Sociology of Child Development

Chapter Eight

Play, Games,

and Interaction

In this chapter, I will briefly review and critique existing theories and definitions of play, games, and socialization. A view of interaction and socialization consistent with the formulations of G. H. Mead and the symbolic interactionist perspective will be offered.

Play, Games, and Childhood Socialization

The import of play and games for early childhood socialization experiences has long occupied the interests of sociologists, social psychologists, psychologists, and students of child development, among them, Mead, Cooley, Groos, Hall, Huizinga, Caillois, Piaget, Freud, Herron, Sutton-Smith, Lueschen, Ellis, I. and P. Opie, Aries, Singer, Stone, Goffman, Waller, Berlyne, and Merton. Research and theory has typically been concerned with one or more of the following questions: How can play be defined and what is unique to childhood play? (Huizinga, 1939; Piaget, 1962; Mead, 1934). Are there different forms of play? (Caillois, 1961; Groos, 1901; Piaget, 1962). Why do people, and children in particular, play? (Huizinga, 1939; Freud, [1922] 1959). What are the consequences of play and games for the socialization process? (Stone, 1965; Merton, 1957).

In general, play is seen as serving anticipatory socialization purposes for the young child. It is conceptualized as a form of social behavior that results in children learning to cooperate and interact with others. It is variously seen as infantile or sophisticated in nature. Play and games are seen as establishing the vehicles for teaching the child how to take the perspective of others (Mead, 1934). Play is seen as displaying shifting degrees of the fantastic and the real. Piaget (1962), for example, views the play of young children as lacking in accommodative abilities and in this sense as revealing the limited cognitive abilities of the child player.

Classical, contemporary, and recent theories compete for attention, yet by and large no consistent theory or definition of play and games relevant to the socialization process currently exists in the literature (Berlyne, 1969). Central to current misunderstanding in this area is the failure of all theorists to distinguish systematically play from games. Furthermore, existing formulations typically attempt to bracket or separate play from work. It will be necessary to make a distinction between these interactional forms in the discussion that follows. Playing a game of marbles, for example, is clearly different from playing dolls or pretending that a blanket is a baby that will not go to sleep. Yet current formulations confuse these two processes: The first is rule governed and gamelike; the second is not. A set of distinctions separating play, games, work, playing at play, and playing at games, will therefore be made, and the impact of play on early childhood socialization will be discussed. Finally, data drawn from naturalistic studies of young children will be offered in support of the perspective presented.

Existing Theories of Play and Games. In an excellent review of play and games, Ellis (1973; also see Gilmore, 1971) lists some thirteen different definitions and theories of play and games. These include the surplus energy theory of Groos (1901), the play-as-relaxation theory of Sapora and Mitchell (1961), the instinct theory of McDougall (1923), the recapitulation theory of Gulick (1902), the psychoanalytic theory of Freud ([1922] 1959) and Erikson (1950), and, more recently, the cognitive-developmental theories of Piaget (1962). Underlying each of these separate formulations are such arguments as that people play because they have too much or too little energy to expend; that they play because they have inherited instincts that make them play; that in play the player recapitulates the history of his species; that in play people play out crises or reverse social roles, so as to resolve unpleasant experiences. In play, people thus confront or avoid problems, produce or lose energy. Ellis summarizes his review by arguing that no existing definition of play or games is workable. He proposes to define play as arousal-seeking activity, yet admits that his definition is too broad. He concludes by suggesting that it is perhaps impossible to separate play from games and work. His review provides a convenient point of departure, for it points to the current state of conceptual confusion in this area. I turn next to perhaps the two most influential theorists on play and games: Caillois and Piaget. (I have cited these several definitions of play so as to establish this state of conceptual confusion. I have not analyzed or discussed these formulations in detail, for my main concern is to establish an interactionist framework for the study of play and games in early childhood. My remarks are also intended to point in the direction of a more general theory of games and sports. The interested reader should consult Ellis, 1973, for his systematic critique and review of these other perspectives.)

In all of the existing formulations of play, the following attributes are commonly ascribed to this interactional form: It is seen as an end in itself, it is spontaneous, it is an activity pursued for pleasure, it has a relative lack of organization, it is characterized by an absence of conflict, and it produces no economic gain or loss for its players (see Berlyne, 1969; Piaget, 1962; Caillois, 1961; Huizinga, 1939). So play is an autotelic or pleasurable and self-contained form of activity. Yet exceptions to each of these criteria can be found. Some play is not spontaneous, some play is pursued so as to achieve an end beyond the playing episode, and some forms of play are highly organized. Also, many playing episodes display high degrees of conflict, emotion, and tension, and some players (for example, gamblers) do lose commodities of some form or another during their playing activities.

It is problems of this order that led Caillois (1961, pp. 9–10) to define play as an activity that incorporates the following essential elements. First, it is free—playing is not obligatory; if it were, it would at once lose its attractive and joyous quality as diversion. Second, play is separate; that is, it is circumscribed within limits of time and space and is fixed in advance. Third, play is uncertain. Its course cannot be predetermined nor its result attained beforehand, and some latitude for innovations is left to the player's initiative. Fourth, it is unproductive, in that it creates neither goods, nor wealth, nor new elements of any kind. Except for the exchange of property among the players, play ends in a situation identical to that prevailing at the beginning of the game. Fifth, play is governed by rules. These rules operate under conditions that suspend ordinary laws and for the moment make new legislation, which alone counts. Sixth, play incorporates an element of make-believe, accompanied by a special awareness of a second reality or of a free unreality, as against real life. Caillois remarks (1961, p. 10) that “These diverse qualities are purely formal. They do not prejudge the content of games.”

Herein lies the problem, for Caillois has confused play with games. He begins by listing the attributes of play and concludes by offering a now-familiar typology of games: games of competition, change, simulation, and vertigo. These four game forms are analyzed along a continuum of “structuredness”: from paidia (unstructured) to ludus (highly regulated). It is not clear, then, whether Caillois offers a theory of play, or a theory of games, or a typology of neither one.

Caillois’ failure to separate play from games is also evident in Piaget's highly original formulations, which postulate a stagelike, cognitive-developmental model of childhood play. He begins by assuming that the thought processes of the young child move between the polar extremes of assimilation and accommodation. He also establishes four major phases in the development of the child: sensorimotor (birth to two years), preoperational (two to seven years), concrete operational (seven to eleven years), and formal operational. With increasing age, the child's cognitive abilities increase, and the preceding schemata of the stages is reordered. The processes of assimilation and accommodation underlie the child's intellectual development. In assimilative thought, children impose their own reality on the physical world. They do not adapt to it but force reality to fit their limited cognitive conceptualizations of it. In the process of accommodation, children alter their cognitive structures to meet the obdurateness of reality. Piaget defines intelligence as the primacy of accommodation over assimilation. He defines play as assimilative behavior, and he distinguishes three forms of play: sensorimotor practice games, symbolic games, and games with rules. Games with rules, Piaget argues (1962), rarely appear before the ages of four to seven, and they belong mainly in the third period of development (see Flavell, 1963; Ellis, 1973). Like Caillois, Piaget fails to separate play from games. Thus he confuses what occurs in a game, which is properly termed playing at a game, with what occurs when a child “pretends to eat a green leaf and calls it spinach” (1962), which should be termed playing at play.

Piaget's model has been criticized on several scores. His data are primarily drawn from the behaviors of western European children playing during their free time at school. His model gives no room for games of chance (Caillois, 1961). He treats play as an aberration on the thought processes and sees in it none of the elaborate pretense elements Stone (1965) and others have observed in the play of little children (Sutton-Smith, 1966). His model of cognitive development assumes that intelligence and thought are relatively unresponsive to the influences of social context and social interaction. He postulates that social behavior originates in the organism (thus he can speak of a movement from autism to egocentricism to sociocentricism). He views the child under the age of seven as an egocentric interactant who is unable to fully take the views and attitudes of others (Piaget, 1926). It is not until after the child reaches the age of seven that Piaget views socialized thought and action as appearing. His choice of the term socialized is unfortunate, and, as Markey (1928) observes, it implies that social behavior does not appear prior to the age of seven. This clearly prejudges the cognitive and interactive skills of the young child and overlooks the fact that the young child is, in a fundamental sense, the center of his own universe of discourse. That he occupies such a position in no way implies that he is unable to take the positions and attitudes of others. The observations that are reported hereafter reveal young children to be differentially skilled interactants. By neglecting these qualities, Piaget denies the symbolic interactionist view of mind and self, for it is the social context of socialization experiences that shapes the thought processes of the young child (see Vygotsky, 1962; Mead, 1934).

This brief sketch of existing theories of play and games serves to support the following conclusions: With the exceptions of Huizinga, Caillois, Mead, and recent symbolic interactionists, students of play and games have tended to divorce these interactional forms from the interpersonal contexts that produce them. Thus the majority of existing formulations are context-free typologies, divorcing play and games from social setting. Furthermore, they seldom take account of the player's perspective in the playing or gaming episode, young child or otherwise. We have typologies of games, but few theorists make a distinction between play, work, and games. (The major exception is Huizinga, 1939.) Too often games are confused with play, and play is confused with games. There exists in these current theories no firm discussion of the uses to which persons put play and games. There is little treatment of the implications that play and games carry for their players. The data base of existing studies tends toward the study of formally structured games observed in laboratories or child development settings. With the exception of the Opies’ (1969) large catalogue of games in Great Britain, few students have recorded and analyzed the spontaneous play and games of young children. The language or medium through which the gaming encounter is produced and presented seldom receives empirical attention. The verbal and nonverbal gestures that go into the focused action of play and games have seldom been studied. In summary, existing theory gives no firm interactional basis for a comparative study of these social forms.

Elements of a Theory of Play, Games, and Interaction

In their decontextualization of games and play, contemporary theorists have failed to grasp the fundamental fact that all instances of play and games involve the interactions (which may not be face to face) of one or more persons who are orienting their cognitive, physical, and symbolic behaviors toward themselves, one another, or animate and inanimate objects (the case of solitary play with animate and inanimate objects was brought to my attention by two anonymous reviewers). There are forms of play in which the child does not directly and physically interact with another person. They may, however, symbolically mock or imitate that individual, perhaps taking on the mannerisms of a mother, father, or hero or idol. Yet the physical absence of the other does not stop the child from bringing elements of that person's line of action into play. In play that is not face to face, the child imaginatively acts out the attitude of the other. In face-to-face play, of course, the child reacts directly to the actions and utterances of the other. A definition of play, then, must be capable of handling the solitary player. Similarly, such a definition must include playful and playlike actions with animate and inanimate objects that the child has “personalized”—perhaps given a name to or assigned a role to in a game. The child transforms the object into a playmate and may or may not anthropomorphize that object in the process.

Accordingly, a theory of play and games must rest on a consistent image of the interaction process, and it must address the place of persons and situations in that interaction. Also, as the foregoing suggests, it must treat solitary as well as joint players, and the objects of play (whether animate or inanimate) must be considered.

Interaction (whether playful or worklike) occurs in place (Goffman, 1971). Place consists of the following interrelated elements: differentially self-reflexive actors; place or setting itself (that is, the physical territory); social objects that fill the setting and are acted on by the actors in question; a set of rules of a civil-legal, polite-ceremonial, and relationally specific nature that explicitly or tacitly guides and shapes interaction; a set of relationships that binds the interactants to one another; and a shifting set of definitions reflective of each actor's co-ordination to self and other during the interaction sequence. The total amount of time two or more actors spend in place shall be termed the occasion of interaction. Every focused exchange between those actors shall be termed an encounter. Thus, places furnish the occasions for interaction, which in turn provides the conditions for encounters (see Goffman, 1961a).

Any interactional episode between one or more persons can be studied in terms of the dimensions of place that are acted on and played out. Every instance of focused interaction displays a unique configuration of these six elements of place. Each element can assume different parameters over time, within and across the same encounter. The players can stand in varying degrees of “knowingness” to one another. The relationships that bind them together will be of shifting equality, moving between the extremes of subordinate and superordinate, and they may be of the intimate, stranger, friend, colleague, or enemy variety (see Davis, 1973). The setting may be rigid or flexible, known or foreign, public or private, rented or owned. The objects may or may not be taken as concrete extensions of the actors’ selves. The objects may be owned, rented, borrowed, sacred, or irrelevant to the action at hand. The rules that shape interaction will be differentially flexible, and they will vary in their bindingness on the actors. They will be differentially reciprocal. The degree to which rule violation influences the interaction process will be specific to each place, occasion, and encounter.

The action that occurs within place will be of shifting temporal length, differentially reciprocal. The degree to which rule violation influences the interaction process will be specific to each place, occasion, and encounter.

The interaction that occurs within place will be differentially focused or unfocused. Its biographical implications for the players will also vary. Some episodes carry no implications, while others (turning point encounters) are heavily consequential (Strauss, 1969). The degree of seriousness or “pretense” that hovers over place, occasion, and encounter will be variable (Glaser and Strauss, 1964; Stone, 1965). The players may be seriously embedded in the action or take it lightheartedly. They may pretend to be serious or seriously pretend not to be serious.

Play and Games Defined. Play, work, and games are interactional forms: They are situated productions that differentially mold the elements of acts, interaction, and place that have been outlined. Definitions are in order.

Play and games, as Caillois noted, exist along a continuum of interactional structuredness. A game shall be defined as an interactional activity of a competitive or cooperative nature involving one or more players who play by a set of rules that define the content of the game. Skill and chance are the essential elements that are played over. Played for the amusement of the players and perhaps also for a set of spectators, games are focused around rules that determine the role that skill and chance will have. The objects of action are predetermined (checkers, kites, chess players, dice). Furthermore, the elements of pretense are known beforehand and are specified by the rules of conduct deemed appropriate to the game at hand. These rules specify the number of players, tell the players how to play, and determine how they will relate to one another as game players. Typically these rules hold constant outside game relationships and make each player a co-equal (subject to the variations of skill and chance) during the game itself. (However, some games deliberately build on inequality; they assign in sequence, or by chance, undesirable roles to their players—for example, “wolf-wolf,” “hide-and-seek.”) And of course, many games have winners and losers.

Yet location in those roles is determined by the game, not by relationships that exist between the players when they are not gaming (see I. Opie and P. Opie, 1969). While the factor of pretense is known before the game is entered, the players may exaggerate pretense during the game—for example, players of monopoly, poker, or “button-button.” In this sense, emotion is central to the game, and players may enter the game precisely because its outcome is problematic and emotional, if not pretenselike in nature.

The temporal duration of the game is typically not known; only its outcome is known—it will produce a winner and a loser. Thus the length of a game will be determined solely by how long the players take to play it and by how long they are willing to play. Some games are suspended, broken productions that never reach a rule-determined conclusion. Rule violation is crucial to the game, and all games have players who cheat, or break, or distort the rules. (Highly structured games—for example, professional football—have their own judges or referees who determine when a rule has been broken. Less structured games rely on their players to be one another's judge—and jury.) On rule violation, Caillois (1961, p. 7) has observed: “The confused and intricate laws of ordinary life are replaced in this fixed space and for this given time by precise, arbitrary, unexceptionable rules that must be accepted as such and that govern the correct playing of the game. If the cheat violates the rules, he at least pretends to respect them…. one must agree with the writers who have stressed the fact that the cheat's dishonesty does not destroy the game. The game is ruined by the nihilist who denounces the rules as absurd and conventional, who refuses to play because the game is meaningless. His arguments are irrefutable. The game has no other but an intrinsic meaning. That is why its rules are imperative and absolute, beyond discussion.”

The biographical consequences of any given game, for any player, will vary by the nature of the game and by the player's attitude toward what the game furnishes him or her. The seemingly inconsequential nature of games can establish for the player a sense of self not otherwise attainable in daily life. Goffman (1967, pp. 269-270) observes: “When persons go to where the action is, they often go to a place where there is an increase, not in the chances taken, but in the chances that they will be obliged to take chances. … On the arcade strips of urban settlements and summer resorts, scenes are available for hire where the customer can be a star performer in gambles enlivened by being slightly consequential. Here a person concurrently without social connections can insert coins in skill machines to demonstrate to the other machines that he has socially approved qualities of character. These naked little spasms of the self occur at the end of the world, but there at the end is action and character.”

The place or places of the games are differentially fixed. Some games can be moved from place to place, as in checkers or chess or hide-and-seek. However, once the place of the game has been set, it is typically seen as “playable” only in that place, at least for its current duration. It may be played elsewhere on another occasion. Other games require a fixed place for play that is borrowed or loaned, rented, or owned. Such places require that players come to them. The more formalized the game, the more likely it will be fixed by place or by classes of place. And the more fixed the place, the more likely that the game is played for the amusement of an audience as well as for the players themselves. In either case (movable-negotiated place or fixed place), games can be seen as relatively intransient productions. (This point will be elaborated later, when it is shown that play—? as an interactional form—is relatively impervious to the needs of a fixed place.) During the production of the game, players are granted a “place legitimacy” that is often not granted the player of play.

In the present context, play shall be given two meanings. On the one hand, it will describe the activities that occur during the gaming encounter. Persons as players play games; they are playing at a game. However, some persons play at games, but are not playing a game. Witness the following episode: Three daughters—ages six years, four months (Sara), six years, one month (Jody), five years, one month (Ramona)—are seated in the center of a family living room.

Jody says to Sara and Ramona:

“Sara and Ramona, let's play cards. I'll deal.”

Sara:

“Okay, but do I have to sit next to the fireplace?”

Ramona:

“I don't want the joker.”

(Jody sorts the cards into three equal stacks, and play begins. Each player has to get an ace to win a hand.)

Jody:

“I've just changed the rules; if you get a pair, you win.”

Sara and Ramona:

“No, no [in unison]. That's not fair.”

Jody:

“Okay, Sara, want to have an ace?”

Sara:

“No. Let's play by your rules.”

(The game goes to conclusion. Each player wins one hand. The cards are stacked up, and the girls leave to watch television in another room of the house.)

In other instances, these three players have played bridge and poker, slightly altering the rules on each playing occasion. Here the players have a sense of a game, have a knowledge of playing in turn, can understand the concept of the deal and the shuffle, but are not playing a specific game of cards. They are “playing at playing at a game of cards,” playing within a constructed pretense context that permits them to be something other than what they ordinarily are. (Further exceptions to our formulations can be noted. A player can play with himself and focus the effects of play solely on himself. And many players play with no sense of a game in mind.)

When persons engage in the production of a pretense awareness context that is not framed by the specific rules of a game, they shall be said to be “playing at play.” Playing at play, unlike playing at a game, involves the use of flexible rules, and it is much less tied to any specific place. Playing at playing involves players engaged in a “free-floating” set of interactions that takes them from place to place as the focus, needs, and demands of whatever is being played shift and take on new forms.

Consider the following interaction, based on thirty minutes of observation. Jody and Ramona and Sara enter the kitchen. To Jody and Ramona's mother (Emily), they ask:

Jody, Ramona, Sara:

“May we have juice and crackers?”

Emily:

“Yes, wash your hands first.” (The three comply and then go to the kitchen table.)

Sara:

“Let's play we're princesses. Jody, you be the king. Ramona, you be the queen.”

Ramona and Jody:

“Alright.”

Ramona:

“Let's play the king died last night. Oh dear, your husband is dead.”

Jody:

“What shall we do? Let's go on a walk through the forest and look for another king.”

(The three leave the kitchen, enter the dining room, and crawl under the dining room table.)

Sara:

“King, king, where are you?” (She moves a chair and places a stuffed dog in front of her.) “Bow-wow, King, King, Bow-wow.”

Ramona:

“Oh no! Your doggie has to go to the bathroom.” (The three exit and go upstairs to the bathroom. They come back downstairs.) To Emily, they state: “The king is dead, and we found him.”

Jody:

“Come, Madame, we will go to the desert island.” (They move to a green circular rug near the front door of the living room.)

In unison:

“Oh no, the crocodiles are out again; quick, get on the island.”

Sara:

“Bow-wow, my dog will kill them.”

Ramona:

“Oh boy! Thank you.”

Three different playing at play behaviors are displayed in this specimen, and they move from setting to setting, irrespective of the actual place of play. At the kitchen table, over juice and crackers, they moved into playing princess. The dining room table saw the dog enter the picture. The upstairs bathroom permitted them to elaborate on the playing of princess, and finally the dog saved them from the crocodiles who were attacking them on the desert island. In this sense, players of play can pick up their play and play it anywhere. (Some forms of playing at play do become place specific. In many nursery schools, the play involving mother, daddy, and baby often attaches itself to the school's playhouse. Construction play similarly becomes attached to the block corner or the sand box. This attachment is typically encouraged by the designers, owners, and managers of these establishments. The child players are taught to attach their various forms of play to these settings. This often gives rise to place invasions, and battles over territory ensue. See Joffe, 1973; Suttles, 1968; and Thrasher, [1927] 1963, on play and territorial invasions.)

Play is an expansive and expandable interactional form. It is not tied to the demand of time or place. Unlike games, which have concrete rules specifying who may play and how many players there will be, the number of players who can play is limited by the number of persons present and by the relationships the players have with one another. Often these relationships stand outside the moment of play itself, unlike the game as noted earlier.

The following specimen is drawn from a West Coast university preschool. Female Cindy (three years, two months), male Wally (three years, six months), and John (three years, one month) are making pumpkin muffins the day before Halloween. They are seated around a table mixing their batter and filling their baking cups.

John:

“My mommy has a grandma, and she lives in L.A., and her name is Grandma Kaye.”

Wally:

“I have another grandmother, and she lives in New York.”

(Two children pass by the door to the kitchen.)

In unison:

“Hi, Mary; hi, Gary, come on in,”

Mary:

“Hi, John; hi, Wally, hi Cindy.” (Each child says hi to every other child.)

Wally:

“Hi, John.” (He says this to Cindy. He then proceeds to alter the name of each child around the table, calling each one by the name of the child seated next to them.)

John:

“Hey, that's a mistake, I'm not Cindy!”

Cindy:

“We're just pretending, silly, we don't mean it.” (They then go through name alterations for all of the children sitting around the table. The number has now climbed to six.)

Two points are relevant in this specimen. First, making pumpkin muffins, a form of work and play, provides the children with an opportunity for focused interaction. Thus making pumpkin muffins is an ancillary activity, and telling about grandmothers is more important. Notice there is nothing about making pumpkin muffins that would necessarily lead one to talk about grandmothers. (This point will be developed later.) Second, as indicated earlier, play is an “incorporating” activity—it expands to fit the number of possible players who are present. Thus, three persons are added to the gathering, and an emergent form of play is produced—the name alteration game (see Wolfenstein, 1954, on name games). It must be noted, however, that play can be an exclusionary form of activity, as when two seven-year-olds refuse to play with three four-year-olds. This also sets play apart from games, for once a game is started its addition or exclusion (loss) of players is predetermined.

Pretense is central to play, more so than in games. Kendon (1930, pp. 57–58) in his autobiography, comments on pretending: “Pretending is a child's art alone, and grown-ups cannot practice it; they can only act. The pretender has no audience but himself, he is not trying to appear like the character he presents, he is that character and feels like it. If he has put a towel around the muzzle of a rocking horse and is driving through rain and storm to catch a train, his feelings will be very nearly the real feelings which he would enjoy if the horse were a real horse, the towel a real bridle, the rain wet, and the storm loud.” Play permits the child to symbolically transform reality into new forms and into new configurations. The pretense nature of play rests on a set of tacitly known rules that permit objects to become anything the player wants them to be. The object is what it is during the moment of play. When play is over, the object may be returned to its rightful place, but during play it is what the player says it is. The young child has yet to assign a fixed character or meaning to these objects of play. A towel can become a bridle, a stuffed dog a real dog, a dining room table a cave, scrawls on a page can depict an elegant castle, a female can become a male or a king, a chair can become a bed. Isaacs (1933, p. 35) reports the following behavior of Frank (age four years, eleven months): “Frank was making a ‘bed’ with chairs when Harold took one of the chairs. Frank threw a cylinder at Harold, which Harold caught and threw back at Frank. Frank was going to throw it again, when Mrs. I intervened. He protested, saying, It didn't hit him!’ as if that gave him the right to throw it.”

The failure to support and validate another player's definition of an object, especially when the object is a scarce resource, provokes object battles, and these battles are often seen by their players as another form of play. While games have object-transference rules (for example, checkers can be played with bottle caps, Goffman, 1961a), these rules are fixed before entry into the games. In play, the rules are made up as play is produced. They may assume a fixed form over time, especially if the same players play with the same objects again and again.

Rule violation, which focuses on one or more players’ refusal to treat an object(s) in a desired way, potentially destroys play. Like the game, play cannot tolerate the nihilist; it is, however, much more tolerant of rule violations.

Because play is not tied to specific places, it is often performed in the unwilling presence of others. And, while the player typically plays only for his and his fellow player's benefit, his play is often seen as an intrusion into the worlds of others. Persons come to see players play games, but seldom do they come to see persons play at play. These actions are regarded as intrusions. The child is told to go play but is admonished to play elsewhere.

Ramona and Jody are playing on the living room cocktail table. Ramona has a Barbie doll, who is talking to her rabbit, who is named Bunny. Jody is counting doubloons and is dropping them on the table. They are watching a television show that the parents Emily (mother) and Nick (father)—Nick in particular—are engrossed in. Ramona complains, “No! Jody, you hit my doll.” Jody denies it. “I did not. Get out of my way.” Ramona appeals to her mother: “Mommy, Jody hit my doll!” Nick interjects, “Go upstairs and play!”

The biographical consequences of play are more difficult to assess. They are not predetermined, as they are in games, but refusals to play with a particular player can be seen as especially devastating, at least by the denied player. He has not been permitted to enter the pretense world of the other players. Thus, it is often regarded (by the child player at least) as a privilege to play, a privilege that carries biographical implications into the future. One knows that if one has played in the past with a certain player the likelihood of playing together in the future is increased. In this way, players become friends. The following excerpt from an interview cited in Chapter Four, with Fred (age four years, seven months), supports this conclusion.

Interviewer:

“Which guys do you know here?”

Fred:

“Well, I don't know all the guys, Bob, Chuck, Eddie, Andy, Tom, Sam.”

Interviewer:

“Who are your best friends? Who do you know best?”

Fred:

“Chuck and Eddie.”

Interviewer:

“Why Eddie?”

Fred:

“He said one day, ‘If you let me play with one of these puppets, I'll be your best friend.’ If he wants to play with me, he can.”

Interviewer:

“What's a best friend?”

Fred:

“He wants to be a friend real bad. But you see it's in his head why he wants to be my best friend, and it's in my head, and I can't know what he feels and thinks. I don't know. He wants to be friends real badly. It's in my head. We just play together.”

The matters of skill and chance are less relevant in play. Indeed, unlike games, which are built around the role of skill and chance, playing at play often reveals the players cooperating with one another. Willy (four years, three months, male) and Bonnie (three years, eleven months, female) are working with puzzles. Willy's mother is observing them helping one another. The mother says, “Oh, good, Willy, you need only one more piece.” Bonnie repeats, “Willy only needs one more piece.” The mother says to Bonnie, “Yeah, you need a whole bunch more, better get busy.” In this instance, each player has a puzzle. Initially, they cooperate in this individual activity by helping one another, so it becomes a joint endeavor, to end when both of them have succeeded in completing their respective puzzles. But now the mother has introduced elements of adult “gamelike” competition into their play.

If competition and skill and chance are kept at a minimum in the child's game, boasting about one's abilities is often taken quite seriously. While players of games can boast after a game is over, they are under constraints not to do so during the playing of the game, because one play does not constitute a game and one who boasts too early may later be the loser. Since play seldom establishes winners and losers (childhood play, at least), boasting and bragging about one's abilities take on central import—the player seeks praise for his or her accomplishments, abilities, and pretensions. Three males (Andy, three years, eight months; Bob, three years, ten months; Ken, four years) are sitting in the corner of the playroom in the preschool described earlier. They are assembling small pieces of plastic into ships and airplanes. Andy starts:

Andy:

“I'm Superman.” (He was wearing a Batman tee-shirt.)

Bob:

“I'm strong man, I'm so strong I can blow the whole world into pieces.”

Ken:

“I'm so strong I can blow the United States into pieces.”

Andy:

“Oh yeah?”

Bob:

“Oh yeah?”

Ken:

“Yeah!”

Three forms of play have been distinguished and they range along a continuum of rule-embeddedness, or degree of structure and formality. Playing at play describes the free-floating interactions of the young child who moves from place to place, playing at whatever concerns him at the moment. Playing at a game describes the young card player who draws and deals and forms pairs, yet has no firm knowledge of how a specific game of cards is played. A good deal of the play of young children from the ages of three to seven falls in this category. Playing a game describes the skilled game player who works within the rule-constructed boundaries of a specific game wherein matters of skill and chance are rigidly controlled.

If our earlier statements concerning solitary and joint play are reconsidered, six types of play are actually produced. Thus, the child players can play at play alone, as when they stage a conversation with a stuffed bear, or they can conjointly produce a conversation with another player who has a stuffed animal. Similar considerations hold for playing at a game and playing a game. Prior typologies of games, in particular Piaget's, Huizinga's, Caillois's and Mead's, have confused these play forms, and in so doing they have failed to dissect the unique configurations of place and action that occur within each of them. They also prejudged or underestimated the interactional skills of the young child player of play and games.

Further complications are introduced when the topic of work is considered. It is necessary to introduce a conceptual discussion of work at this point for, as noted earlier, existing theories of play and games typically begin by first defining work (see Caillois, 1961). That is, work is seen as a form of activity that can be neatly separated from play and games. However, such a separation seldom stands the test of critical analysis, as I hope to show in the following discussion. This is especially the case if work is considered in processual, not static terms. Like play and games, work is a situated production, and it may be approached from a playful or gamelike perspective. Indeed, one person's work is another person's play. And it often takes a class of workers to produce for other persons the context of play. Furthermore, the moment of play can be transformed into the actions of work, just as work can be transformed into play (Bernstein, 1973).

Like play and games, work is a situated production that involves the six elements of place. Unlike play, it is situationally fixed (although some workers can do their work wherever they are—for example, painters and novelists). The rules of work, as in games, are known beforehand. Work involves the transformation of objects from one state to another—ideas into printed words, images into pictures, steel into nails. Games and play are different in this respect, for when play or the game is over players typically return to their preplay and pregame conditions. Thus, play and games seldom produce lasting economic effects or rewards. And they are less likely to affect the biographies of their players, while work does. The persons that one works with assume a fixedness that need not be present with the players of games. Players of play tend to establish fixed playmates. Work builds up a set of “working relationships” that are differentially rule grounded—subordinate, superordinate, equals, colleagues, buyers, and sellers. These relationships are assumed to hold during the work transaction. We have shown that this need not be the case with play, and games work on the deliberate suspension of existing relational ties. Work is entered into with a known and often predetermined degree of pretense and illusion. This is not to say that the worker cannot approach work as an autotelic act (as the player does), but he is less likely to sustain the autotelic spirit for the duration of any working occasion. Work is structured so as to attempt to control the play of skill and chance. While workers are ranked by skill, it is assumed that the degree of chance that enters into the ranking is kept to a minimum.

Play, Work, and Games: Process. So far I have regarded play, games, and work as more or less static interactional forms. In this section, I treat them as processual forms, and use the verb, not the noun form, for each word. There are six interrelationships between play, work, and games. Persons can work at work, work at playing, or work at producing a game. They can play at work—the young child helping a mother fix a salad or the university employee who plays at being a faculty member—or they can work at work. They can play at play—our young children discussed earlier. They can play at games, and they can actually play games (Caillois’ and Huizinga's typologies only fit persons who play games). Each of these conditions constitute different variations on place. Each is a situated production that will vary by the attitudes brought to it by the persons or players in question. And each varies by the self-other relationships in question. Furthermore, any occasion of working at work, working at play, working at a game, playing at work, playing at play, and playing a game can be transformed—during its production—into any of the remaining other forms. A game can be approached with enthusiasm, as can play, but during the interaction one or more players can alter their definitions of the situation. The effects of the game, or the play on them, is suddenly (or gradually) defined as undesirable. The consequence is a bad game or a bad instance of play.

Play, work, and games must be approached historically. That is, they must be viewed as temporal productions that have a life within the lifetime of their participants. The temporal dimension of these interactional forms are especially crucial when the topic of interest is the socialization experiences of young children. With increased interactional skills, they become sophisticated players of games and work. Indeed, it is not uncommon to observe the two-, three-, and four-year-old child engaged in the work of the adult (see Kuhn, 1954, on the Iowa Amish; this social collectivity sets its children to work on adult chores at the age of two). The matter of history raises three interrelated questions: What is the actual chronological or biological age of the player? What is the actual age of the interactional form the player works or plays at? and What is the interactional age of the player? Piaget's (1962 and elsewhere) formulations of child's play treat only the biological age of the player. Thus, with conviction he argues that only after the age of seven does the child know how to play rule-bounded games—for example, marbles. In no instance does Piaget consider the interactional age of the play form, nor does he trace out the interactional experiences the player has with that play form. Therefore, his model is irrevocably embedded in a biological image of cognitive development. Interactional age references the actual amount of time the player has spent in the play or work form under consideration.

Interactional age is not an age per se, for it is not grounded in the usual psychological conceptions of physical or mental age (see Markey, 1928). Rather, it is based on the number of interactions the child (or adult) has had with a specific unit of behavior. Thus, it is quite easily measured, although measurement is confounded by the fact that the person can have symbolic interactions with the behavior in question while never displaying physical actions toward it. This, in a qualified sense, would reference what Merton (1957, p. 265) terms “anticipatory socialization.” In this sense, it might be more appropriate to label this concept interactional experience. It is introduced at this point so as to direct the researcher's attention away from a singular concern with just physical or mental age. Such concerns have led investigators away from the actual study of social interaction in early child development. As a consequence, they have not adopted a historical view of the interactional process. Nor have they considered how that process is built up and incorporated in the behavioral repertoires of the child. A rather blinded view of child development has emerged in the social and psychological literature as a result of this neglect. Mead's (1934) model of the play, game, and generalized other phases of development suggests that interactional age is crucial in the socialization process. Furthermore, nowhere does Mead attach age specifications to these three phases. Implicit in his formulations is the suggestion that some persons may never progress to the generalized other phase of taking the other's attitude. Such a view is accepted in this discussion.

Accordingly, a child's expertise in playing or gaming would be expected to be a direct function of the actual amount of time that child had spent in that play form. Thus, the longer our three card players play at playing at cards, the better skilled they will become. Consequently, if biological age is held constant and if interactional age or experience is allowed to vary, it would be expected that a biologically younger child might show greater play skills than would a biologically older child who had a low interactional age with the play form. Unfortunately, existing data on player age seldom take account of interactional age (see, for example, Herron and Sutton-Smith, 1971). Some studies have considered the interactional age of games (for example, I. Opie and P. Opie, 1969), yet they fail to connect those ages with the interactional ages of the players. As a consequence, there are two bodies of data that exist side by side, but the crucial variable that would bring them together into some meaningful interpretative network is pitifully absent. By way of illustration, consider the following example of playing at a game. The players are the three girls (Jody, Ramona, and Sara) discussed earlier and an eight-year-old boy named David. The game that is played at is termed “animals,” and it appeared in the behavioral repertoires of the three girls on March 10, 1972. At that time, the players were approximately four, four, and three years of age, respectively. David was introduced to the game fourteen months ago. At that time, the following behavior specimen was recorded. The game consists of each child taking on an animal identity (Hon, tiger, and bear) and then crawling through the house as quietly as possible so as to come up behind an adult and then roar as loudly as possible. If any player is seen by an adult before the other two players are in a position to roar, she is chastized by the other playmates. At issue, then, is quietness, and the aim of the game is to provoke a startled reaction from the adults. This play form can be repeated over and over again. The act began as follows:

Jody:

“Come on, Ramona and Sara. Let's playanimals. David, do you want to play?”

David:

“Sure, what do I do?”

Sara:

“You can be the big lion. Get down on thefloor. You can follow Jody.”

Ramona:

“You have to be real quiet. Wait until wegrowl.”

At this point the players began to work their way down the stairs. They entered the living room. Emily, the mother, was in the kitchen. They moved across the living room. Jody accidentally coughed. David roared the sound of a lion. Jody and Ramona and Sara turned on him.

Jody, Ramona, Sara:

“David, you ruined it. You weren't supposed to roar until Jody did. You ruined it!”

David:

“I'm sorry, can we start over again?”

This is not a particularly elaborate example, and unfortunately David's failure to keep quiet was based more on accident, as on a failure to understand fully the rules of the play form.

The following episode is more to the point. A birthday party has been given for Ramona on the occasion of her fifth birthday. Present are Jody, Sara, David, and Linda and Larry (ages ten and eleven) and Elsa (age six). The game “button-button, who-has-the-button?” is proposed. Elsa has never played the game. Linda and Larry coach her on the rules. She says she understands. On the first play of the game, when the player (David) in the center yelled “stop!” Elsa, who had the button, stuck her hand out and gave David the button—a clear violation of the rules. Play was started again, and this time Linda was in the center. The button ended up in David's hand, and Jody knew it. As Linda looked around the circle, Jody smiled and giggled, as did Larry. Linda pointed to Jody, who exposed two empty palms. Next Linda pointed to Larry; she was fooled a second time, and as a consequence lost the game. Here the ability of the player to fool another player is seen as a function of past experience with the game and with face-to-face interaction itself. Linda argued that because she had not played the game since the last birthday party she should be permitted another turn.

These two episodes point in the direction of needed research on children's play and games. Until researchers have combined the conceptions of interactional age just outlined, they will be in no position to make arguments connecting biological age to rate of cognitive and moral development (see Kohlberg's 1968 efforts in this direction).

