Chapter published in Comparative Psychology: A Handbook (G. Greenberg and M. M. Haraway, eds.), Garland Publishing Co., New York, 1999, pp. 725-735.

PLAY

Gordon M. Burghardt

Play, playfulness, or play-like behavior has not proven easy to define or study. But identifying it as a "real" phenomenon is another matter. Play has been noted in books on animal behavior and comparative psychology for many years, as in the 19th century writings of Thompson (1851), Darwin (1874), Lindsay (1879), Büchner (1880), and Romanes (1892). For these writers a definition seemed almost unnecessary: play was behavior that was not serious (in the sense of being directed to accomplishing some immediate task), was often highly energetic, most characteristic of healthy individuals, and was probably pleasurable. But then, as now, play was an oddly discontinuous category: prominent in some species (especially mammals) and absent or rare in most (especially invertebrates and poikilothermic vertebrates). Thus play was, and still is, rarely incorporated into any general system of comparative psychology.

Today, nonhuman animal play is classified into three types: locomotor/rotational play (running, leaping, sliding, brachiating), object play (manipulating, pulling, pushing, chewing) and social play (rough and tumble, chasing) (e.g. Fagen, 1981). Although the first two (locomotor and object)are often studied in solitary individuals, there is no reason that they cannot also occur in a social context - as chasing implies. In the child play literature, a normally solitary play activity, such as using building blocks, performed in close proximity to another child is called parallel play and contrasted with cooperative play (Pellegrinni & Boyd, 1993). Still, the tripartite distinction is useful.

 

Early examples and theories of play

Play was initially viewed using the criteria listed in the first paragraph above. Examples of play were reported in crabs, ants, fish, bower birds, dogs, cats, and primates. Currently examples from the first three are excluded from the pantheon of "true" play (e.g. Fagen, 1981) while the latter three are still the focus of extensive play research. To see why let us look at some examples.

In an early book by Thompson (1851), probably the first "encyclopedia" of comparative psychology (64 topics), the entry on "playfulness" first defined play as an "exuberance of animal spirits" in which the animal "abandons itself" to "the performance of some one of its passions, whether of joy or mischief, defiance or fear" (p. 61). Thus crabs "play with little round stones, and empty shells, as cats do with a cork or small ball" (p. 62). Horses, hares, whales and fish gallop, gambol, and leap. Birds, deer, lambs, dogs, beaver, and orangutans play wrestle or chase one another. In this early treatment many classic play examples are already laid out.

The fate of play in ants is instructive. The observations of wrestling and hide and seek by Huber, Forel and other early observers favorably cited by Büchner (1880) and Romanes (1892) are now interpreted as aggressive competition (Wilson, 1971). They are dismissed from the category of play because they are presumed to be serious and functional. This dismissal is accepted by many play researchers (e.g. Fagen, 1981; Hole & Einon, 1984) but may prove premature as the immediate benefits of play become more recognized (Burghardt, 1988).

Darwin (1874) was aware of play but it occupied only a small role in The Descent of Man (Darwin, 1874), not even appearing in the exhaustive index. Typically, however, Darwin gave two prescient examples that went beyond "mere play" (p. 72) in his discussion of complex emotions. One was the presumed sense of humor dogs show in retrieving a stick. A dog may sit down next to the stick after returning it, but when the master comes to pick up the stick, at the last moment the dog grabs the stick and runs away in "triumph", repeating the game, and "enjoying the practical joke" (see Mitchell & Thompson, 1991, for a sophisticated analysis of similar phenomena in dogs). Darwin also noted the complex mixture of fear and attraction shown by monkeys in response to a live snake.

By the end of the 19th century play was viewed as largely restricted to the higher animals (i.e. birds and mammals). Although there were other explanations of human and animal play (see Baldwin, 1902b; Müller-Schwarze, 1978), two theories predominated in discussions of animal play and have cast their shadows, or perhaps their bright insight, down to the present. The controversy centered on these views reflects the confusion that arises when theories do not distinguish among the different levels of analysis and kinds of questions asked in attempts to explain a) the different kinds of play, b) the diversity of play among animals, c) the causal mechanisms of play, and d) the functions of play.

The first of these two theories was the popular and then much maligned surplus energy view popularized by Herbert Spencer (1872) in but a few paragraphs of a massive treatise on all of psychology. This view derived play from excess energy accumulated by the "higher animals" in the course of obtaining better nutrition than "inferior" species during ontogeny. This excess energy had to be expressed in some manner, hence play. This view confused motivational, metabolic, and behavioral "energy." Thus Spencer's formulation, while providing a phylogenetic rationale for observed differences, tended to be seen as supporting a view of play as just the nonfunctional "letting off of steam." Any similarity of juvenile play to adult "serious" behavior was viewed by Spencer as due to some ill-defined process of imitation.

