Agner Fog: Cultural selection © 1999

12. Play, games and sport

12.1 Play

Play may seem a useless activity, but it has an important function. Through play children learn and develop their brain. Babies babble in order to learn to pronounce the sounds of the language, they shake their rattles or throw their dummies in order to train their motor functions, to understand how things move, and to explore the law of gravity. Bigger children play role games where they learn to understand and master the roles which are necessary in the daily life of adults. The importance of play for children's development is indicated by how much time and energy they spend on playing (Fagen 1981:272). Children may occasionally be conscious that they are learning, for example when practicing to catch a ball. But playing may very well be functional without the child being conscious of this function.

Playing may sometimes be quite risky. Boys, especially, love games involving speed, excitement, fighting, or daring acrobatics. These dangerous games have an important function which outweighs the risk: The child learns to understand dangerous situations and gets the chance to experiment with alternative responses. He thereby learns fighting techniques, techniques of controlled falling, evasive actions, strategy, and other skills which later in life may turn out to be of vital importance in a hazardous situation. It may be a question of life and death to be able to react speedily to a sudden danger. Intellectual knowledge is not sufficient here, because it takes too long to activate it from memory. A prompt reaction can only be achieved by learned reflexes, and these reflexes can only be learned through 'violent' games. Children learn to know the dangerous situations before they become real in the relatively protective atmosphere of play. This has been called buffered learning (Roberts & Sutton-Smith 1962, Sutton-Smith & Roberts 1970).

Play is far from unique to humans. Most mammals play, as do many birds (Fagen 1981). In most primate species it is common to see young females play mother towards infant conspecifics. They hereby learn how to treat baby monkeys, and this knowledge may be vital when they later get their own young (Lancaster 1971; Fairbanks 1990; Gould, L. 1992)28. Observations of apes raised in captivity have shown that females who have had no opportunity of learning how to treat infants are unable to take care of their own young (Harlow & Harlow 1969). Similar experiments have obviously not been carried out with humans, but there is reason to believe that children's playing with dolls and playing mother towards smaller children has an equally vital function (Eibl-Eibesfeldt 1989:589ff).

The theory that play is a method of learning was formulated clearly by Karl Groos (1896, 1899) a hundred years ago, and the recognition that play can be educational is undoubtedly much older. However obvious this learning theory may seem, it is still disputed, and there are many alternative theories which space does not permit to be discussed here (Travick-Smith 1989, Smith & Syddall 1978, Bekoff 1976, Reilly 1974a).

Developmental psychologist Jean Piaget regards play as hardly more than an epiphenomenon to the cognitive processes of a child (Piaget 1945). Piaget sometimes criticizes, sometimes praises, the learning theory of Karl Groos. This inconsistency is concealed behind a high level of abstraction to such a degree that some authors cite Piaget for being a proponent of the learning theory, whereas many others cite him for the opposite. Piaget is traditionally considered one of the most important authorities on play theory alongside the culture historian Johan Huizinga. The latter may be said to turn the theory upside down. Where most others would say that children imitate the culture of adults in their play, Huizinga attempts throughout his book to prove that culture imitates play. He thereby sees play as fundamental to our entire culture with almost metaphysical importance (Huizinga 1938). His thinking is an example of the classical idealist philosophy and theology which opposes the rational and mechanistic conception of humans and emphasizes the autonomous and creative aspects of human nature (Gruneau 1983:24; Norbeck 1977).

Philosophy and science as such are also subjects of cultural selection, and not even the most reputable scientist is unaffected by the culture in which he lives. This not only applies to Huizinga, but also to Piaget. When reading Piaget's descriptions of the observations on which his theory is based, it is apparent that he yields to the preconception that everything which is pleasurable or funny is useless and a waste of time, whereas serious activities are useful. This preconception stems from puritanism - a regal philosophy which suppresses individual self-determination in favor of social control. One of the means of puritanism is to claim that any pleasurable act is useless or dangerous, so as to suppress the individual pursuit of pleasure. (I have explained the function of pleasure in chapter 13.3).

Despite the fact that the books of Piaget and Huizinga are the most often cited works in play theory, it is a prevalent opinion within the areas of pedagogy, psychology, anthropology, and ethology, that play has an important learning function, and this theory is supported by several studies of children as well as by animal experiments (Weininger 1978, Reilly 1974b, Roberts & Sutton-Smith 1962, Sutton-Smith & Roberts 1970, Robinson 1978, Fagen 1981, Caro 1988, Corsaro & Tomlinson 1980, Rosenstiel 1977, Lancy 1977, Golomb & Cornelius 1977, Sylva 1977, Garvey 1977, Feitelson & Ross 1973, Dansky & Silverman 1973, Boulton & Smith 1992, Eibl-Eibesfeldt 1989, Fagen 1981, Oakley & Reynolds 1977).

A learning process which is genetically controlled is called programmed learning (Gould, J. & Marler 1987; Mayr 1974). Genetic programs are called open or closed depending on the flexibility of the learning process. Let me explain this difference with a few examples. Humans are not born with the ability to walk, but they are born with the ability to learn how to walk, and any normal child will sooner or later learn to walk. The genetic program not only gives the child the ability to learn, but also determines rather precisely what to learn. The ability to learn to walk is a closed genetic program. Another program, the ability to learn to talk, is much more open or flexible in the sense that it is not fixed to one particular language. The cultural differences in people's languages are much bigger than the differences in their way of walking. But the flexibility is not unlimited. If a couple of children are left to themselves without sufficient verbal stimulation from adults, then they will invent their own language containing nouns, verbs, etc. The communicative function of language is predetermined to a high degree.

Let us return to play as a learning process. Is this process controlled by an open or closed program? There are important differences between children's way of playing in different cultures, but there are also striking similarities, and children in different cultures go through the same stages of development in their playing behavior (Seagoe 1970). The program has certain degrees of freedom, but also certain limitations. Children imitate the behavior of adults in their play, and this activity will therefore necessarily reflect cultural differences in behavior. But there are certain themes which are repeated in the playing behavior of children in widely different cultures, and some of the skills that children learn through play are quite superfluous for the culture in which they live. For example, children love to build caves although it is thousands of years ago that humans lived in caves. And even in the most peaceful societies children love to play war. War toys are popular despite parents' and teachers' disapproval. Themes like fighting, hunting, building huts, child care, and sexuality, are so universal in children's play that there is reason to believe that there are instincts or genetic programs behind these behavior forms and that these programs penetrate into the playing behavior.

Homo sapiens is the most flexible animal on Earth, and the ability to adapt to varying environments and ways of life is an important factor in the ecological success of the humans. But flexibility requires learning. A behavior which is controlled only by genes will always be robot-like and inflexible. The more flexible an animal is, the more time it must spend on learning the necessary behaviors while growing up. Humans therefore have an unusually long childhood compared with other animals. The more complicated the society, the more skills the children have to learn, and the longer the period of education.

This brings us back to the genetic r/K-theory (see chapt. 4.3). A genetic r-strategy is a strategy where each individual spends a maximum of resources on breeding as fast as possible and producing numerous offspring, but uses no energy on caring for or protecting its progeny. The opposite is a K-strategy, where each individual gets very few young, but spends a maximum of resources on giving each young the best possible conditions. The reproductive strategy of humans is a typical K-strategy. The long childhood requires that the parents spend a lot of energy on childcare and education. It has been observed that K-selected animals generally play more than r-selected animals, in perfect accordance with the r/K-theory (Fagen 1981:489).

There is a close connection between the genetic and the cultural r/k-dimension here. Humans in regal cultures get many children. The children are set to work from an early age, and also become able to take care of themselves early. Not so in kalyptic cultures where large families are rare and where the children are cared for by the parents for many years. This difference is also reflected in the different attitudes towards play.

Ariel and Sever (1980) have compared children's play among bedouins and kibbutz children in Israel. The difference between these two cultures is enormous with regard to the treatment of children. Bedouin children get no formal education and are set to work in the household from an early age. The children have no toys, and any kind of play is unwelcome and sometimes punished severely. The kibbutz children, on the other hand, get twelve years of school education. They have lots of toys, their kindergartens are richly equipped, and the adults take considerable care of the children. Not surprisingly, the play of the kibbutz children is much more varied and imaginative than that of the bedouins (Ariel & Sever 1980). These observations are in accordance with the fact that the bedouin culture is much more regal than the kibbutz culture. There is no difference between the predictions of the genetic r/K-theory and the cultural r/k-theory here, but since the two populations are genetically related, it is concluded that the difference must be due to cultural factors.

It is obvious that children's play reflects the culture in which they live since playing is often an imitation of the adult world. David Lancy has studied children's play in Liberia and found that children learn the behavior patterns of adults in four ways: by watching the adults work, by imitating the activities of adults in their play, by helping the adults, and by direct instruction (Lancy 1977). In particular, children imitate what they have no access to do for real. In the imitating play, children learn the motor and cognitive aspects of those activities to which they have no access in reality. Observations in cultures where from an early age children are set to do housework, show that these children do not imitate housework in their play. There is no reason to play what you can do for real (Storey 1977, Ariel & Sever 1980). But the children imitate what they cannot easily obtain. A very common play theme among the bedouin children in Israel as well as among liberian children is imitation of motor vehicles. Cars and motorcycles are attractive to the children because they represent a rich and privileged way of life which few bedouins have a chance of achieving (Ariel & Sever 1980, Lancy 1977). These observations come from relatively primitive cultures which have contact with a more technologically and economically advanced culture which may make them atypical because of the dramatic cultural contrasts.

Childishness and play are often associated with fantasy in our culture, but not all children are equally imaginative. Fantasy play is found in cultures which value flexibility, inventiveness and individual initiative, but not in cultures where everything is controlled by fixed systems and strict rules (Storey 1977, Ariel & Sever 1980, Feitelson 1977). There is therefore reason to assume that the function of fantasy play is to learn inventiveness and independent initiative.

Historical studies of the play of american children reveal that there have been considerable changes in the playing patterns during the twentieth century. Formalized and organized group games have dropped in popularity, and singing games have almost disappeared. Informal group activities like swimming, fishing, hunting, sailing, cycling, and roller skating have become more popular, probably because society has become less hierarchical and formalized (Sutton-Smith & Rosenberg 1961). The altered sex roles in the society are also reflected in the play. The difference between boys' play and girls' play has become smaller, but it is remarkable that it is the playing pattern of the girls that has approached that of the boys, whereas boys' play has not moved in the direction of girls' (Sutton-Smith & Rosenberg 1961).


12.2 Games

If you want to compare playing patterns in different cultures, then you may have to classify plays into categories. Caillois (1955) has defined four motifs in play and games: competition, games of chance, imitation, and the element of speed and excitement as is found for example in roller coasters and motor sport. These categories are hardly exhaustive, as you could add motifs like exploration, experimenting, fantasy and creativity. Another dimension in Caillois' classification is a graduation from spontaneous and improvised play to organized and regular playing, i.e. games. Games are easier to categorize and systematize than fantasy plays, imitation, and exploration plays, and it is therefore easier to make cross-cultural comparisons of games than of other kinds of play.

A particular classification system for games, introduced by John Roberts and his colleagues, distinguishes between whether the outcome of a game is determined by the physical skills of the players, or of strategy, or pure randomness. These three motifs are connected with important aspects of child education and fundamental structural principles in the society, according to cross-cultural studies. Games based on physical skill are most prevalent in societies that attach importance to individual achievements, self-control, and control over the physical environment. Games of strategy represent social complexity, stratification, obedience, independence, symbolic rewards, and psychological control. Finally, games of chance are connected with religion and magic, and the belief that supernatural beings can be influenced through rituals. Apparently, the game patterns of both children and adults reflect sex, class, and cultural differences. The authors explain these observations partially with reference to the theory that the games symbolize intrapsychic conflicts, and partially by assuming that children learn social skills by using the game as a model (Roberts, Arth & Bush 1959; Roberts & Sutton-Smith 1962, 1966; Sutton-Smith, Roberts & Kozelka 1963; Sutton-Smith & Roberts 1970).

At the end of the 19th century, american children were playing various games of skill where the looser was punished. Interestingly, these games have disappeared in the modern society, and have been replaced by games where the winner is honored (Sutton-Smith & Rosenberg 1961). This reflects the fundamental principle in the modern capitalist society where economic rewards to skilful and efficient persons are more important as controlling means than punishment of the lazy and incompetent. It is also interesting to observe that solidaric countries, where competition has little importance, have games where no winner or looser is singled out, but where the process rather than the result is the important thing (Calhoun 1987:60. For specific examples see: Ager 1977, Rosenstiel 1977).

The theories of a connection between playing and social structure have gained wide recognition (Calhoun 1987) despite some criticism (Townshend 1980). Unfortunately, most anthropological studies of play have concentrated on games rather than unorganized play. This probably gives a bias since rule-governed games are more prevalent in societies controlled by complex rules than in societies where flexibility is important. Some of the studies do not distinguish between the games of children and the games of adults, although the latter may have other functions. Similar games may also have different functions in different societies (Lüschen 1970).


12.3 Sport

Sport is a form of game, played by children and adults alike, where the players train physical skills. The distribution of various sports in different countries can best be explained by the theory of culture centers from where new phenomena spread by diffusion (see chapt. 2.5).