One final issue on age needs to be considered. The longer a play form is played, the more gamelike it becomes. The child player, then, is seen as moving along a continuum of play complexity; yet this continuum must be viewed in multidimensional terms. That is, the child's first form of play will be of the elementary sensorimotor nature observed by Piaget. Here the child manipulates its own body and surrounding objects in its immediate environment. It talks to play dogs, to favorite blankets, yet it displays no complex mix of rules concerning how the dog or the blanket is to behave. Soon thereafter, with the acquisition of more sophisticated linguistic skills—the elaboration of inner speech into longer thought episodes—he or she will be observed playing not only at play but at games as well (see Vygotsky, 1962). The second form of play will be playing at play and playing at games. The final form will be playing games. Yet, at any point after the age of three, if the child has had sufficient interactional experience, he or she can engage in any of these three play forms. Consequently, depending on the child's experience with the form under consideration (that is, the amount of time spent with it) and depending on the age of the form, he or she can display sophisticated or ill worked-out actions toward that form. Thus, the players of any game or play form will display differential role-taking skills that are directly related to their amount of familiarity with that form. The critical variable in this model is interactional experience, not biological age or level of cognitive development. A thirty-year-old male with a Ph.D. in sociology may be a quite inept player of a play form developed and played over and over again by three children ages four, five, and six.

One final topic remains to be treated. This involves the symbolic and interactional effects of play and games on the young child's socialization experiences. In this respect, I wish to discuss the attitudes of the young player toward play and games. Here work is of utmost relevance, for players can approach play and games with a studied seriousness that suggests they are actually working at play. Thus, play as a quasi-autotelic interactional form transcends itself and provides the contexts for working out such important matters as pride, deceit, self-respect, and friendship formation. Paradoxically, then, when players enter the pretense world of play, they enter a world that allows them to become what they will become in everyday, “real” life: differentially skilled participants in the business of face-to-face interaction. By learning to play, young children learn what elements make up place. They learn to attach different meanings and interpretations to self, other, and object; and to take the point of view of civil-legal, polite-ceremonial, and relationally specific rules. They learn how to form, break, and challenge social relationships; how to measure time and its passage; and how to assume (or avoid) the biographical consequences of any set of actions. The player at play is seen as acquiring the skills requisite for future moments of focused interaction. Playing at play, playing at games, and playing games give child players the interactional experience of taking one another's perspective in concrete, fpgused interpersonal exchanges. In this sense, play as an autotelic interaction form transcends itself, for play becomes life and life becomes play.

What play does for the child player is to give him or her interactional experiences. Socialization quite simply turns on such experiences.

In the present context, socialization represents the progressively cultivated ability of the individual to take and act on the attitudes or lines of action of others. In the words of Simmel (1921, p. 349), “Socialization is thus the form, actualizing itself in countless various types, in which the individuals—on the basis of those interests, sensous or ideal, momentary or permanent, conscious or unconscious, causally driving or purposefully leading—grow together into a unity and within which these interests come to realization.” The present argument suggests that play is but one of many interactional forms crucial to early childhood socialization experiences. While it is clear that Simmel's view of this concept is no longer in prominent use in the social sciences (see Parsons and Shills, 1959; Clausen, 1968), it is also evident that any functional and psychological catalogue of play effects will always be lacking. For play certainly teaches roles, clarifies rules, produces and reduces tension, creates and recreates the problematic. It exhausts energy, and it produces energy. Yet, play only sets the contexts of conjoint interaction, and once that context is entered the focus of the child's actions shifts beyond the initial focus of play to other emergent activities. Play, then, is a socially legitimate activity that is seen as appropriate for its players and, as long as it does not splash over into the worlds of the unappreciative audience, the child will be encouraged to play and play and play. Unwittingly, this carefree activity called play constitutes the most important interactional experiences of the young child, for in play he or she becomes a more sophisticated member of the outside social world where work is taken seriously and play is regarded as inconsequential.

Children enter the world with no conception of self, other, object, situation, or place (Lindesmith, Strauss and Denzin, 1978). Out of the symbolic processes of interaction, they soon acquire an ability to take the attitude of the other (Markey, 1928). A matrix of self-other relationships, at first only dimly sensed, make up the child's universe of discourse. Viewed as a semiincompetent actor, the young child is encouraged to play and is typically provided with a stock of objects to play with (Ball, 1967). The adult caretaker teaches the child to play and in doing so produces and provides the contexts for solitary and conjoint play and interaction. In solitary play, the child learns to attach labels and meanings to objects that previously were meaningless. In play, the child brings into his own universe of experience other objects of discourse and in so doing learns to transcend his own immediate physical and symbolic reality. At first the child's ability to take the attitude of the other is limited, and this is at least partially determined by his linguistic skills and by his amount of interactional experience with the object or other in question. The child's early actions thus display what has been termed playing at play. The child is learning how to play. As interactional experience increases, the child moves from the playing-at-play stage to a stage termed playing at games. Now the child is grasping the various elements of place developed in this chapter. He or she has an understanding of rules, of objects, of others, and of himself or herself. Although an understanding of sequence, propriety, and rule violation has been developed, the child has yet to understand firmly all of the elements that go into proper “gamelike” behavior.

As their interactional experience increases, children become more competent game players, and soon (I have presented evidence from three- and four-year-olds) they actually become players of games. At this point, they have advanced to Mead's (1934) conception of the generalized other, for they are now able to call out the collective attitudes of a group of individuals. The concept of interactional age or experience suggests a “multidimensional” approach to the study of child development since a child may be quite sophisticated in one interactional form and quite ill-equipped to handle another. Thus, single, linear models of socialization, or development, gloss over the variously developed interactional skills of the young child.

Charged by their caretakers to play, children often approach play with the studied seriousness of work. Thus, the work of the young child may, on occasion at least, be termed play. It would be erroneous to argue that play is the single necessary condition for the production of self-competent interactants. Other interactional forms of a routine and ritual nature clearly enter into this process. Yet play, in its solitary and conjoint forms, assumes a dominant position in the cycle of activities of the young child.