The confusion about energy along with a growing evolutionary adaptationist stance and functionalism in psychology led to the search for an adaptive reason why animals should devote time and effort to play. The most celebrated view was the practice theory of Groos (1898), popularized in America through the efforts of James Mark Baldwin, one of the first psychologists to take evolution and development seriously (Baldwin, 1902a). In this view play was an instinctive ontogenetic process preparing the animal for later life by perfecting the performance of critical behavior patterns. This view could explain why play was most prominent in young animals and occurred in isolated animals not able to imitate others. This practice view, still highly influential, unfortunately helped promulgate another kind of confusion. In this case the confusion was between behavior patterns that have current value versus those with "delayed" benefits. As noted above, this contrast led to arguments that if a current function could be demonstrated, or even convincingly proposed, for performing a playful act, other than practice, then the behavior was by definition not play. There was also uncertainty as to whether play should be viewed as a separate instinct or drive or as an unrefined version of the later serious behavior. This brought in motivational issues. For example, since play seemed "fun" it was often viewed as a distinct motivational category. But then how was it linked to the later serious context? This issue was more of a problem for the Groos position than for Spencer's.

Although it is often claimed that the surplus energy view was untenable after Groos (e.g. Harré & Lamb, 1983), many influential writers continued to see some merit in energy formulations (e.g. McDougall, 1924; Morgan, 1920). But the conceptual problems posed by both theories were largely unrecognized or ignored: confusion of the function, adaptiveness, or delayed benefits of play with the causal mechanisms that lead to play either in an individual's ontogeny or in a species' phylogeny.

Of the other early views on play two will be mentioned that a comprehensive theory of play may also have to address. One was the recuperation theory, which held that play, like sleep, was a welcome, even necessary, respite from the boring or onerous day to day demands of existence. This provides play with a function without mandating specific delayed benefits. G. Stanley Hall (1904), in attacking Groos, championed the recapitulation theory: many play activities in humans, such as ball, stick, and kicking games, were remnants of our hunting and fighting ancestry and not necessarily functional for future activities.

In short, many of the major ideas about play informing current thinking had arisen by around 1900. Each may have explained some important aspect of playfulness better than the others. But since play was viewed as unitary only one theory could be true. The conceptual tools to test the theories adequately were not available, the conception of play itself was often unclear, the levels of discourse were confused, and methods of recording and analyzing behavior were far from adequate. While this situation still persists in some degree today, we now can clarify some of the issues involved.

 

Current approaches to play

Fifty years ago the most prominent comparative psychologist of his day began a review of play with this sentence: "Present-day understanding of animal play is regrettably limited, and current views on the subject are considerably confused" (Beach, 1945). In several pages Beach raises and critiques many ideas about play still debated today. His final sentence presciently presages what is now accepted as the primary context in which play should be understood (e.g. Smith, 1982): "An evolutionary approach has proven fruitful in advancing our knowledge of many phases of human behavior; and it is not too much to hope that the careful study of animal play will offer potentially significant results in the increased understanding of similar behavior in man" (Beach, 1945:540).

Regrettably, today we still have no comprehensive model of play, evolutionary or not (Burghardt, 1984). However, the last decade has seen some developments indicating that play will receive its due as both a crucible and product of ontogenetic and phylogenetic processes. Detailed observations of play along with new developments in comparative and observational methods, physiology, and neuroscience promise advances that may remove the conundrums and embarrassment that the topic of play often elicits.

The only monograph length comparative survey of animal play since Groos (1898) is Fagen (1981). He reviews the myriad definitions of animal play, theoretical issues, and, most usefully, provides a comprehensive survey arranged phylogenetically. Fagen traced the failure of prior research to provide a solid grasp on this elusive topic to the absence of an integrative theory. While he praised the increasing literature of careful and detailed descriptions of play, both field and laboratory, since Beach (1945), Fagen accused ethologists and comparative psychologists of being too provincial to see the merits of mathematical modeling and recent evolutionary theory (i.e. sociobiology). These claims have been critiqued by Burghardt (1984). While the evolutionary approach writ large has had a salutary effect in encouraging an integrative approach, evolutionary modeling in isolation from detailed knowledge of the topography, physiology, and ontogeny of playfulness can only provide the illusion of understanding. Fagen and most commentators on animal play bemoan the lack of data on play in many species. Until we obtain such data even the application of the new generation of comparative phylogenetic methods (Brooks and McLennan, 1991; Harvey and Pagel, 1991) cannot be successfully deployed. In short, we need to proceed on several fronts in studying play.

Play is most in need of a serious application of the ethological research aims that derive from Tinbergen's (1963) influential statement: causation, adaptiveness, ontogeny, and evolution. Recently a fifth aim has been suggested by Burghardt (in press): the experiential aspect of behavior from the participant's point of view, called private experience. This added aim is particularly relevant in play because of the common incorporation of terms such as pleasure, fun, enjoyment, rewarding, and positive affect in many definitions of play. Admittedly, this is a controversial issue even in humans and is difficult, but not impossible, to assess in nonhuman animals.

