Play's the thing (From New Scientist magazine, 09 June 2001.)
Annabel Baker
Kids need the playground just as much as the classroom. Having fun builds bigger, better brains, says Bryant Furlow
PLAYING is a serious business.Children engrossed in a make-believe world, fox cubs play fighting, or kittens teasing a ball of string aren't just having fun. Play may look like a carefree and exuberant way to pass the time before the hard work of adulthood comes along, but there's much more to it than that.
For a start, play can be dangerous, and even costs some animals their lives. For example, 80 per cent of deaths among juvenile fur seals occur because playing pups fail to spot predators approaching. It is also extremely expensive in terms of energy. Playful young animals use around 2 or 3 per cent of their energy cavorting, and in children that figure can be closer to 15 per cent. "For evolutionary biologists, even 2 or 3 per cent is huge," says John Byers from the University of Idaho. "You just don't find animals wasting energy like that," he adds. There must be a reason for this dangerous and expensive activity.
But if play is not simply a developmental hiccup, as biologists once thought, why did it evolve? There are scores of theories, but none is totally convincing. The latest idea is perhaps the most audacious--it suggests that play has evolved to build big brains. In other words, playing makes you intelligent.
Playfulness is quite a rare trait. It is common only among the mammals, although a few of the larger-brained birds such as magpies and crows also indulge. Animals at play often use unique signals--tail-wagging in dogs, for example--to indicate that activity superficially resembling adult behaviour is not really in earnest.
One of the most popular explanations of play is that it helps juveniles develop the skills they will need to hunt, mate and socialise as adults. Another is that it allows young animals to get in shape for adult life by improving their respiratory endurance. Both these ideas have been questioned in recent years.
Take the exercise theory. If play evolved to build muscle or as a kind of endurance training, then you would expect to see permanent benefits. But Byers points out that the benefits of increased exercise disappear rapidly after training stops, so any improvement in endurance resulting from juvenile play would be lost by adulthood. "If the function of play was to get into shape," says Byers, "I would expect the age distribution of play to vary widely." The optimum time for playing would depend on when it was most advantageous for the young of a particular species to get in shape. But it doesn't work like that. Across species, play tends to peak about halfway through the suckling stage and then decline to a low at weaning.
Then there's the skills-training hypothesis. At first glance, playing animals do appear to be practising the complex manoeuvres they will need in adulthood. But a closer inspection reveals this interpretation as too simplistic. In one study, behavioural ecologist Tim Caro from the University of California, Davis, looked at the predatory play of kittens and their predatory behaviour when they reached adulthood. He found that the way the cats played had no significant effect on their hunting prowess in later life.
In another study, neuroscientist Sergio Pellis of the University of Lethbridge in Alberta, Canada, scrutinised videos of rodents play fighting--the most common form of social play in rodents. Despite superficial similarities between this and the social, sexual and fighting behaviour of adult animals, Pellis's close examination of the play bouts revealed no compelling link between play manoeuvres and adult tactics. "For rats, and probably other rodents," says Pellis, "the primary function of play fighting does not appear to be to provide practice for either sex or aggression."
So what is going on? Prompted by the observation that play seems confined to the most intelligent animals, Byers looked at the behaviour and brain size of various marsupials. He found that playful species such as the wombat have bigger brains for their body size compared with their lazier kin, which include the docile koala. More recently, Pellis has teamed up with Andrew Iwaniuk of Monash University in Melbourne to show that in primates, the amount the brain grows between birth and maturity reflects the amount of play in which each species engages.
And earlier this year, Pellis, Iwaniuk and biologist John Nelson, also of Monash University, reported that there is a strong positive link between brain size and playfulness for mammals in general. It is the most extensive quantitative comparative study of juvenile play ever published. Comparing measurements for 15 orders of mammals--from canids to dolphins, rodents to marsupials--the team found larger brains (for a given body size) are linked to greater levels of play. Likewise, animals with relatively small brains tend to play less.
Byers believes that because large brains are less hard-wired and more sensitive to developmental stimuli than smaller brains, they require more play to help mould them for adulthood. Evolutionary neurobiologist Robert Barton of the University of Durham agrees. "I suspect it's to do with learning, and probably specifically with the importance of environmental input to the neocortex and cerebellum during development," he says.
According to Byers, the timing of the playful stage in young animals provides an important clue to what's going on. If you plot the amount of time a juvenile spends playing each day over the course of its development, you end up with an inverted-U-shaped curve. This is the classic signature of a "sensitive period"--a brief developmental window during which the brain can be modified in ways that are not easily replicated earlier or later in life. Think of the relative ease with which young children--but not infants or adults--absorb language.
Byers suspected that these play curves might coincide with a particular phase of brain development known as terminal synaptogenesis. "In many parts of the brain, there is an overproduction of synapses [the connections between neighbouring neurons] and then a specific culling," he says. "Synapses that are active are retained, while the ones that are less active end up being destroyed."
To test this idea, Byers teamed up with biologist Curt Walker from Dixie State College in St George, Utah, to see how the distribution of play with age in cats, rats and mice fitted with the development of a part of the brain called the cerebellum. Among other things, the cerebellum controls the fine motor skills needed for eye tracking, stalking, pouncing and fleeing--the adult activities that most closely resemble the play of kittens and rodent pups.