In the mid 19th century, Great Britain was the leading nation in the world for industrialization and colonization. Britain was also a cultural center for sport. It was here that sports evolved into their modern form with their associations, rules, and competitions. Football or soccer, which was the most popular sport in Britain, spread in the late 19th century to several other countries, followed by cricket, hockey, tennis, rugby, golf, and track and field athletics (Stokvis 1989).

In Germany, gymnastics was part of the well organized educational system, and was promoted with the purpose of strengthening the physical and mental health of the population to the benefit of the industrial development. Gymnastics has spread to those countries which were within the german sphere of influence, in particular the Netherlands, Scandinavia, and Eastern Europe.

Sport in the USA underwent its own development in the 19th century rather independently of Europe. During the 1840's baseball and cricket, which was at least equally popular, were organized in sports associations which laid down the rules of the games. However, the anti-british atmosphere connected with the civil war made cricket less popular because it was regarded as english. But baseball, which was regarded as a typical american sport, soon became the most popular sport in the USA. In the late 1870's, american football was introduced as a variant of the english rugby. Prestigious universities like Yale and Harvard played an important role for the spreading of this game. Basketball and volleyball arose because there was a need for some more exciting activities to put into the already existing gyms. Team handball arose for the same reason in Germany (Stokvis 1989).

From these few cultural centers various sports have since spread far and wide to countries with very different cultures. British sports have spread as far as to Africa and South America, while the north american sports patterns have spread to Japan, France, and Australia (Stokvis 1989).

Various psychological theories have been proposed for explaining the difference between people's sports preferences in different countries. Arens (1978) and Duthie (1980) think that the marked division of labor in american football reflects the industrial society, whereas Calhoun (1987:232) notes the element of territorial war in this game, which he regards as symbolic of the conquest of America. Some researchers think that football is a typical urban sport whereas baseball belongs in the rural districts (Guttmann 1978, Wagner 1988). Such theories have the obvious weakness that they cannot explain why a sport like soccer is the most popular sport in cultures as different as Europe, Africa, and South America, while baseball is more popular in North America and Japan.

History shows that the distribution of sports depend more on political power structures and spheres of influence than on psychological and cultural preferences. Soccer has spread because of the influence of Britain as a leading colonial power. Baseball has spread because of american nationalism. American football has spread because it was played on prestigious US universities. It seems that diffusion is more important than selection for the distribution of sports, and several sociologists have remarked that sport is an autonomous phenomenon which is rather independent of art, religion, economy, or other cultural phenomena (Robins 1982, Bourdieu 1978, Nixon 1982, Lüschen 1967, Maheu 1962).

I would not claim, though, that no selection takes place at all. Comparing the american sports, it is easy to see that not all sports spread equally effectively. The japanese, for example, have accepted baseball but not american football. Basketball and volleyball have been well received by the europeans, whereas baseball is less popular in europe, and american football is very seldom played here. There are two possible explanations why europeans and japanese do not play american football. One possibility is that this game has been deselected because it is too violent (Stokvis 1989). The other possibility is that this game is complicated to learn and requires special equipment that is difficult to obtain, which makes a barrier against diffusion. If the diffusion barrier hypothesis is correct then the europeans would continue to play american football if they had learned it, but if psychological selection is at work here, then the europeans would loose the interest in american football even after having learned to play it. Which of the two possible mechanisms is more influential here cannot be determined on the existing basis, but the two mechanisms are not mutually exclusive.

If we assume that sport, like other kinds of play, functions as a model for social phenomena or everyday doings, then it should be possible to find connections between sport preferences and social structure. Many sociologists have studied such connections. The relationship between sport and religion is discussed by Lüschen (1967) who notes that the interest in sports is higher among protestants than catholics, especially when it comes to individual sports (curiously enough, he does not mention the low sports activity in muslim countries). Lüschen explains this difference with reference to how much importance the different cultures attach to individual accomplishments, and he finds that competitive sports are more popular in societies where social status depends on individual performance (Lüschen 1967).

This is confirmed by historical studies. In the primitive societies of the past, bodily exercises took the form of rituals, and often with a religious meaning. In the modern society this has changed into a secular sport where competition, quantification, and records have a prominent importance (Guttmann 1978; see also Ibrahim 1976). Critics of this characteristic claim that there are also examples of quantification and competition in ancient times, and that there are still rituals in modern sport (Carter & Krüger 1990, Eichberg 1986). Obviously, the technological evolution has had an important influence on sport. Just think of what the invention of vulcanized rubber has meant for balls and wheels, or what modern mass communication has meant for spectator sports and sports journalism (Betts 1953).

A very refined kind of sport was played in the 17th century. Geometrical precision and elaborate formations was a common feature in riding and fencing. This very aristocratic manner was a parallel to the minuets in dance and the baroque in art (Eichberg 1986).