What is play?

It now seems clear that play is a heterogeneous category of diverse phenomena derived from different sources in spite of sharing a few common themes. Thus it is understandable that no unitary view will encompass all kinds of play: locomotor, object, and social. Still, characterizations of these common themes is frequently attempted in the form of definitions of play (Fagen, 1981).

Probably the definition of animal play that captures as objectively and comprehensively as possible the current position was stated by Martin and Caro (1985:65), who modified an earlier definition by Bekoff and Byers (1981). "Play is all locomotor activity performed postnatally that appears to an observer to have no obvious immediate benefits for the player, in which motor patterns resembling those used in serious functional contexts may be used in modified forms. The motor acts constituting play have some or all of the following structural characteristics: exaggeration of movements, repetition of motor acts, and fragmentation or disordering of sequences of motor acts."

This definition avoids many of the more contentious aspects of play such as motivation, affect, and intentionality (Allen & Bekoff, 1994). It does not, however, include the observation that some of the most drammatic examples of social play include marker signals that may serve to both delimit an interaction as play or serve to punctuate or maintain a play bout (Bekoff, 1975; Rasa, 1984). The bow, which occurs almost exclusively in play in canids, is a highly stereotyped action that shows species differences. Juvenile coyotes (Canis latrans) are more aggressive to one another and engage in dominance fights more than dogs or wolves (Canis lupus). There is suggestive evidence that coyotes need to work harder at maintaining a prey atmosphere than the other two species, since they perform more bows later in play bouts (Bekoff, 1995).

Additionally, most workers today, regardless of their attitude towards play, eschew uncritical anecdotal reports and call for careful descriptive analysis of what is actually being observed (c.f Hole & Einon, 1984; Müller-Schwarze, 1984; Coppinger & Smith, 1989). As the heterogeneous nature of play becomes taken more seriously, it will become more acceptable to see play as a set of transitional behavior patterns linking a variety of behavioral phenomena with diverse qualitative and continuous features. It will also preclude often useless arguments as to whether a given species (e.g. ferrets, Lazar & Bekhorn, 1974) "really" play or show "true" play.

 

Who plays?

Generally accepted evidence for play is confined to birds and especially mammals (Bekoff & Byers, 1981; Fagen, 1981). Every family of mammals shows some evidence of play in some species. There is a rough progression within mammals with primates being overall most playful but there is also much play in aquatic mammals, carnivores (especially otters) and rodents. Phylogenetic factors are less strong than might be expected because there is so much variation in the extent and complexity of play within taxa that can be related to differences in ecology, development, physiology, importance of learning, and the normal behavioral repertoire. Thus predatory species (e.g. cats) will show much object play related to predation while vegetarian prey species (e.g. deer) may show much locomotor play related to speed and agility in flight.

Since Fagen (1981), the literature continues to accumulate descriptions of playful behavior in a wide range of species, primarily mammalian and avian. The behavior of nonavian reptiles is still rarely identified as play (Burghardt, 1988; but see Burghardt et al., in press). Often play studies take place in zoos and the main focus is on environmental or behavioral "enrichment" (e.g. Markowitz, 1982) rather than on play. Here "voluntary" interaction with objects is used as an indicator that an animal is not bored, engaged in pathological stereotyped behavior, or just plain vegetating. But such studies can also elucidate the nature of both stimuli and the context (environmental, social) that facilitate, direct, or modulate the occurrence of play.

 

Comparative studies of play in rodents

 

As an example of a valuable set of studies that show the importance of looking at the details of play behavior rather than just measures of frequency, duration, and context, the work of Sergio and Vivian Pellis and their colleagues on rodents is exemplary (review in Pellis, 1993). Play fighting in young rodents is a major category of social play (Panksepp, et al., 1984). A major main difference from adult fighting is its lack of "seriousness," that is, the lack of injurious bites. A series of studies have revealed that the targets of play fighting are most often the nape of the neck while serious fighting involves targets such as the lower dorsum and top of the head (Pellis and Pellis, 1987). Indeed, the targets in play fighting are those most involved in amicable behavior and this observation holds across a number of species.

Comparing play in a related group of animals that show qualitative and quantitative variation in the kinds and amount of play are essential for testing phylogenetic hypotheses as well as identifying the most useful measures. A careful analysis of play-fighting in muroid rodents by Pellis, et al. (1989) illustrates the value of this approach. Deer mice (Peromyscus maniculatus bairdii), prairie voles (Microtus ochrogaster), and montane voles (M. montanus) were compared to laboratory rats in an attempt to move from subjective statements about species differences in play to quantitative (frame by frame film analysis) measures of play complexity. Both play attacks and resulting play defense and counter-attacks and, importantly, the body parts that were targets differed across species. Basically, all four species had all the subcomponents of attack, defense, and counterattack, but differed in the frequency of their combination. Rat play appears the most complex because lab rats show the highest rates of defense, counterattacks, and role reversals. Montane voles showed high rates of defense while prairie voles showed high rates of counterattacks. Deer mice showed low rates of both.

Other useful comparative studies of play since Fagen (1981) include Biben (1982; 1983), Byers (1984), and Müller-Schwarze (1984).

 

Evolution

Comparative analyzes should lead to evolutionary hypotheses about the evolution of play as well as a portrayal of the phylogeny of play. The development of new comparative methods in systematics and behavior (e.g. Brooks and McLennan, 1991; Harvey and Pagel, 1991) was predicted to help rejuvenate a true comparative psychology (Burghardt & Gittleman, 1990) but this has so far not generally occurred. In the case of play the major lack may be in the quantity, quality and detail of the information available over closely related taxa.

Burghardt (1984, 1988) argued that play developed from precocial behavior evolutionarily modified by energetic constraints and the onset of parental care, leading to a diverse set of derived activities that could become, ultimately, incorporated into normal behavioral development in various ways. This theoretical outline is supported by an evolutionary model of play fighting in muroid rodents (Pellis, 1993). In reviewing extensive comparative results, Pellis concludes that play fighting is derived from precocial sexual behavior during juvenile development. Pellis elaborated a four stage model in which this precocial behavior becomes progressively necessary for the maturation of normal adult sexual performance (rudimentary play fighting or sexual play), then elaborated to differ from adult sexual behavior (true play fighting), and finally emancipated to also serve nonsexual functions (emancipated play fighting). In this model different species of rodents are at different stages along this path with house mice at the starting gate (no juvenile play fighting nor adult precopulatory behavior) through guinea pigs, gerbils, deer mice, voles, and rats. Although not explicitly phylogenetic, this model makes a number of predictions about changing functions, motivations, and neural control of play and play-like behavior that show how the complexity and diversity of play can be sorted out and understood.

 

Genetics and individual differences

Although play appears early in life and is often species characteristic, indicating that, as Groos held, play is instinctive or innate, there are few genetic studies focusing on determining if the individual variability seen in this behavior is genetically based. Two recent studies address this.

Locomotor neonatal jumping or popcorn behavior in house mice is often considered play and has been shown convincingly shown to be heritable (Walker & Byers, 1991). This behavior is normally seen in mice confronted with an attacking predator, such as a snake. Another study looked at the early behavior of horses born to different mares but with a small number of sires. Offspring behavior such as number of play invitations varied across sires indicating some genetic influence (Wolff & Hausberger, 1994). However, due to the small number of animals no genetic parameters could be estimated. Further, the variation in play could have been a consequence of variation in other measures such as mother - foal distance.

Since temperament in general is now a focus of intensive genetic study, it seems only a matter of time before more detailed information on genetic components of playfulness are reported.

 

Ontogeny

Most studies of ontogeny have looked at the course of play throughout the juvenile period: the type, context, frequency, duration, and targets. Among the animals that have been studied are rats (Panksepp, 1980), olive baboons, Papio anubis (Chalmers, 1980), cats (Bateson & Young, 1981), canids (Biben, 1983), gazelles (Gomendio, 1988), and fur seals (Harcourt, 1991a). Different kinds of play often wax and wane independently, indicating that they are under different ontogenetic pathways. These differences can provide clues as to possible functional and physiological roles. Studies that manipulate variables such as social contact, presence of mother, nutritional status and other variables are subsequent steps in the analysis of ontogeny that have much promise but are still too rare.

 

Function and adaptive value

Play, being primarily a behavior of young and growing animals, is often considered a major developmental factor (Smith, 1982) while at the same time being ignored. The major reason is the limited evidence that play has a demonstrable delayed adaptive function (Martin & Caro, 1985). Attempts to correlate individual differences in play among littermates with later ability to court, fight, hunt, or capture prey have been generally unsuccessful. But gender differences and species differences that can be associated with different demands of adult behavior are still often considered evidence that play has specific adaptive functions. For example, males in many polygynous species generally show more agonistic play than females and this is suggested to be necessary for improving the fighting behavior of males (Smith, 1982; Caro, 1988). Lack of opportunity to play in isolation-reared polecats (Mustela putorius) has been invoked as the reason they do not know the proper orientation to the neck, which is necessary both for killing dangerous prey such as rats and in subduing females during mating (Eibl-Eibesfeldt, 1963). The orientation of the neck bite must then be learned in considerably more risky contexts.

There are numerous postulated functions of play besides practice including its role in social bonding, dominance hierarchies, cardiovascular fitness, cognition, creativity, mastery, self-assessment, and parental assessment. A survey of these views is beyond this entry but Bekoff (1978), Fagen (1981), Smith (1982; 1984), Chiszar (1985), Caro (1988), and Thompson (in press) provide entree to the literature.

Recently the costs of play have received renewed attention (Caro, 1995). Play has time and energy costs, to be sure, but these have been argued to be minimal (Martin, 1984; but see Bekoff & Byers, 1992). In any event, one of the strongest relationships found in play research is that play is remarkably sensitive to environmental context and the players' physical and nutritional state (references in Fagen, 1981; Burghardt, 1984; Caro 1988). Thus these costs are buffered "naturally." This being the case, efforts are now being directed at costs that occur during play itself such as increased risk of predation, risk of physical injury from preforming the vigorous play activities, and time diverted from other seemingly more important activities. For example, Harcourt (1991b) has shown that 22 of 26 fur seals he observed killed by sea lions were attacked while playing, although play took up but about 6% of their waking hours.

A fine example of long-term quantitative field study of play in nature is Caro's (1995) study of play in cheetahs based on 2600 hours of observation of 40 families over the first year of life. This study shows both the strengths and weakness of field study based on correlational and design feature analysis. Caro found that risks of injury, maternal separation, and predation were minimal in cub play. In fact, mothers seemed to show no increased "unease" when their cubs played. That energy expenditure may be a cost was shown by the positive relationship between time spent eating and playing. Possible benefits of play were modest: there were positive relationships between the amount of time spent in object play and contact social play with the amount of contacts made with live prey released by the mother for her offspring to play with. Non-contact social play measures such as stalking, crouching, and chasing showed some subtle age-related changes.

In this study locomotor, object, contact social play and noncontact social play were distinguished. The developmental pattern of all varied. Locomotor play peaked first and Caro argues this could be because of the need to perfect flight responses at a time of maximal vulnerability to predators.

 

Physiological underpinnings

Vigorous play, the typical kind, takes energy. Thus a relationship between metabolic rate (basal, resting, and scope) and the extent of play has been noted (e.g. Fagen, 1981; Burghardt, 1984; Barber, 1991). In animals both terrestrial and aquatic, play should and often does occur most frequently in the more energy efficient medium (water) (Burghardt, 1988). Furthermore, the decline in play when animals lack adequate nutritious food supports this contention (references in Burghardt, 1984; Caro, 1988).

Although play is often considered an essential shaper of cognitive ability, even creativity, by both students of animal (e.g. Fagen, 1984) and human play, the effect of play on brain function is uncertain. Enriched environments that may provide opportunity for play with objects have been shown to lead to increased cortical growth in rats (Rosenzweig, et al., 1972; Greenough & Juraska, 1979). Pellis et al. (1992) performed experiments showing that neonatally decorticated rats played as much as control rats but more readily escalated to adult style defense. They suggest that cortical growth inhibits adult-style precocial behavior and provides a period for improving adult behavioral skills.

Recently Walker and Byers (1995) have documented a relationship between the timing of vigorous play in the ontogeny of three species (mouse, rat, and cat) and the time that most modification of muscle fiber type and cerebellar synapse formation is occurring. Of all the physiological and anatomical measures reviewed, only these have permanent effects and are associated with the waxing of playfulness. Thus play may facilitate the experience of balance and quickly adjusting motor movements so essential in the later lives of many animals. But note that this refinement of the motor training theory is agnostic on the specific abilities animals will use as adults.

The role of both neurotransmitters and hormones in play is an active area. Androgens have been implicated in the increased social and especially rough and tumble play of male mammals, especially primates and rodents. Spotted hyenas (Crocuta crocuta) have a social system where females dominate males and females have higher levels of circulating androgens relative to males than is typical in mammals (Pedersen, et al., 1990). In a study of social, locomotor, and object play in captivity, females were markedly more playful than males in several measures of social and locomotor play but did not differ in object play. In no context did males play more than females. This might suggest that androgens play a causal role except for the finding that gonadectomized animals did not differ in playfulness from normal animals (Pederson, et al., 1990). Much more work is needed on the role of hormones in play.

Since social play in rats is prominent and complex it is a prime candidate for drug manipulation. The administration of opiates (morphine) increased play whereas the antagonist naloxone decreased play fighting as measured by number of pins in rats (Panksepp, 1986). It appears that reduced serotonin and norepinephrine activity enhances play fighting. Since such reductions also accompany social deprivation, which also increases social play, we may be getting closer to some neural underpinnings of changing readiness to play (Panksepp, et al., 1987).

In a more recent experiment on the neural underpinnings of the play reward system, neither morphine nor naloxone interfered with learning a spatial task rewarded with the opportunity to play. However, the morphine injected subjects played significantly more in the goal box. This suggested to the authors that opiates regulate the expression of play rather than modulating its motivation (Normansell & Panksepp, 1990). This along with the other evidence cited above suggests that playfulness is a fairly intrinsically motivated behavioral phenotype.

 

Motivation

Early attempts to show a relatively autonomous drive for play by depriving animals of play had mixed results and were challenged on methodological and other grounds (Fagen, 1981; but see Müller-Schwarze, 1984; Ikemoto & Panksepp, 1992). Although motivation has been relatively ignored recently it may bear more analysis. Dwarf mongooses have a distinctive play vocalization that pulses 3-12 times per second throughout a play bout with an object. The repetition rate is related to the intensity of play and the associated behavior patterns (Rasa, 1984). Many observers have noted that social play increases after feeding. Pellis (1991), in addressing the motivational distinctness of play, noted that changes in the object and social play occurred as a function of hunger in a group of captive oriental small-clawed otters (Anonyx cinerea). The typical object play sequence appeared similar to the gathering, handling, fragmenting, and chewing of food. Social play involved attempts to gently bite a conspecifics cheek. As feeding time approached the otters increased object play and decreased social play (indeed even threatened each other over objects). But once satiated, object play declined and social play markedly increased. Although this is not definitive evidence that different kinds of play may be linked to primary motivations rather than separate play "drives" or instincts, it greatly implies non-independence.

 

Need for Integration

Research on play involves many diverse issues. At this point, the heterogeneous category of play does not allow a comprehensive model such as those proposed over the years for learning, instinct, motivation, altruism, or territoriality. But, of course, none of those models has endured. The direct practice theory and its close relation, cognitive development, are still the most intuitively attractive and easily grasped. Their hold on researchers in spite of a dearth of experimental evidence is intriguing.

There are other models of play worthy of consideration. These include the energy and thermal regulation model (Barber, 1991); metamorphosis theory (Coppinger and Smith, 1989); surplus resource theory (Burghardt 1984, 1988), precocial ritualization theory (Pellis, 1993), and motor training theory (Byers & Walker, 1995). These models have much in common and combine physiological, life history, and ontogenetic factors to various degrees. They go beyond simple delayed benefit models of all stripes (practice, social, cognitive, flexibility). These delayed benefits will most likely be shown to be secondary benefits only possible after proximate processes developed and modified the maturation and experiential support of complex adult behavior during the long evolutionary history of vertebrates.

 

References

Allen, C. and Bekoff, M. (1994). Intentionality, social play, and definition. Biology and Philosophy, 9, 63-74.

Baldwin, J. M. (1902a). Development and evolution. New York: Macmillan.

Baldwin, J. M. (Ed.). (1902b). Dictionary of philosophy and psychology. Vol. 2, New York: Macmillan.

Barber, N. (1991). Play and energy regulation in mammals. Quarterly Review of Biology, 66, 129-147.

Bateson, P., & Young, M. (1981). Separation from the mother and the development of play in cats. Animal Behaviour, 29, 173-180.

Beach, F. A. (1945). Current concepts of play in animals. American Naturalist, 79, 523-541.

Bekoff, M. (1975). Social play and play-soliciting by infant canids. American Zoologist, 14, 323-340.

Bekoff, M. (1978). Structure, function, and the evolution of cooperative social behavior. In G. M. Burghardt & M. Bekoff (Eds.), The Development of Behavior: Comparative and Evolutionary Aspects (pp. 367-383). New York: Garland STPM.

Bekoff, M. (1995). Play signals as punctuation: The structure of social play in canids. Behaviour, 132, 419-429.

Bekoff, M., & Byers, J. (1981). A critical reanalysis of the ontogeny and phylogeny of mammalian social and locomotor play: An ethological hornet's nest. In K. Immelmann, G. W. Barlow, L. Petrinovich, & M. Main (Eds.), Behavioral Development: The Bielefeld Interdisciplinary Project (pp. 269-337). Cambridge: Cambridge University.

Bekoff, M., & Byers, J. (1992). Time, energy and play. Animal Behaviour, 44, 981-982.

Biben, M. (1982). Object play and social treatment of prey in bush dogs and crab-eating foxes. Behaviour, 79, 201-211.

Biben, M. (1983). Comparative ontogeny of social behavior in three South American canids: the maned wolf, crab-eating fox, and bush dog: Implications for sociality. Animal Behaviour, 31, 814-826.

Brooks, D. R., & McLennan, D. A. (1991). Phylogeny, Ecology, and Behavior. Chicago: University of Chicago Press.

Büchner, L. (1880). Mind in animals. London: Freethought.

Burghardt, G. M. (1984). On the origins of play. In P. K. Smith (Ed.), Play in Animals and Humans (pp. 5-41). Oxford: Basil Blackwell.

