor of Semmelweis University in Budapest, Hungary, and her colleagues have recently conducted their own research into the dreams of young children. They used a method that aimed to take the best parts of home-based research (i.e. the familiar setting and known adults questioning them) and infuse them with as much control as they could. Parents acted as interviewers, but were carefully trained, and the interviews were recorded to allow the researchers to remove any answers that came from suggestive questioning. Using these methods with children from ages 4 to 8, they found starkly different results to Foulkes’. Although they did find that dreaming reflecting development in some ways, such as an increase in dream length and in cognition (e.g. planning or decision-making within the dream), other aspects of dream content did not change with age, including bizarreness, emotionality, kinematic imagery/actions, and human characters. What’s more is that they found that these aspects of dreams were relatively prominent in even the youngest children’s dreams, suggesting that they are more similar to adult dreams than was concluded by Foulkes. The authors suggest again that perhaps there is something about the laboratory setting in Foulkes’ research that influences dream reporting: whether it is the unfamiliar setting, the unfamiliar adult questioning them, being woken up in the night and being too fatigued at that time to report dreams in detail (and this being more of an issue with the younger children), or another reason.

However, partially in response to such research as this, in 2017 Foulkes published a new paper reminding that his research did in fact include home-based studies as well, which replicated his lab-based research. He suggests that the resistance to accepting his conclusions may be because they seem so counter-intuitive; most people would assume, or based on their own experiences of hearing their children’s dreams believe, that complex, story-like, emotional, and bizarre dreaming happens from a young age, but Foulkes’ research indicates that until the age of about seven this just isn’t the case – at least in the most part. That’s not to say that some younger children can’t have these kinds of dreams occasionally, but most of the time they aren’t.

Currently Foulkes is the only researcher to have conducted research both in the lab and at home, and thus make the comparison between the two methods. Until we have more data on children’s dreams both in the lab and outside of it, the question remains unanswered. We can be fairly sure that dreaming does follow a developmental pathway, increasing in complexity and length with age, but whether very young children can have these kind of dreams, and if so how often, is subject to further research.


Foulkes, D. (1999). Children’s dreaming and the development of consciousness. London: Harvard University Press.

Foulkes, D. (2017). Dreaming, reflective consciousness, and feelings in the preschool child. Dreaming, 27(1), 1–13.

Resnick, J., Stickgold, R., Rittenhouse, C. D., & Hobson, J. A. (1994). Self-representation and bizarreness in children′s dream reports collected in the home setting. Consciousness and Cognition,
3(1), 30-45.

Sándor, P., Szakadát, S., & Bódizs, R. (2014). Ontogeny of dreaming: A review of empirical studies. Sleep Medicine Reviews, 18(5), 435–449.

Sándor, P., Szakadát, S., Kertész, K., & Bódizs, R. (2015). Content analysis of 4 to 8 year-old children’s dream reports. Frontiers in Psychology, 6.

%d bloggers like this: