WHAT COULD BE MORE HUMAN than recreation? A game of chess, a mystery novel, a video game, collecting stamps, playing tennis, a night of drunken debauchery. There are an infinite number of ways that humans have fun, and they all seem very human-specific. Who can imagine raccoons sitting around playing a game of Texas Hold’em? While dolphins are undoubtedly intelligent, I am pretty sure they would not play chess for kicks, even if we could teach them how. So, do animals have fun? Is there an equivalent of recreation in the animal world? Hold that thought for a second. Before we look at animals having fun, we will explore why humans have fun.
At first glance, there does not seem to be any evolutionary advantage to having fun. How could it have ever been advantageous for humans to be so preoccupied with games and play? In fact, during the long and dark pre-agricultural period in which the very existence of our species hung by a thread and every day was a struggle, would playing have been a detriment to survival? In humans, as well as other animals, time spent playing is time not spent looking for food. Play could be a distraction from looking out for predators. Play could cause needless injuries—even traumatic death. Viewed from the lens of species survival, the drive to play seems like it would be a huge disadvantage. But yet, all humans in all cultures like to play.
The universality of play indicates that it is an innate feature of humanity. It had to be present in our ancestors, even while they were struggling during those dark times. Furthermore, playing is not exactly a minor part of our lives. It is a big and important part of who we are. This behavior cannot simply be an evolutionary side effect or a genetic accident. We are not talking about a tiny appendix in our abdomen. Play is a huge part of the human experience, especially among the young, and thus there must be value in it.
Just like organs or tissues, behaviors will be carefully honed over time by natural selection. Behaviors that enhance survival and/or reproduction of individuals will develop through the generations and become part of the innate “nature” of the species. Such as the suckling of all infant mammals, many of our behavioral urges are driven purely by instinct. On the other hand, behaviors that detract from survival or reproduction will quickly be eliminated through the occasional death of those inclined to perform them, provided there is some genetic influence on the behavior. Further, the more time that an animal spends engaging in a behavior, the more certain we are that it plays an important role in survival or reproduction.
Similar to seemingly useless anatomy, one could reason that some occasional behaviors are mere flukes. Maybe they are carryovers from a previous environment in which such behavior, or some earlier form of it, was necessary. After all, natural selection does not work overnight. Just because a behavioral drive is no longer needed does not mean that it will disappear from a species instantly. It takes time and negative selection. Individuals have to die and take these no-longer-helpful behaviors with them in order for the species to evolve away from them. This cannot be the case with play, even in humans. All humans in all cultures have a drive to play, and we spend a lot of time doing it. With this in mind, it is simply not possible that recreation is not somehow important for human health, survival, or reproduction. But how?
Now that we have properly framed the question, we can go back to animals as a starting point. Do animals play? Right off the bat, any dog owner will tell us that of course they do. They play fetch, they do tricks, they play tug-of-war. Dogs chase each other, they wrestle, they roll on the ground together. Some dogs like to swim; some enjoy jumping into the water from a boat or a dock. So the answer is yes, dogs are playful animals.
Actually, we cannot let ourselves off that easy. Dogs may be a special case because, through their unique breeding, they have a strong, innate desire to work and please their human companions.1 Stronger in some breeds than others, this is a well-documented feature of dogs that is unique to them. It is a result of thousands of years of their evolving alongside humans, as well as selective breeding by those same humans. Dogs really are programmed to “enjoy” working with and for humans, and the only required reward is praise and acceptance. Yes, dogs will also do tricks and perform tasks for food or privileges, but with the proper training, the natural desire for dogs to be accepted and praised by their masters is enough to entice many to perform virtually any task or behavior. In fact, one study has shown that dogs are physiologically “aroused” when they perform trained tasks.2 The point here is that we cannot take the easy road and say that animals know what “fun” is just because dogs seem to play with their human companions. We have to dig a little deeper.
What about when dogs play with each other? Undoubtedly, dogs are naturally given to wrestling with and chasing each other. However, rather than purely for fun, these behaviors could be attempts to establish dominance within a social context and a means to establish trust and familiarity. This does not necessarily mean that these behaviors are not also fun for the dogs, but it does mean that we cannot claim that they are, per se, evidence that animals play just for the fun of it, like humans do. That is the kind of play that we are talking about—just plain old having fun. That is what we humans do. We just have fun. For humans, play is for play’s sake.
Or is it? Maybe our definition of play is the sticking point. If we think of play as doing something enjoyable that has no other purpose, then of course we will not find another purpose—we have made not having a purpose part of the definition. Instead, we need to keep an open mind about hidden purposes of play. We play because it is fun, but it may also be serving other purposes. After all, if we discover biological benefits to playing, then it really is not “just for the fun of it.” It only seems that way. Perhaps the example of dogs playing together really is a clue to the function of play for other species, including us. Perhaps the secret to understanding the function of play is the realization that playful acts can be fun and serve some other purpose for the species.
DEFINING PLAY
Before going further, we must supply a definition of play, which is more difficult than one might think. Any attempt at a simple definition fails rather quickly. One problem is that any subjective features are not helpful. Words like “fun” and “enjoy” are hard to define in other species and could also be applied to things that are clearly not play. For example, eating, sex, and even scratching an itch are things that we enjoy and often have fun while doing, but they are very different behavioral phenomena than things like low-stakes games, sports, make-believe, and so on. We need a definition that is more restrictive than simply things we enjoy or things we do that are fun. Another problem is that a definition along the lines of “serving no purpose” dooms our whole discussion because we are looking for what the benefits of play might be.
In the comprehensive and erudite book The Genesis of Animal Play, reptile behaviorist Gordon Burghardt spends well over sixty pages discussing the problems of defining what play is.3 He demolishes most prior attempts to define play by pointing out how the definitions are either overly broad or overly restrictive. Instead, he offers a framework of the five principle features of play and argues that all five standards must be met, at least minimally, in order to label a behavior as play. The five features are:
1. The performance of [play] behavior is not fully functional in the form or context in which it is expressed; that is, it includes elements, or is directed toward stimuli, that do not [directly] contribute to current survival.
2. The [play] behavior is spontaneous, voluntary, intentional, pleasurable, rewarding, reinforcing, or autotelic (“done for its own sake”).
3. [Play] differs from the “serious” performance of ethotypic behavior structurally or temporally in at least one respect: it is incomplete (generally through inhibited or dropped final elements), exaggerated, awkward, or precocious; or it involves behavior patterns with modified form, sequencing, or targeting.