Burghardt, G. M. (1988). Precocity, play, and the ectotherm-endotherm transition. In E. M. Blass (Ed.), Handbook of Behavioral Neurobiology: Vol. 9, (pp. 107-148). New York: Plenum.

Burghardt, G. M. (i1997). Amending Tinbergen: a fifth aim for ethology. In R. W. Mitchell, N. V. Thompson, & L. Miles (Eds.). Anthropomorphism, Anecdotes, and Animals: The Emperor's New Clothes. Albany, NY: SUNY Press.

Burghardt, G. M. & Gittleman (1990). Comparative and phylogenetic analyses: New wine, old bottles, pp. 192-225. In M. Bekoff & D. Jamieson (Eds.), Interpretation and explanation in the study of behavior: Vol. 2. Comparative perspectives, (pp. 192-225). Boulder, CO: Westview Press.

Burghardt, G. M., Ward, B., & Rosscoe, R. (1996). Environmental enrichment and play behavior in a captive Nile soft-shelled turtle, Trionyx triunguis. Zoo Biology.

Byers, J. A. (1984). Play in ungulates. In P. K. Smith (Ed.), Play in Animals and Humans (pp. 43-65). Oxford: Basil Blackwell.

Byers, J. A. & Walker, C. (1995). Refining the motor training hypothesis for the evolution of play. American Naturalist (in press).

Caro, T. M. (1988). Adaptive significance of play: are we getting closer. Trends in Ecology and Evolution, 3, 50-54.

Caro, T. M. (1995). Short-term cost and correlates of play in cheetahs. Animal Behaviour, 49, 333-345.

Chalmers, N. R. (1980). The ontogeny of play in feral olive baboons. Animal Behaviour, 28, 570-585.

Chiszar, D. (1985). Ontogeny of communicative behaviors. In E. S. Gollin (Ed.), The comparative development of adaptive skills: evolutionary implications. Hillsdale, NJ: Erlbaum.

Coppinger, R. P., & Smith, C. K. (1989). A model for understanding the evolution of mammalian behavior. In H. Genoways (Ed.), Current mammalogy (Vol. 2, pp. 335-374). New York, Plenum.

Darwin, C. (1874). The descent of man (2nd ed.). London: Murray.

Eibl-Eibesfeldt, I. (1963). Angeborenes und Erworbenes im Verhalten einiger Säuger. Zeitschrift für Tierpsychologie 20, 705-754.

Fagen, R. (1981). Animal play behavior. New York: Oxford University.

Fagen, R. (1984). Play and behavioral flexibility. In P. K. Smith (Ed.), Play in Animals and Humans (pp. 159-173). Oxford: Basil Blackwell.

Gomendio, M. (1988). The development of different types of play in gazelles: implications for the nature of the functions of play. Animal Behaviour, 36, 825-836.

Greenough, W. T., & Juraska, J. M. (1979). Experience-induced changes in brain fine structure: their behavioral implications. In M. E. Hahn, C. Jensen, & B. C. Dudek (Eds.), Development and Evolution of Brain Size: Behavioral Implications (pp. 295-320). New York, Academic Press.

Groos, K. (1898). The play of animals. New York: D. Appleton

Hall, G. S. (1904). Adolescence, its psychology and its relations to physiology, anthropology, sex, crime, religion and education (Vol. 1). New York: D. Appleton.

Harcourt, R. (1991a). The development of play in the South American fur seal. Ethology, 88, 191-202.

Harcourt, R. (1991b). Survivorship costs of play in the South American fur seal. Animal Behaviour, 42, 509-511.

Harré R., & Lamb, R. (Eds.) (1983). The Encyclopedic Dictionary of Psychology. Oxford: Basil Blackwell.

Harvey, P. H., & Pagel, M. D. (1991). The Comparative Method in Evolutionary Biology. Oxford: Oxford University Press.

Hole, G. J., & Einon, D. F. (1984). Play in rodents. In P. K. Smith (Ed.), Play in Animals and Humans (pp. 95-117). Oxford: Basil Blackwell.

Ikemoto, S., & Panksepp, J. (1992). The effects of early social isolation on the motivations for social play in juvenile rats. Developmental Psychobiology, 25, 261-274.

Lazar, J., & Beckhorn, G. D. (1974). Social play or the development of social behavior in ferrets. American Zoologist, 14, 405-414.

Lindsay, W. L. (1879). Mind in the lower animals (2 vols.). London: Kegan Paul.

Markowitz, H. (1982). Behavioral Enrichment in the Zoo. New York: Van Nostrand Reinhold.

Martin, P. (1984). The time and energy costs of play behaviour in the cat. Zeitschrift für Tierpsychologie, 64, 298-312.

Martin P., & Caro, T. M. (1985). On the functions of play and its role in behavioral development. Advances in the Study of Behavior, 15, 59-103.

