Abstract
Three children invent a trading game and vanish into it for several days straight. How does such a world come to exist, and why does it matter so much? This article follows children’s play from an infant’s exchange of sounds to the contested worlds of later childhood. Children discover worlds, construct them, and make them matter, and playing generates stakes as well as revealing them. Forms of play accumulate, are reconfigured, fall dormant, or are lost. This is the second part of a three-part series of articles exploring what kinds of worlds are built in playtimes.
I. A World Made Through Playing
A few years ago my family spent two and a half weeks with friends in New Zealand. My children were ten and eight, and our friends’ son was nine. During the visit the three of them developed a trading game. It had money, valuables, rules, bargains, and some kind of running tally of gains, and that is nearly all that the adults ever understood of it. What we could see was that three children were spending whole days inside it. My two had never played like that with another child, and apart from one earlier episode, described below, they have never played like that with each other.
The episode raises three questions that the usual accounts of children’s play do not answer. The first is how such a world comes into existence. Nobody gave the children the game. Its materials, trade, money, the weighing of one thing against another, came from the world around them, but the arrangement was theirs, made during the visit out of what was to hand. The second question is why it mattered so much. Things were valuable inside the game that had no value outside it, and a bargain struck there was plainly a serious matter to the three who struck it. The third is what would have been lost if one of them had walked away. The pretend money would have been no loss to anybody. Something else was at stake, and the adults watching could not have said exactly what.
This article works backwards from that scene. It asks how far the processes involved in building such a world can be traced into earlier childhood, beginning in the first months of life, and what they show about what is at stake for children as they grow. Its hypothesis is that children’s play offers an unusually revealing way to investigate the formation and transformation of their mesocosms, the lived worlds within which they coordinate with their own bodies, with other people, and with places, things, and symbols. From the first exchanges of sound with a caregiver to the imaginary societies, constructions, games, and contests of later childhood, playing makes visible the coordinations that children are becoming able to sustain and the stakes those coordinations carry. Infants belong to a mesocosm from birth. What develops is the differentiation and range of their participation in it and their ability to reorganise it, and play is where much of that participation is first tried.
Two things follow if the hypothesis is right. The worlds of play are neither given to the child by an environment nor invented independently of it. They are established through the uptake, modification, and occasional rejection of arrangements that were there already. And development is no linear succession of types of play, each replacing the last. New forms become possible as capacities and circumstances change, while earlier forms persist, acquire new value, fall dormant, or are lost. The claim towards which the article works is that play does more than make existing stakes observable. By playing, children discover worlds, construct worlds, and make worlds matter. The article ends at the onset of adolescence, with a question it does not try to answer.
II. Play as Mesocosmic Construction
Living Value Theory begins from the principle that life is recursive mediation in a mesocosm. A living being coordinates continuously through its own embodiment, through being-with other living beings, and through dwelling among conditions that nobody made, and human beings coordinate in addition through made things and through symbols, the mediations the theory calls multimateriality and multisymbolization. Counterparts differ in kind. Some do not alter course however they are engaged, as a stone or a rule does not. One’s own body answers to its own ongoing state. Other living beings respond to one’s responses, which the theory calls interrecursivity. Most coordination is muted and proceeds without notice. When its texture changes it becomes salient, and among symbol users it may then be articulated for the occasion, stabilised into portable conventions, and reconsidered. Stakes, finally, are facts about what depends on a coordination: what would be lost, damaged, or made unavailable if it failed.
Two cautions about this architecture govern its use in a developmental argument. The levels of recursivity are analytical distinctions and should not be read as stages with ages attached. For any coordination three questions have to be asked separately: when a child becomes able to take part in it, when she can articulate or reconsider it for herself, and through which levels the wider coordination is already organised by the adults, things, and institutions around her. The answers give three different histories. A baby who is sung to sleep is participating in a coordination that is conventional through and through, long before she could name a lullaby. The same holds for the mediations. An infant’s mesocosm is materially and symbolically mediated from the first day, by clothing, a cot, a bottle, a house, a name, and the speech and routines of those who look after her. Made things and symbols do not arrive in her world at some later date. What changes is her capacity to engage them as differentiated counterparts and to take a growing part in the coordinations through which they are used. Development is accordingly described in what follows as increasingly differentiated participation across mediations and levels that are present from the start. The three types of counterpart are likewise met together and not in sequence: children deal concurrently, from the first months, with what does not answer back, with their own bodies, and with other people, and a good deal of early play consists in finding out which is which.