4. The behavior is performed repeatedly in a similar, but not rigidly stereotyped, form during at least a portion of the animal’s ontogeny.
5. The behavior is initiated when an animal is adequately fed, healthy, and free from stress (e.g., predator threat, harsh microclimate, social instability) or intense competing systems (e.g., feeding, mating, predator avoidance). In other words, the animal is in a “relaxed field.”4
As you can see, defining play is serious business for the scientists that study it. The complexity of this definition reflects the complexity of play forms found in humans and other animals. Notice how many times the word “or” appears. This is to ensure that our criteria for play capture the rich diversity of play found throughout the animal kingdom. However, the requirement that all five conditions are met, in at least one aspect, ensures that nonplay behaviors are not counted as play.
Burghardt’s list is a good example of how scientists attempt to achieve objectivity while studying something that is inherently subjective. The study of animal behavior constantly struggles to develop and apply objective measures for their work, which tend to change over time (hopefully for the better), and the literature on animal play reveals constantly evolving standards and approaches.
Lucky for us, we can get away with a deeply subjective definition of play for the purpose of our discussion. “You know it when you see it” will work just fine. While this is hardly scientific, it is good enough for this chapter because I will only be discussing examples and modes of play that meet the listed criteria and are accepted as play by scientists. If you hunger for the more objective and rigorous analysis of animal play, I urge you to read Professor Burghardt’s book. It is a dense text written for specialists, but at seven-hundred-plus pages, it is the most historically and scientifically comprehensive discussion of animal play out there.
ENJOYMENT DRIVES PLAY
Now that we have covered the scientific definition of play, I will discuss play as we know it: that which is fun. To be a little careful, we can add, “that which is fun but not something else like eating or sex.” The point is that the essential element is that we enjoy play, which is what makes it play. But “fun” and “play” are not synonyms. Play is the behavior itself. Fun is the feeling we get when we do it. In all her creativity, nature has joined the two in order to drive us to do the behavior, just like sex.
Nature’s way of nudging animals to do certain things is to make them enjoy those things. Enjoyment is the feeling of pleasure in our brains caused by the release of neurotransmitters, which is triggered by satisfying the urge to do some behavior.5 All animals have a drive to eat and to drink; it is called hunger and thirst. As we will discuss later, we also have a sex drive. Our brains drive us toward these behaviors through the feelings of pleasure that we feel when we do them. Pleasure is the reward for performing some behavior that is important for survival or reproduction. The fact that we enjoy playing tells us that we have been hardwired to do it.
What are pleasure, fun, and happiness? Again, this could take us down a difficult road, but it is sufficient for us to say that neuroscientists have known for some time that there are chemicals called neurotransmitters that are released in the brain when we experience the feelings of joy and happiness. Many hormones have been implicated in pleasure and the mood we know as happiness, including dopamine, serotonin, norepinephrine, prolactin, oxytocin, and endogenous opioids.6 Different fun or enjoyable activities will cause the release of different combinations of these and other neurotransmitters, and it is a complicated mix.
Through the genius of evolution, these neurotransmitter-driven feelings have evolved as reward systems that underpin our natural drives, urges, and cravings. How exactly the “feeling” of pleasure really came about is something of a mystery, but one thing that is very clear is that the reward system is tremendously similar in humans and our animal relatives. We know this because of animal models of reward and addiction and also because of our ability to measure the release of these neurotransmitters in real time, both in humans and in animals. From this, we know that these pleasure-reward systems can powerfully affect us and drive us toward certain behaviors. We play because we enjoy the neurotransmitter-induced feeling that it gives us when we do. The more interesting question is, why have mammals evolved to enjoy these play behaviors in the first place? Why does nature drive us to play?
PLAY FOR ESTABLISHING SOCIAL RANKING
We mentioned playful dogs earlier, so I will start there. In the case of dogs, wrestling and chasing and rolling around on the ground is how they establish dominance within the social hierarchy in their packs.7 Remember that dogs are descended from wolves and thus retain much of their pack-based social interactions. The survival advantage of this kind of play now becomes clear. Wrestling and chasing are much safer ways to establish dominance than outright fighting. If the dogs were to actually fight in order to establish dominance, the loser could end up dead, and even the winner would likely sustain some injuries. By playing instead, both dogs benefit.
Indeed, some mammal and bird species employ brutal fighting in order to establish dominance, and they suffer losses when they do. Worse, this has led to the bizarre and costly adaptation of antlers and horns in some species, used primarily for fighting among themselves.
How do dogs know when they are playing versus when they are really fighting? As much as it may look like fighting, dogs actually communicate very clearly to each other when they are playing.8 Play begins with an invitation called a play bow. We have all seen a dog do this: they leave their hind legs straight, and with their backside high in the air, they lower their heads and forelegs. This is an invitation to play or, more forcefully, a warning that “play is about to commence; here I come!”
Along with the bow, the tail is held high in the air and wags back and forth before and during the wrestling. This is a way to signal nonthreatening social interaction. You can read a lot about a dog’s emotional state from her tail. High and wagging is a sign of pleasure and play, while a tail held low implies a dog that is scared, threatened, or on edge. Because tails are how dogs and wolves communicate with each other, we can interpret the signs as well.9
Another way that you can tell the difference between playing and fighting is that the dogs do not typically bite very hard during play. Dogs have powerful jaws and can produce a deep puncture wound with just one quick bite. And yet, playing dogs rarely get hurt. This is where the communication and the rules come in.
With this in mind, it seems entirely plausible that, in dogs and wolves, play is a safe alternative to fighting for working out social aggression, and that is the purpose we are looking for! Dogs are programmed to enjoy wrestling with other dogs because that drive helps to safely achieve social harmony. Although humans (except adolescent boys) do not seem to establish social hierarchies through wrestling, this provides a nice model for how to discover why humans play. We need to examine how humans play and then ask what it might accomplish other than simply “having fun.”
PLAY FOR LEARNING SOCIAL RULES
One hypothesis that has emerged is that play is the means by which many social animals learn various rules about which behaviors are acceptable and which are not. This has been proposed most confidently for wolves, which would likely extend to dogs as well.10 The idea is that, through play, animals probe for boundaries. They act out in various ways and indulge their natural impulses, but as they are “corrected” by the social environment, they learn how to behave. This discussion actually delves into the issues of rules, punishment, fairness, and justice. (The notion that the two phenomena of play and fairness could be biologically related is interesting in its own right, and scientist Marc Bekoff has found it impossible to separate the two in his many studies on wolves.)
In this model, among juveniles, playing is not so much about establishing who may be stronger or dominant, although that could happen as well. Instead, playing is about learning how to be social in a social species. As young lions/bears/wolves/zebras play, both with their peers and with adults, their behaviors are refined through coercion, punishment, and reward. Punishment takes the form of ignoring or temporary isolation of some form. An unruly animal will be shunned by the pack or herd, which is felt quite negatively in a social species, particularly by youngsters. This is the equivalent of the dunce cap, time-out, or being sent to one’s room.
Occasionally, adult wolves use mild corporal punishment toward youngsters when an offense is particularly unruly or if urgent correction is needed, such as when they might be hurting another juvenile.11 In other words, they spank. Harsh punishment is important with young animals because injuries from play can occur. However, such injuries are rare because the adults are constantly supervising, ready to intervene as needed. As they grow older, young mammals learn the rules of proper play and begin to monitor each other. Along the way, they become integrated as law-abiding citizens into the social order of the herd.
This type of training is crucial for social species—animals that live in packs or herds with an established hierarchy and interactive behaviors. (Schools of fish are not generally thought to meet this criterion. We are talking mostly about birds and mammals here—the social-cooperative species.) This is why some animals that are rescued from the wild as youths or born in captivity cannot be returned to the wild later. If they are raised without interaction with their own species, they will not be socialized properly according to the rules and norms of their group. This does not necessarily mean that they will be antisocial. It just means that they will have learned by their interactions with their human captors or whatever surrogate raised them, which probably does not prepare them well to be a properly socialized whatever-they-are.
Thus, we can now understand another solid Darwinian selective advantage of a drive to play in mammals. Animals that play will fit better into the social structure of the group. If they do not play, they are at a disadvantage in the hierarchy, which puts them at a disadvantage for reproduction. Further still, playing properly and observing the rules comes with a biological benefit as well, because if you do not do so, you risk being booted from the group altogether. In the ancient lineage of mammals, the collective benefits of herd living were discovered. Through two hundred million years of evolution, group living led to division of labor, sociality, communicative interactions, and dominance hierarchies. It is quite possible that play was very much at the center of all of this sociality as the means through which young mammals were introduced to each other, to the pack, and to the rules and etiquette of social living.
The socialization of young mammals sounds pretty darn similar to the socialization of human children, does it not? As children, we play all the time, we act out, we are constantly seeking entertainment and input, we are drawn to other children and seek to play with them. As adults, we are constantly “correcting” and honing our children’s behavior. Punishments usually involve removal from the social unit—being sent to one’s room, being pulled from the game or the playground, and so on. Even in school, kids are sent to stand in the corner or out in the hallway—they are temporarily shunned from the pack! This is how children learn how to behave and how not to behave. It is really not that different from what wolves do.
We also insist on our children playing with other children, especially if they do not have siblings. We call it “socializing,” and we are eager to set up various playdates. What we are really doing is subjecting our children to social behavioral training. Common sense tells us—and child behaviorists agree—that it is good for our children’s social development to play with other kids, to join the soccer team, even to go away to summer camp. And what happens in these various recreational activities? Children learn to work together with others, follow the rules, recognize and obey authority, and accept punishment for infractions. They also learn the value of the division of labor and cooperative task accomplishment. All while having fun. Think about the first day of summer camp or the first day of soccer practice for very young kids. What is the first thing that is covered? The rules!
Importantly, the lessons learned in Cub Scouts, band camp, and football practice are generalizable to the rest of life. This is the way that our species enjoins the natural instinct to play to the learning of important life lessons. The lessons include how to recognize the pack leader, how to make your own attempt at becoming the alpha, what the rules are and what the punishments are for breaking them, how to make friends and allies, how to avoid or defeat enemies, how to deal with separation from family and loved ones, and how to become emotionally and materially self-sufficient. If you think about it, summer camp is more like survival camp. The lessons learned could rightly be described as savage.
PLAY AS PRACTICE
The summer camp analogy demonstrates the theory that play is useful for teaching social skills and establishing social hierarchies, but there is an even older theory of play called play as practice. These two theories overlap somewhat, and it is important to remember that they are not mutually exclusive. Also, it is time that we look at some species other than wolves and dogs—we have learned plenty from them already.
How about cats? They are not nearly as social as canids (dogs and their relatives). This is quite obvious to any cat owner. Try punishing a cat by ignoring him, and see how far that gets you. It is more like a reward for him. Housecats do not really have a system where they respect “alphas,” and even if they did, good luck trying to convince them that the alpha is you. It is extremely difficult to use rewards to teach them tricks, and any rules that they will obey are based on their own wishes more than anything else. They will poop in the litter box only because you have simulated the natural environment that they prefer. (Cats evolved as desert animals, and they bury their poop to avoid attracting predators.) The question is, “Do cats play?”
I am suddenly reminded of an adorable YouTube video: a tiny little lion cub chasing and playing with a grasshopper. Anybody with a housecat has seen this same thing. Kittens are notorious players. They will play with any little thing that they can find: scraps of paper, small pieces of plastic, dust bunnies, and, of course their absolute favorite, bugs. They bat them around, sneak up on them, and pounce. Why do they do this? They are “pretending” to hunt. All kittens do it, from Bengal tigers to domestic shorthairs.
The phrase that some scientists use to explain this behavior is “play as practice.”12 The reasoning here is that play behaviors employed by young animals serve as a warm-up to things they will have to do as adults. Cats are predators, by nature and by instinct. They simply know how to do it. However, that does not mean that they will be successful on their first try. Hunting takes patience and skill. As any human who game hunts will tell you, far more hunts end in failure than success, and it is no different for cats. Learning to hunt involves a lot of trial and error, and a good hunter learns something with each failed attempt. In other words, hunting takes practice. Thus, for predatory animals, the playing they do as children prepares them for the hunting they will do as adults.
If you want to see an even cuter example of animals “warming up” for adult life, visit YouTube and search for “ducks” and “water slide.” You will find many videos taken from state fairs around the country in which water slides have been erected so that baby ducklings can play on them. These ducks will quickly run up the ramp to reach the top of the slide and then thrust themselves, slipping and squirming, down the slide, plunging into the pool of water below. Once they have done this, they will swim quickly to the side of the pool and repeat the process. They will do this over and over for hours on end until they collapse from exhaustion.
You do not need to be an ornithologist to see that these young birds are having a blast. They enjoy playing in the water and navigating the water slide. And why? I propose that their duck brains are wired to enjoy various forms of water play, as a means to build swimming skills in a nonthreatening environment. If there were actual danger, the ducklings would be stressed and would not play. Since they sense that it is a safe setting, they are free to simply enjoy the water slide.
I urge you to watch the videos. The ducklings display qualities reminiscent of human children on a water slide—namely, they experience an instant of trepidation when they are at the top of the slide. For a moment, they are not sure if they want to slide down. They look a little scared, but they still want to go down. After a second or two, they summon the courage and take the plunge, inching themselves slowly out on the slide. Anyone who has taken a young child to a water park will immediately recognize this nervous exhilaration.
It turns out that many primate species also demonstrate play as practice, but they do not play-hunt like cats do. This is presumably because most primates are not predator-hunters; they eat mostly plants and bugs. However, most primates are communal animals, living in a complex social hierarchy that must be maintained with behavioral patterns that could almost be described as ritual.
Accordingly, the play that primate youngsters engage in is practice for the behaviors that are used in adult interactions, including mating, child-rearing, and dispute resolution.13 It has been reported that Vervet monkeys even engage in “play mothering.”14 Juvenile monkeys will engage in mother-like care of infant monkeys in a sort of stylized play. This is not a simple case of contributing to the good of the troop through babysitting, nor is it equivalent to an older sibling helping Mom with the younger kids. The play that these young monkeys do does not help anyone and probably even subjects the infants to some risk. It certainly does not help the mother, who must watch vigilantly. She is not freed up to hunt or even rest. The play-mothering seems to benefit only the playing child, satisfying his or her need to play-act adult life.
The play of human children is not so different from what our primate cousins do. When we were little, we played house, we had tea parties, we pretended to drive trucks or be police officers. The naughty ones among us played—ahem—doctor. I suspect the play-mothering of Vervet monkeys described in the previous paragraph was familiar to some parents.
Toy makers know this all too well. A quick stroll down the toy aisle at your local department store will yield all kinds of big, colorful, exaggerated plastic versions of implements that adults use in their jobs. From Easy-Bake ovens to motorized miniature cars to Johnny’s first tool set, mimicry of adult behaviors is a pretty big part of children’s play. What looks and feels like playing to us is more than that—it is pretending to be a grown-up and doing grown-up things. What else could this be called except play as practice? How is this any different from the juvenile Vervet monkey pretending to mother a baby or the lion cub hunting the grasshopper?
Is it really so hard to believe that the desire to play-act adult activities is a hardwired genetic drive in human children? It certainly is in other animal species. No one teaches wolves, dogs, cats, or monkeys how to play. They just do it. In the United States, most domestic cats are separated from their mothers and siblings long before they can learn anything from them, but yet, they are playful just the same. It is a natural drive in them, so why is it not a natural drive in us? This makes the evolutionary advantage of play more clear. It is plausible that the playing that we do as children helps to prepare us to be adults—or, at least, that it once did.
Obviously, the reason that children want to play all the time is because it is fun. But why is playing house fun? If you think about it as an adult, it seems like a silly thing to do, but children find it fun. One explanation is that, as children, we are hardwired to enjoy pretending to be adults. The biological benefit of this becomes clearer if we assume that this playfulness actually makes us better at doing adult things. With humans, this is a tricky question. The jury is still out regarding whether or not all that “house” that I played has actually made me better at doing it in real life. But with animals, play-hunting is the first step on the road toward the hunting of real prey.
For full disclosure, the play-as-practice hypothesis has fallen out of favor with some scientists for lack of hard evidence that it truly makes animals better at performing tasks later in life. The reports are conflicting. The biggest problem with attempting to study this in animals is the question of how to deprive them of play without also depriving them of social interactions and other things that may also be important. It is difficult to isolate play behaviors from other locomotor activities and social interactions.
In my opinion, that is the point: play promotes those things by linking them up with the enjoyment-feedback response. It is not so much that the Vervet monkeys truly learn the skills of parenting while engaging in play-mothering any more than human children do. Instead, they gain an outlet to their instincts for parenting and nurture their social interactions at the same time. Nevertheless, some play experts are more skeptical, and that is understandable, given how intertwined all of these behaviors are.
Play as practice is just one working explanation that scientists have developed to explain how and why animals play. There are others as well. The fact that human adults still play argues that there must be additional benefits for play besides play as practice. However, it is true that adults definitely do not play as much as children do, and they play differently. For example, adult play is less about make-believe and more about competition. Sports, chess, poker, and even board games are competitive and skill based. It would be inaccurate to say that adult play has nothing to do with fantasy—just much less so. The fact that the appetite for play wanes during the transition from childhood to adulthood—and has less to do with make-believe—does fit the play-as-practice model very well. Nevertheless, what other explanations for play are there that might apply better to grown-ups?
This is a good time to explain that, when science searches for answers to the mysteries of nature, multiple possible explanations often emerge to explain a natural phenomenon. These differing explanations do not necessarily compete with each other. If there are two possible explanations, both with experimental support, they are not necessarily vying to be the one true explanation. In the case of our study of human play, there might be multiple overlapping explanations, each for different kinds of play or in different contexts and species. The common scientific way to say this is that the multiple theories explaining animal play are not mutually exclusive. They each might be correct for certain play in certain animals. For example, the theory of play as practice works well to explain why human children like to play house, but it may not do as well in explaining why older humans like to play mental games such as chess or physical games such as basketball. There might be other explanations for different kinds of play, and that does not weaken the play-as-practice theory.
PLAY TO HELP ESTABLISH MOTOR COORDINATION
Children and adults both engage in physical play. It is possible that the key to understanding the phenomenon of physical play in humans and other animals is the notion that physical activity promotes brain development, particularly in the areas responsible for skeletal-muscle movements and coordination. In fact, this is a long-known phenomenon and is a key part of childhood development, especially in humans, who are born much less developed than infants of many other species.
Just about the only physical things that babies can do are cry, suck, grasp, and make faces. Incidentally, these are all pretty important for the survival of the babies. Crying is how they communicate their needs, suckling is how they draw milk from the breast or bottle, and making faces is an important part of nonverbal communication. Anyway, besides these few things, babies are born pretty helpless, and their ability to execute complex physical tasks comes slowly over time.
How do humans grow from helpless, uncoordinated babies into suave and elegant ballet dancers? Through practice, of course. As we learn to perform a physical task—whether shooting a basketball or learning a new piece on the piano—we perfect those movements through repetition, and our brains begin to execute them without much conscious effort at all. This is sometimes called muscle memory, but it all happens in the brain. The effect is most dramatically seen when you watch how gracefully and effortlessly a trained basketball player shoots the ball. Compare this with someone who, despite other skills he may have, has never shot a basketball. He will have no form or grace and will clumsily toss the ball in the general direction of the hoop, almost certainly accomplishing the most hilarious of basketball moments—the air ball.
A phenomenon similar to athletic training is at work when infants begin to make their first movements. The connections in the brain that coordinate muscle movements have not been fully developed, and they will not properly develop without many of attempts. Babies’ early efforts at controlled movements are rather pitiable—arms flailing about, legs kicking pointlessly, heads flopping around as if they were only loosely connected to the body. Try this with a newborn one time—if she is facing forward and you are off to one side, clap your hands and draw her attention. Watch how long it takes her to accomplish the simple task of turning her head toward you. It would be a really sad affair if it were not also so cute.
But babies do get better. They improve in all their various movements over time as the brain learns from each effort. The motor impulses from the brain to the muscles are fine-tuned using the sensory information going in the reverse direction. This part is unconscious, but it is part of the incredible beauty and complexity of the brain. As you flex your biceps in order to make some intentional movement of your arm, the sensory nerves up and down your arm are feeding information back to your brain, in real time, regarding how the movement is proceeding. Unconsciously, your nervous system monitors all your movements. That is how we gradually get better at physical tasks. This input-output training is essential for the brain to learn how to send the right commands to our muscles that will result in smooth, coordinated movements.
What does this have to do with play? The drive that kids have toward physical play might be nature’s way of driving us to be as physically active as possible. The more physically active we are, the better we get at controlling and coordinating our movements. Much of this learning takes place in the cerebellum, a structure that looks like a large walnut located in the very back and bottom of the brain case. Movements actually begin elsewhere, in a region called the motor cortex right on the top of our brains, but they are refined and coordinated in the cerebellum, and there is no other way to achieve this refinement than through lots and lots of repetition.
When toddlers first become able to pull themselves up and stand, they usually spend a lot of time bouncing and dancing, especially when they hear music. This continues pretty much all through childhood—kids are in constant motion. It seems like they never just sit still, as any parent or kindergarten teacher will tell you. This is probably not just a weird quirk. It is likely that we are programmed to incessantly move about in order to ensure the constant refinement of our physical coordination. More movement equals better movement; a physically active childhood yields a more agile and well-balanced adult.
It is well known that the brains of youngsters are much more plastic than those of adults and also more susceptible to the input-output training that develops key brain areas. The acquisition of language is a good example of the plasticity of the young brain versus the inflexibility of the older brain. Childhood physical playfulness may well be nature’s way of ensuring that key parts of the brain are properly exercised and developed while the brain is still growing.
If this hypothesis of the benefit of play is true for humans, it should be true for other animals as well. For this, I turn back to the example of cats. Everyone knows how agile cats are. They can jump and scurry and sneak with extraordinary precision and grace—and do so silently. It is truly impressive and awe-inspiring to behold the dexterity of cats.
Anyone who has raised a kitten, however, will know that cats are not born this way. I will never forget watching my first kitten, Sofia, as she was exploring her new home. She would walk around slowly and deliberately. When she would jump up on the coffee table, she did so inelegantly and almost never landed on her feet. “Agility” was not the word that came to mind for describing her clunky movements. I remember very clearly one such event when she struggled to hurl herself up on the coffee table. After a rough landing, she picked herself back up and walked awkwardly along the edge of the table for a few steps before tripping over her own feet and tumbling headfirst back to the ground.
It was a pathetic scene but one that I would never see repeated. Each day of practice served as training for the connections between her muscles, her sensory systems, and her brain. The feedback cycle of inputs and outputs honed these connections until she gradually attained the fine-tuned motor control we all associate with cats. Within a few months, she was jumping from the floor to the kitchen counter, with or without a running start, landing lightly and easily. There is no way that she would have attained that poise and grace had she not worked at it so much as a kitten. Therein is the value of play—training the brain to work with the muscles smoothly and seamlessly.
Physical play is not limited to humans and cats, of course. Mice seem to really enjoy running wheels, and we can find no direct benefit of their doing this. It was previously thought that this was a neurotic or stereotypic behavior, a side effect of captivity, but a recent report showed that mice in the wild were drawn to the running wheels with no reward or incentive and ran on them in sessions that were of similar duration to those of captive mice.15 Why do mice spend their valuable time running in these silly little wheels? For the same reasons that humans will jump on pogo sticks and play with Hula-hoops: It’s fun!
In order to fully appreciate the evolutionary benefit of play as a means to obtaining physical coordination and dexterity, it is important to remember that most species rely on physical agility for their very survival. The same was true for early humans before the invention of agriculture and subsequent civilization. Anatomically modern humans spent about two hundred thousand years living off the land. We had to not only outsmart but also outrun and outfight both our prey and our predators. We were not squarely at the top of the food chain as we are now, and the world was a savage place. The difference between life and death often hinged on how well you moved your arms and your legs. During all that time, the evolutionary drive to play while we were young certainly saved our lives by helping us grow into quick and agile adults.
Two hundred thousand years sounds like a long time, but it is nothing compared with the eighty million years that primates have been evolving or the two hundred million years that mammals have been. All through those eons, physical strength, speed, and agility were vital to the survival of all of our ancestor species and still are for most of our fellow animals. Thus, the drive to acquire impressive physical prowess through playing is ancient and strong. That need and that drive is not limited to childhood, though that is when it is most vital. The persistence of physical play into adulthood is how our ancestors kept their neural connections in top shape. Until recently, couch potatoes would have been eaten by lions.
Even though civilization has recently removed the survival necessity of being in top physical form, the biological drive to play will not necessarily disappear. Why would it? Just because something is no longer necessary or useful does not mean it will simply go away overnight. In fact, a useless behavior might hang around in a species indefinitely unless it is actively selected against. What that means is that unless pointless play starts causing people to die (and thus, not leave offspring), the drive to do it likely will remain. Richard Dawkins once explained it like this: sexual lust derives, at least in part, from the biological drive to reproduce. But that does not mean that lust goes away when we use birth control. Our bodies do not know that safe sex has little chance of producing offspring. Similarly, our bodies do not know that most of us do not really need physical agility in order to survive.