McDougall, W. (1924). An outline of psychology, 2nd ed., London: Methuen.

Mitchell, R. W., & Thompson, N. S. (1991). Projects, routines, and enticements in dog-human play. Perspectives in Ethology, 9, 189-216.

Morgan, C. L. (1920). Animal behaviour (2nd. ed.). London: Kegan Paul.

Müller-Schwarze, D. (Ed.) (1978). Evolution of play behavior. Benchmark papers in animal behavior, Vol. 10. Stroudsburg, PA: Dowden, Hutchinson & Ross.

Müller-Schwarze, D. (1984). Analysis of play behaviour: What do we measure and when? In P. K. Smith (Ed.), Play in animals and humans (pp. 147-158). Oxford: Basil Blackwell.

Normansell, L., & Panksepp, J. (1990). Effects of morphine and naloxone on play-rewarded spatial discrimination in juvenile rates. Developmental Psychobiology, 23, 75-83.

Panksepp, J. (1980). The ontogeny of play in rats. Developmental Psychobiology, 14, 327-332.

Panksepp, J. (1986). The neurochemistry of behavior. Annual Reviews of Psychology, 37, 77-107.

Panksepp, J., Normansell, L., Cox, J. F., Crepeau, L. J., & Sacks, D. S. (1987). Psychopharmacology of social play. In B. Olivier, J. Mos, & P.F. Brain (Eds.), Ethopharmacology of agonistic behaviour in animals and humans (pp. 132-143). Dordrecht, Netherlands: Martinus Nijhoff.

Panksepp, J., Siviy, S., & Normansell, L. (1984). The psychobiology of play: Theoretical and methodological perspectives. Neuroscience and Biobehavioral Reviews, 8, 465-492.

Pedersen, J. M., Glickman, S. E., Frank, L. G., & Beach, F. A. (1990). Sex differences in the play behavior of immature spotted hyenas, Crocuta crocuta. Hormones and Behavior, 24, 403-420.

Pellegrinni, A. D., & Boyd, B. (1993). The role of play in early childhood development and education: Issues in definition and function. In B. Spodele (Ed.), Handbook of research in childhood education (pp. 105-121). New York: Macmillan.

Pellis, S. M. (1991). How motivationally distinct is play? A preliminary case study. Animal Behaviour, 42, 851-853.

Pellis, S. M. (1993). Sex and the evolution of play fighting: A review and model based on the behavior of muroid rodents. Play Theory and Research, 1, 55-75.

Pellis, S. M., & Pellis, V. C. (1987). Play-fighting differs from serious fighting in both target of attack and tactics of fighting in the laboratory rat Rattus norvegicus. Aggressive Behavior, 13, 227-242.

Pellis, S. M., Pellis, V. C., & Dewsbury, D. A. (1989). Different levels of complexity in the play fighting by muroid rodents appear to result from different levels of intensity of attack and defense. Aggressive Behavior, 15, 297-310.

Pellis, S. M., Pellis, V. C., & Whishaw, I. Q. (1992). The role of the cortex in play fighting by rats: Developmental and evolutionary implications. Brain, Behavior and Evolution, 39, 270-284.

Rasa, O. A. E. (1984). A motivational analysis of object play in juvenile dwarf mongooses (Helogale undulata rufula). Animal Behaviour, 32, 579-589.

Romanes, G. J. (1892). Animal intelligence (5th ed.). London: Kegan, Paul, Trench, Trübner.

Rosenzweig, M. R., Bennett, E. L., & Diamond, M. C. (1972). Brain changes in response to experience. Scientific American, 226, 22-29.

Smith, P. K. (1982). Does play matter? Functional and evolutionary aspects of animal and human play. Behavioral and Brain Sciences, 5, 139-184.

Smith, P. K. (Ed.) (1984). Play in Animals and Humans. Oxford: Basil Blackwell.

Spencer, H. (1872). Principles of psychology (Vol. 2 pt. 2). New York: D. Appleton.

Thompson, E. P. (1851). The passions of animals. London: Chapman & Hall.

Thompson, K. V. (in press). Behavioral development and play. In D. Kleiman, M. E. Allen, K. V. Thompson, & S. Lumpkin (Eds.), Wild Mammals in Captivity, Principles and Techniques, Chicago: University of Chicago Press.

Tinbergen, N. (1963). On the aims and methods of ethology. Zeitschrift für Tierpsychologie, 20, 410-433.

Walker, C. and Byers, J. A. (1991). Heritability of locomotor play in house mice, Mus domesticus. Animal Behaviour, 42, 891-898.

Wilson, E. O. (1971). The insect societies, Cambridge: Harvard University (Belknap).

Wolff, A. and Hausberger, M. (1994). Behaviour of foals before weaning may have some genetic basis. Ethology, 96, 1-10.