It may seem natural to compare competitive sport with war. In discussions on sport you sometimes hear the hypothesis that sport provides an outlet for accumulated aggressions which otherwise may have found their expression in violence or war. The opposite of this catharsis theory is the hypothesis that aggressive behavior is learned, and that warlike sports therefore are more likely to be found in militant societies. In order to settle this theoretical debate, Richard Sipes has analyzed the prevalence of war and sports in various societies. He found a positive correlation between war and war-like sports, and on this basis he rejects the catharsis theory in favor of a theory of cultural learning (Sipes 1973, 1975; Hietanen 1982). According to this theory, we would expect competitive sports to be more prevalent in regal than in kalyptic cultures, but it is hard to find support for this claim in modern societies. On the contrary, it is easy to find examples of kalyptic societies where competitive sports are popular and regal societies where they are not. There is therefore reason to assume that competitive sports not only serve as a model for war, but also as a model for economic competition. The reason for the discrepancy with Sipes' statistics may be that his studies are based solely on primitive societies where economic competition is unimportant or non-existent.

While it is difficult to find specific connections between sport and social structure, it is easier to see a connection between the social status of a person and his sporting preferences. Golf and tennis, for example, are typical upper class sports. In general, it has been observed that sports where the player increases his power and reach by means of a tool are mostly preferred by members of the higher social classes who are used to controlling big flows of resources without physical effort. Contact sports and team sports, on the other hand, are more popular among the working classes (Metheny 1968:72, Bourdieu 1978, Sack 1988, Lüschen 1969, Sibley 1988). These observations support the theory that sport functions as a model for the work situation of a human.

When speaking of the function of sport, you must also mention the conscious use of sport for political purposes. Sport has often been promoted with the purpose of raising the international prestige of a country, to improve the health of the population, to control and socialize, to create solidarity and common identity, or to create international understanding (Riordan 1974, 1982; Miracle 1980; Eichberg 1973; Brohm 1976; Hietanen 1982; Whitson 1984).


12.4 Comparison with other cultural phenomena

I have mentioned that playing and other learning mechanisms are necessary for the flexibility and adaptability of the human race. When a playing child imitates an adult, he learns the behavior of the adult, and we may say that a cultural transmission has taken place. Play may also be experimentation and exploration, where the child does not learn from others but learns to know his physical environment. Or play may be an exercise where the child learns to control his own body. Thus only some of the playing activities of a child are part of the cultural transmission process.

As explained in the previous chapters, religion and art are important transmission channels for cultural information. Comparing play and art as media for cultural transmission, we find some fundamental differences which are worth mentioning. In music, dance, pictorial art, etc., the sender of the message plays an active role. The receiver may sit passively and watch and listen, or he may join in the dancing and singing in a mutual exchange of cultural information. But in the imitative play, the receiver of the information is always the most active part. The adult does not have to do anything special to be imitated by the child. The child may even watch the adult without the latter noticing it, and may later repeat in play what he has seen. The adult may be completely ignorant of what the child has learned from him. You may therefore say that imitative play has the character more of assimilation than communication. Another difference between art and play is that art is a self contained bearer of information separate from the transmitted behavior, just like the genetic code. But by learning through watching and imitation there is no separate bearer of information.

The message transmitted through art is usually about social structures and interaction patterns. But in play, there is a much broader spectrum of knowledge that is acquired, including individual as well as social skills. Art, music, dance, etc. are therefore mostly social phenomena, whereas playing not always is so.

Returning to the discussion of programmed learning and its degrees of openness (see above), you may say that playing has a significant degree of freedom in the sense that, through play, the child may learn a wide range of skills. On the other hand, playing is not so flexible when it comes to cultural differences. There are significant similarities between children's play in different cultures, and the child often learns skills that he has no need for in the culture in which he lives. Physical exercise or sport, in particular, represents a relatively closed program in this respect. Among the cultural phenomena I have studied, sport is the one that depends least on the social structure. Most sports are immediately intelligible and acceptable to anyone regardless of cultural background. This is the reason why sport is such an effective means of creating international understanding and cooperation, and of integrating an inhomogeneous population.

As a contrast to play and sport stands art as a rather specific communication channel for social structure and interaction-patterns. Art is unsuited for transmitting anything other than cultural information, but has quite wide limits with respect to the possible social structures. Art therefore depends much more on culture than does sport, and this is also true of religious rituals, dogma and myths.


28. This behavior may also have other functions in certain species, according to Tanaka (1989) and Stanford (1992).