PLAY FOR BUILDING SOCIAL BONDS AND TRUST
A third explanation for the benefit of play is in the development of bonds and trust in social species. Relationships that develop through childhood play will last and help form a closely knit community. If this is true, this explanation for play should function in social animals that live in packs, herds, pods, flocks, or gaggles. Indeed, it has been observed that juvenile primates that play together have been shown to be far less prone to violent conflict as adults, regardless of relatedness.16 Cubs in a pride of lions are not always siblings or half-siblings and yet, as they play together, they form relationships that will last a lifetime.
Is this true for humans as well? Think about your oldest friends. The bonds that we form playing with other children can last a lifetime and often trump conflicts or disagreements we may have. I can think of several friends with whom I “go way back,” and no matter the different turns our lives have taken, the geographic distance, or changes in values or worldviews, I know that we will always be friends and share a special bond for having grown up together. When we spend time together, though many years have passed, we can slip right back into our friendship as we talk for hours about the good ol’ days. In addition, these same friends can get away with saying or doing things that I would never tolerate from a stranger or even a “newer” friend. Our bond can survive political disagreements that would normally infuriate me to the point of ending a friendship—simply because of my attachment to our shared history. After all, we played hide-and-seek together when we were six years old.
This sort of social imprinting has some other effects that are also interesting. While these have little to do with play, I want to mention them because they underscore the point that experiences we have in childhood imprint us in a way that affects our adult relationships and behaviors. I know this is not exactly an earth-shattering revelation, but I am not invoking Freud or any deep psychoanalysis. I am talking about more primal human behaviors that have more explicit connections with those of animals.
One of these primal behaviors is mate choice. It is a well-established principle that many higher mammals and birds will actively avoid mating with individuals that they spent a great deal of time with as juveniles as a means to avoid inbreeding.17 For example, in chimpanzees, our closest relatives, females begin to actively avoid the males that they had previously associated with once they reach sexual maturity.18 Why would natural selection have favored this behavior? Well, for a young female chimpanzee, all or nearly all of the males that she will associate with during her childhood will be close relatives, most of them brothers or half-brothers. When it is time to start mating, her offspring will fare better if their daddies are outside of the family.
Is this true for humans as well? The incest taboo has some biological basis and is often extended not just to blood relatives but to people that we grew up closely with. We all have people in our lives that, all things being equal, we should be attracted to. Maybe people even ask if we have ever been in a relationship with those people. Because we grew up closely with them, our reaction is something like, “Ew, no. She’s like my sister!”
The similarity to the inbreeding avoidance behaviors of our chimpanzee cousins suggests something deeper than a cultural taboo. Studies in humans, most famously the Jewish kibbutz, have attempted to demonstrate what is called the Westermarck effect, named after the famous Finnish sociologist Edvard Westermarck. Often called the first socio-biologist, Westermarck held that humans have a reduced sexual attraction toward individuals with whom they were in close proximity during formative years.
In addition to forming specific bonds, juvenile playing has also been shown to reduce general aggressiveness in rats.19 In this experiment, scientists raised some rats in an environment in which they were sedated during their social interactions, so they could not play with the other rats, and compared this with rats who could play as normal (but also experienced some sedation, to control for possible side effects of sedation). The play-deprived rats grew up to be easily frightened, defensive, and agitated adults. However, the researchers found that if they provided just one hour of play per day, rats would grow up to be properly socialized with other rats as adults. They also grow up to be smarter, but more on that later.
Playing as a means to establish bonds and reduce aggression is not just for youngsters, either. It is well known that both playing and grooming are used by chimpanzees and other apes of all ages to establish social cohesion and bonding. In fact, a population of chimpanzees in a French zoo is known to engage in pre-feeding rituals that include playing.20 Why? To reduce the tension and competition that comes with mealtime. Both in the wild and in captivity, meals can be a stressful and competitive event for chimpanzee troops, and fights over food often break out. Thus, playing right before mealtime seems to be an effective strategy to pacify the troop and allow a more harmonious dining experience. If you think that is weird, bonobos sometimes engage in group sex before mealtime, as discussed in chapter 4.
We humans do the same thing. Well, not exactly the same, but similar. While mealtime is not as conflict-prone for us, humans are still very much a competitive species. Competitions can lead to conflict, and conflict is counterproductive. Since the dawn of civilization, within-group interpersonal competition has frequently reared its counterproductive head in the workplace. To mitigate this, it seems that employers have taken a cue from those French chimpanzees. Anyone who has worked in a corporate or academic setting can tell you about periodic “retreats,” team-building exercises, and the like. What do the employees do at these events? They play together. The stated purpose of these activities is to promote social bonding and cooperation, to reduce conflict and tension, and to facilitate teamwork and personal relationships. Just like chimpanzees.
Playful competitions for reducing tension and aggression can even be scaled up to the global level. I am talking about the Olympic Games, of course. The Olympian spirit is explicitly oriented toward reducing global conflict through sport. It is up for debate whether or not it actually works, but there is a solid biological basis for how it might. Playing builds bonds of trust, in children and adults, and the mechanism could be that the release of all those joyful neurotransmitters in our pleasure center creates an opportunity for intimacy and understanding.
PLAY FOR MANAGING STRESS
We all recognize that some forms of play and recreation are good for relaxing. We use our favorite pastimes as welcome escapes from the pressures of our daily lives. They relieve our stress, at least for a little while. Is it possible that our perceptions about recreation relieving stress are not just superficial? Could it be more than just a psychological effect? Is there a medical benefit to getting some quality playtime now and then? Is this medical benefit strong enough to have resulted in survival advantage? If so, there is a good case that we have evolved to enjoy recreation at least in part because of its therapeutic value.
When it comes to relieving stress, not all types of play are the same. For relaxing, we are usually talking about things like reading, watching a funny or interesting movie, gardening, tending to a collection of some sort, and other hobby-type recreations. We all have a few activities that we do purely because they relax us. Most physicians agree that stress is bad for your health and that the periodic relief of stress through relaxation has medical benefit. We all know this intuitively, but there is good science behind it, too.