Play has been characterised as the sought opening of a coordination that was already working, under an allocation of stakes that makes the opening safe enough to explore. For the purposes of this article the characterisation has to be enlarged, because the children in New Zealand were not reopening any one settled activity. They established an arrangement that had not existed in that form. Five processes should be distinguished. In opening, a settled coordination becomes salient and available for variation. In construction, play generates and stabilises an arrangement of coordination that was not previously available. In maintenance, the participants keep the arrangement operative by repeated uptake and adjustment. In transformation, the established arrangement is itself modified. And in closure, the activity ends, goes quiet, leaves deposits, or falls dormant. Construction always draws on existing coordination, so that nothing is made from nothing, and yet what emerges cannot be reduced to the opening of any of its antecedents. Play is accordingly treated here as a mode of living coordination in which established possibilities are deliberately varied, recombined, or opened to further coordination, under conditions that make the activity worth sustaining, and whose achievements include new coordinations with stakes and histories of their own. This is an enlargement of the original formula and should be tested as one. A game, in these terms, is an arrangement of constraints, in rules, equipment, or routine, that holds some conditions of play steady and can be taken up again, and neither concept requires that the players themselves command symbols.
Much of what this article attempts has been done before, and the debt should be stated. William Corsaro (1992) argued that children do not simply internalise an adult world. They take it up, interpret it, and reproduce it with alterations in peer cultures of their own, a process he called interpretive reproduction, and with Ann-Carita Evaldsson he showed in detail how such cultures are produced and maintained through play and games (Corsaro and Evaldsson 1998). The cultural-historical tradition that descends from Vygotsky has studied jointly constructed playworlds (Lindqvist 1995) and has argued that motives are formed in play through participation and do not precede it as fixed needs (Fleer 2012). Marjorie Taylor and her colleagues (2020) have documented paracosms, the elaborate imaginary worlds that a minority of children maintain in middle childhood. And Ellen Sandseter (2007) has catalogued what children actually put at risk in physical play. That children build worlds by playing is therefore no discovery of this article. What is offered here is a way of bringing these lines of work into one architecture, by asking of any playful arrangement through which mediations it runs, what kinds of counterpart it makes available, what becomes a stake within it and for whom, how it stabilises through repetition, and how it changes or ends.
Some terms need fixing. A child’s mesocosm is the whole field within which she lives. Playworld is used here as an ordinary descriptive word for what children make when they play together over time. Within that range three degrees can be told apart, with no sharp boundaries between them. A playful encounter is a temporarily patterned coordination, such as an exchange of sounds between a baby and a parent, which takes place within an existing relation of care and makes no separate world. A stabilised playful practice is a repeated coordination with recognisable routines, expectations, and a history. A playful submesocosm, a term used provisionally and reserved for the strongest cases, is a field differentiated enough that a network of local counterparts, distinctions, constraints, and dependencies has become consequential for those inside it. The trading game plausibly qualifies. What marks the third degree is the emergence of locally consequential differences within a field of coordination that was already there: a piece of paper becomes currency, a rule that did not exist last week becomes a matter of dispute, and one child’s continued presence becomes indispensable to the others. Such a field is never detached from the wider mesocosm. The players bring their language, their materials, their earlier relationships, and their bodily needs into it, and what happens in the game can alter their relations afterwards. Its boundaries are partly practical, partly symbolic, and permeable in both directions.
The method that follows from this may be called developmental mesocosmography through play. Its techniques are those of naturalistic observation and the ethnography of childhood, and what is new is the organisation of the findings. For any playful arrangement the observer reconstructs the coordinations that existed before the play began; its constitution, meaning the bodies, living counterparts, places, things, and symbols involved; the way its participants mark off what counts within the activity; who can invite, begin, alter, interrupt, or end it; what in it is muted, salient, articulated, conventional, or open to reconsideration; what becomes vulnerable and how far the consequences reach beyond the game; what is carried from one session to the next in memory, routine, material arrangement, or rule; how it changes as the participants change; and how it ends, whether by declining relevance, outside disruption, withdrawal, or the loss of a necessary condition. The strategy is longitudinal, following the same children across time. Moments of failure, an exclusion, a broken rule, an interrupted game, are especially informative, because stakes that had gone quiet become visible in them, and untroubled play needs equal attention, since a method that watched only breakdowns would miss what makes shared play pleasurable and durable. The observer’s position is part of the situation. Some playworlds depend on being opaque to adults, as the trading game partly did, and an adult’s presence may alter what is played.