The researchers found that in all three species play was at its most intense just as terminal synaptogenesis in the cerebellum reached its peak. Evolutionary anthropologist Kerrie Lewis from University College London points out that since new brain cells are seldom produced after birth, synaptogenesis is the most likely way in which play could sculpt the developing brain.
But there are other possible mechanisms. "It might also include things that influence processing efficiency, like myelination," Lewis says. Myelin is a fatty sheath that insulates the tentacle-like axons of nerve cells, improving their ability to conduct electrical signals. Either way, play shapes the overall architecture of the brain rather than individual circuits connected with specific activities. "Most likely, [animals at play] are directing their own brain assembly," says Byers.
"People have not paid enough attention to the amount of the brain activated by play," says Marc Bekoff from the University of Colorado. Bekoff studied coyote pups at play and found that their behaviour was markedly more variable and unpredictable than that of adults. Behaving this way activates many different parts of the brain, he reasons. Bekoff likens it to a behavioural kaleidoscope, with animals at play jumping rapidly from one activity to another. "They use behaviour from a lot of different contexts--predation, aggression, reproduction," he says. "Their developing brain is getting all sorts of stimulation."
Not only is more of the brain involved in play than was suspected, but it also seems to activate higher cognitive processes. "There's enormous cognitive involvement in play," says Bekoff. He points out that play often involves complex assessments of playmates, ideas of reciprocity and the use of specialised signals and rules. He believes that play creates a brain that has greater behavioural flexibility and improved potential for learning later in life. "It's about more connectedness throughout the brain," he says.
The idea is backed up by the work of neuropsychologist Stephen Siviy of Gettysburg College in Pennsylvania. Siviy studied how bouts of play affect the brain's levels of a protein called c-FOS--a substance associated with the stimulation and growth of nerve cells. He was surprised by the extent of the activation. "Play just lights everything up," he says. He speculates that by allowing connections between brain areas that might not normally be connected, play may be enhancing creativity.
All these findings paint a picture of how play might have originated. The comparative study reported earlier this year by Pellis and his colleagues suggests a "stepwise" relationship between increasing brain volume and the evolution of play. The researchers suggest that minor changes in brain size might not have required evolutionary changes in play behaviour, but at certain threshold increases in volume, greater levels of playfulness evolved.
Lewis's recent findings point to the intriguing possibility that different types of play may have evolved at different stages in evolutionary history, to allow the development of distinct regions of the brain. She looked at the relative size of the neocortex--which is responsible for social reasoning, among other things--in primate species, and found that the larger the neocortex in each species, the more social play they indulged in. But this relationship did not extend to object or motion-based play. By implication, Lewis believes, social play may help wire up the social brain, while other forms of play do not. "I think it's reasonably safe to assume that different types of play did emerge at different points in time, but possibly with some overlap," she says.
The idea that play has evolved to build big brains certainly has its critics. Like much of behavioural ecology, it rests on a scaffolding of correlations. "The problem with correlations is that they don't consider unknown third variables," cautions Caro. "So maybe brain size and play are both correlated with metabolic rate or some other factor. Certainly, something about being [warm-blooded] seems important for promoting play."
Even some of the researchers whose results seem to support the link between brain building and play are cautious in their assessment of the theory. Siviy believes there is not yet enough evidence to settle the question. But he thinks the timing of play is convincing. "It's an ideal time to do some learning, to make some modifications to brain circuitry," he says.
One of the strengths of the idea is its testability. Magnetic resonance imaging techniques that identify myelin by-products, for example, should be able to show whether play boosts myelination, as Lewis has suggested. What's more, measuring the volume and activity of certain parts of the brain is becoming increasingly easy due to advances in non-invasive imaging.
If the theory is backed by experiment, what would it say about the way many of us in affluent societies raise our children? We already know that rat pups denied the opportunity to play grow smaller neocortices and lose the ability to apply social rules when they do interact with their peers. Bekoff says play is a sign of healthy development. "When play drops out, something is wrong," he says. Children destined to suffer mental illnesses such as schizophrenia as adults, for example, engage in precious little social play early in life. But can a lack of play affect the creativity and learning abilities of normal children?
The answer is that nobody knows. When Byers searched the literature for information on the relationship between childhood play and development in different cultures, he found that no studies have been done. "There's not even any great data on rate of play for any culture across ages," he says. Until such information is available, assessing the importance of play will be slow going. Meanwhile, our ideas about what constitutes a normal childhood are changing fast.
"Kids are discouraged from playing because they've got to go to school," says Bekoff. "They have all these things to do after school that adults think of as play--but Little League isn't play, in many ways." Organised sports are too structured to emulate spontaneous play, and there's often so much pressure involved that after-school activities aren't even fun. With schooling beginning earlier and becoming increasingly exam-oriented, play is likely to get even less of a look-in. "We have basically become a playless society," says Bekoff. Who knows what the result of that will be?
Further reading:
Bryant Furlow is a science writer based in northern California