A great deal of evidence has indeed linked the levels of circulating stress hormones (mostly cortisol, but also epinephrine and aldosterone) to general sickliness. Cortisol is known to directly inhibit cells of the immune system, so it is no surprise that chronic stress leads to colds, the flu, and other minor respiratory and digestive infections. Even minor cuts and scrapes heal more slowly in a stressed person. The general immunosuppressive effect of stress works through many mechanisms, including reduced antibody production, reactivation of latent viral infections, and the general inhibition of white blood cells, which are our primary infection fighters.21 Scientists have actually observed stress hormones restrain white blood cells in a laboratory setting.
Since stress hormones make you sick, and relaxation reduces stress hormones, it stands to reason that putting time aside to enjoy relaxing activities would be good for your health. Science has confirmed the old wives’ tale that working long hours will weaken your immune system and make you more susceptible to seasonal colds and flu.22 Indeed, the sickest I have ever been was when I got mononucleosis at the end of my first year in college. Lest anyone wonder if it was something else that led to my getting the “kissing disease,” in my defense, I had been up late studying for finals every night for a week beforehand. The stress, combined with the lack of sleep, made me a prime host for the Epstein-Barr virus that likely caused the mono. (No matter how I was exposed to it.)
However, not all forms of play and recreation can be described as relaxing. Think about a paintball competition, an intense game of basketball, most video games, a high-stakes poker game, even a round of golf. Far from relieving stress, these activities actually induce a great deal of stress. Are these forms of play actually bad for our health? If so, why would we have evolved to want to do them? These forms of competitive play are probably the most similar to animals that wrestle and joust for fun. It almost looks like fighting, but it is not. It is competitive play. But if it causes stress, why would this have evolved?
It turns out that even stressful forms of play are actually relaxing. The secret to this apparent paradox is in the difference between short- and long-term stress. Short-term stress, sometimes referred to as sympathetic neural stimulation or the fight-or-flight response, involves a different set of hormones than long-term stress. While long-term stress releases cortisol and aldosterone (among others), short-term stress releases norepinephrine and epinephrine (adrenaline). The result is that any inhibition of the immune system during acute stress is only temporary. The body returns to normal very quickly. In addition, your brain and body seem to know the difference between real life-or-death stress and the playful form of stress experienced during a friendly competition. This is why this form of play does not violate the last criterion of play, which requires animals to be in a “relaxed field” and free from (real) stress.
Even more important, following the fight-or-flight stress of a game, our long-term stress hormones are actually reduced.23 It is as if the burst of “safe stress” is opening some sort of stress valve in our bodies, lowering our baseline level of stress over a longer period of time. We even sometimes refer to this as “blowing off steam.” Similar to a cardiac defibrillator, by delivering an enormous depolarizing shock, a normal resting heart rhythm can be reestablished. I know I am not the only one that feels much more relaxed after a long run or a round of golf than I did beforehand. The point here is that both relaxing and intense forms of recreation can reduce the hormones involved in long-term stress and thus both are good for our health.
This health benefit of play is almost certainly not unique to humans. First of all, the effect of stress hormones on the immune system is well documented in all sorts of laboratory animals. Second, in addition to the effects on the immune system, physical play and exercise are well known to promote good cardiovascular health in both humans and animals. Plenty of research shows that getting enough rest and minimizing stress is a recipe for longevity in both laboratory and zoo animals. This is somewhat difficult to dissect. Does the health benefit come from the exercise itself by way of increased physical fitness? Or is the reduction in stress hormones responsible for most of the health benefit?
In a certain sense, it does not matter. Whether it is the cardiovascular improvements or the reduced stress—or both—the point is that it enhances your survival. And if it enhances your survival, natural selection could act to promote these behaviors through a natural drive to play or exercise. I suppose that we may never really know whether we have a natural drive to exercise as adults or if that drive comes purely from conscious health or beauty concerns. The bottom line is that play is good for us—and always has been.
There is one more aspect to the stress theory of play. Some scientists have theorized that, in children, the stressful aspects of competition and intense play help condition us to manage stress later in life. By exposing us to this safe form of stress, “tense play” helps prepare us to deal with real stress and keep our cool while doing it.24 Real-life stress can involve matters of life and death, which was especially true in our prehistoric forebears. If acute stress made someone freeze or panic, that could have meant his demise. Similarly, if chronic stress causes someone to gradually go to pieces and mismanage her life, she is in deep trouble.
In contrast, the most successful among us are those who rise to meet challenges and never shy away from stress. In fact, many high-performing people report that they are at their best under conditions of stress, such as deadlines, contests, and intensity. It is conceivable that this is a skill gained while playing games. By easing into a life of stress through playful “safe stress,” we may condition our bodies to cope with real stress and even thrive under it.
Once again, it is the lab rat that provides strong evidence for this possible benefit of play. While not specifically addressing play, researchers have found that rats that have been periodically stressed when they were young cope much better with stressful situations as adults. When placed in novel situations, the stress-accustomed rats “froze” less, explored more, and were generally less fearful than rats that had not been forced to deal with stress when they were younger. Interestingly, these “stress-primed” rats had a less-pronounced surge of cortisol when they were placed in stressful situations later on.25 Thus, most scientists agree that mild stress can actually be good for us, if engaged safely and with moderation. That is exactly what competitive play does: it gives us a safe outlet for learning to manage stress.
PLAY FOR DEVELOPING CERTAIN COGNITIVE AND CREATIVE SKILLS
Does playing make us smarter? The answer appears to be yes. As I will discuss shortly, playing as children helps develop certain abilities and overall intelligence, while playing as adults helps to “refresh” some skills that may have become exhausted and gives a burst of creativity. In fact, there is a pretty good correlation between the size of the forebrain, where higher cognitive functions are housed, and the tendency toward playfulness among animal species. In other words, the bigger the brain of the animal (relative to body size), the more the animal will play. This begs the question of cause and effect: do big brains make us play, or does play give us big brains? As usual in biology, the answer is likely a heavily nuanced “both.”
Beginning with animals, as I alluded to before, laboratory rats that are denied the opportunity to play suffer deficits in the development of their brains.26 Other studies have shown that rats with certain damage in their prefrontal cortex will not be driven to play as much, and when they do, they do not perform as well in competitive play. In turn, these rats have developmental defects in skills that would normally have been honed through play.27 These are not just a few isolated studies that can be interpreted any number of ways. It turns out that the importance of play to the cognitive and neural development of rats has been studied many times and in many ways.28 The summary of all of this research is that play is a key aspect for the proper development of rats, not just socially but cognitively as well. Playing is required for full and normal development of the brain and overall intelligence in rats.
Similar observations have been made in other animals as well. Play helps animals develop spatial mapping skills—the ability to visualize objects and places and their relative position to other objects and places. It helps them solve problems such as mazes and perform other learned skills. Playful animals will be better at manipulating objects successfully as tools. The list goes on and on. There is every reason to believe that playful behaviors are important for the normal brain development of all animals.29 The best animal in which to demonstrate this principle is the one with the most cognitive development to worry about in the first place: humans.
When it comes to humans, there is a staggering amount of evidence that playing is important for cognitive development. Many books have been written on the subject, and the educational psychology of play is a vibrant academic subdiscipline unto itself. For constraints of space, I will attempt only to briefly summarize the major points of what we know about the value of play in human children. Experts in play generally talk about seven modalities (types) of play, each with its own benefits:
Attunement refers to play between infants and their parents/caregivers. This type of play involves babies just looking and being entertained, but it is crucial for emotional development because at this stage emotional states are almost purely “contagious” for the babies (see chapter 4). When you hold an infant and smile and express joy, the infant often “catches” the joy and shares in it. Similarly, fear and stress can be conferred upon the baby as well. In a sense, young childhood is when we develop our most basic emotions. Neglected children almost always suffer emotional difficulties, and this is part of the reason why.30
Body play is the physical play that we have already discussed for its role in developing coordination and locomotor precision. It also helps with visualizing spatial relationships and with understanding how our body relates to the world around us, which is important for learning to avoid physical dangers and for respecting natural boundaries.
Object play is playing with toys and other physical things, mostly with our hands but also with other body parts. This, too, helps with understanding spatial orientation and visualization in three dimensions. It also develops reflexes and fine motor control.
Social play is that which involves others and is useful for learning social structures, etiquette, and rules.
Imaginative play is among the most intriguing forms of play. It was touched on earlier in the play-as-practice section, but psychologists tell us that it is much more than that. It is believed that imaginative play is key to developing the broad range of cognitive abilities in humans that involve abstract concepts, independent thought, and creativity.31 Children use make-believe to explore and interact with the social roles that they are gradually learning in their environment. These include gender, professions, and various activities that they see around them. Children will vociferously engage in imaginative play with few or no props needed.
Through make-believe, children develop their own imaginations and learn the ability to perceive things that do not actually exist, both figuratively and literally. In the figurative sense, they can imagine scenarios that do not exist, like being a mother, a professional athlete, or a teacher. By literally, I mean that they practice the skill of mental imagery where they close their eyes and conjure an image. When we do that, we perceive something visually, even though we do not actually see it. Did play lead to imagination, or did imagination lead to play?
Narrative play is the process of storytelling, both in reception and delivery. This, too, exercises the imagination, but it also promotes the development of language skills. Language development will be discussed in chapter 10.
And finally, transformational play occurs when we take on a new identity and then solve problems and challenges as that new identity. This often involves a different set of rules and surroundings from our actual environment and thus requires creative thinking and “transcendence” from some of the limitations of our actual world. In a sense, we repeat the process of our earlier “probing” of the limits and rules of our actual world and apply that same process of discovery in an imaginary setting that we do not totally control (as in purely imaginative play). Video games are good examples of this type of play. Sorry, parents, many studies have shown that video games can promote development of certain cognitive skills and even raise children’s intelligence in some contexts.32 Of course, these benefits should be balanced with the benefits from physical and social play in order to shape a well-rounded and well-adjusted child.
The bottom line here is that there are many benefits to the various types of play when it comes to developing the full potential of the human brain. The sensory-motor stimulation, the imagination, the socialization—all of this is good for human development.
What about adults? Is playing good for adults as well? The adult brain is much more hardwired and less able to pick up new cognitive abilities, so it is at least possible that play is not as important for adults. Indeed, the drive to play is much weaker in adults than in children, but it does exist. We adults play, too. In addition to the stress relief and cardiovascular health benefits, are there also cognitive benefits for adults?
Play has been found to be a very effective way to enhance learning in students of all ages.33 Many studies have shown that physical exercise is good for maintaining cognitive function in senior citizens, but what is often overlooked is that the exercises that are most effective for this are games and team sports, which combine body play with social play and object play. One study even showed how the Nintendo Wii improved the well-being of the residents of a retirement home.34 Finally, sessions of play are also associated with bursts of creativity and successful problem solving.
* * *
While researchers are divided as to the origin of play, there is no doubting the biological benefit. Squirrels that play more are better coordinated and better parents.35 Play has been shown to be beneficial for rats, brown bears,36 and wild horses.37 Play has been shown, in humans and animals, to promote intelligence and, specifically in humans, original and creative thinking. Stuart Brown, a physician and one of the world’s leading experts on play, put it this way: “The opposite of play is not work—the opposite of play is depression.”38
Even in the time of Plato, it was known that humans are hardwired for play. This propensity gradually, but not completely, fades with maturity. Because there are multiple benefits of play, it is no surprise that the behavior is widespread in nature. However, as with everything else in life, there is a natural tension between the advantages of play and the disadvantages of overdoing it. Play may be good for a young lion cub, but only up to a point. If she gets so distracted so as to become unaware of her surroundings, a hyena might make an easy meal of her. And while all work and no play may make a bonobo dull, all play and no work will get him cut out of the food-sharing hierarchy or shunned altogether.
These opposing forces of play and more “serious” endeavors have been fighting it out for millions of years. This, too, sounds familiar. In a sort of reenactment of evolution, many of us will spend our adult lives trying to strike the perfect balance between work and play. Many educational psychologists believe that our children would learn better if we employed more play in our schools—playing to learn. Similarly, in today’s very corporate culture, most experts of play agree that we have suppressed fun and games too much among adults. Maybe that is why we have not seen as much of a decline in chronic illnesses as we might expect, given our advances in preventive and curative medicine. In other words, “When we stop playing, we start dying.”39
FURTHER READING
Brown, Stuart L. Play: How It Shapes the Brain, Opens the Imagination, and Invigorates the Soul. New York: Penguin, 2009.
Burghardt, Gordon M. The Genesis of Animal Play: Testing the Limits. Cambridge, Mass.: The MIT Press, 2005.
Byers, John A. Animal Play: Evolutionary, Comparative and Ecological Perspectives. Cambridge: Cambridge University Press, 1998.