Three distinctions discipline the answers. The first separates three things that are easily run together. Game objectives are the outcomes an activity defines as success or failure. Local commitments are possibilities that players come to care about through playing, such as keeping a fictional world going or finishing a construction. Vital stakes are the wider coordinations whose viability depends on what happens, such as bodily security, access to companions, or the continuation of an important relationship. Accumulating pretend money was an objective of the trading game, and nobody had a vital stake in pretend wealth. Losing a race costs a child nothing vital, unless the defeat belongs to a pattern of humiliation that threatens her place among her friends. Conversely an imaginary world becomes vitally consequential when it has become one of the main ways in which a friendship is carried on. How objectives and commitments inside a game become attached to coordinations outside it is one of the most interesting questions that play raises. The second distinction is between stakes that exist before a game and stakes that the playing generates, since an improvised activity, repeated, can become something a child very much does not want to lose. The third is between what becomes possible for a child and what becomes important to her. Developmental psychology has mostly asked when a capacity arrives. The question here is when a capacity becomes relevant to a living coordination, and what is then opened or put at risk by its uptake. A child may learn to climb and take no interest in climbing, and another may be captivated by it long before she is any good.
The developmental proposition that organises the rest is this. Children are born into mesocosms that precede them and are integrated into relations of care and material dependence from the start. Through play they come to participate, by degrees, in modifying, differentiating, and constructing the worlds they live in, and the shift from coordination largely organised by caregivers to worlds organised by children is the main line of the story. The process does not end in independence. An adult depends on a mesocosm as much as an infant does and is more skilfully embedded in it.
III. The First Playful World: Cooing and Reciprocal Delight
When my son was about four months old, he and I had a game. I made a sound, he made one back, I answered him, and we could keep the exchange going.
This was in no way precocious, since reciprocal vocal exchange appears very early. Maya Gratier and her colleagues (2015), recording mothers with infants between eight and twenty-one weeks old, found conversation-like turn-taking already in place. Colwyn Trevarthen (1979) called such exchanges protoconversation and treated them as the first form of intersubjectivity. Daniel Stern (1977) described the games of mothers and infants as a theme with variations, in which the adult repeats a sequence while altering its timing by small amounts and the infant’s excitement follows the alterations. By the second half of the first year there are structured routines such as peekaboo, whose pleasure depends on the face returning a little later, or from a slightly different place, than it did the last time (Bruner and Sherwood 1976).
These exchanges arise within a condition that has to be described first. D. W. Winnicott (1960) remarked that there is no such thing as an infant, meaning that a baby exists only as part of a unit of care. Her temperature, arousal, and sleep are regulated through the caregiver’s body, and the commonest event of her day is a small arc in which a felt misalignment, hunger or cold or loneliness, issues in a cry, the cry is answered, and calm returns. The arc runs from fluent coordination through salience and back again, and it is remedial, since something was wrong and has been put right. The cooing game runs the same arc for its own sake. Nothing was wrong. A sound is offered where none was needed, the small excitement of waiting for the answer is sought and enjoyed, and the game is the first instance in a life of an opening that is wanted for itself.
The question of interest is what kind of coordination becomes available in such an exchange. A caregiver and an infant establish a temporally organised pattern of mutual response. A sound makes an answer possible, the answer comes to be anticipated, and a variation on it produces excitement or surprise. All of this takes place as fluent coordination and as felt salience, with no symbols involved. The infant does not need to understand the exchange as a game, to recognise a rule, or to represent the parent as a playmate. Yet the elements of all later social play are present: an opening is offered, an answer comes, each party adjusts to the other in time, a variation is introduced and taken up, and both are delighted.
Put to this small world, the five questions give instructive answers. It is constituted by two bodies, two voices, and a stretch of time in which the infant is fed, rested, and alert. Its boundary is drawn almost entirely by the infant’s state, and the game is over the moment she turns her head away. Its coordination consists in timing. Its history is short and real, since a routine that has been played a dozen times is anticipated in a way that a new one is not. The game as I remember it was dyadic, a matter of two voices with no object between us, although exchanges arranged by adults around a rattle or a picture book are found in the first months as well.
Its stakes lie in two mediations at once. In embodiment, what is at stake is the regulation of arousal: too little stimulation and the infant’s attention drifts, too much and she is overwhelmed, and an overtired baby cannot be played with at all. In being-with, what is at stake is the continuation of reciprocal engagement itself. The infant is discovering that another living being responds, that the response can be invited, and that an encounter can be kept going by what she does. Playful exchange at this age is entangled with feeding, soothing, and the other necessary coordinations of care, from which it can be distinguished without being separated.
The arrangement is asymmetric, and the asymmetry matters for what follows. The adult provides most of the conditions: the occasion, the pacing, the degree of variation, and much of the buffering of stakes, since it is usually the adult who notices that the baby has had enough. The regulation is nevertheless mutual. The infant steers the exchange by her gaze, her attention, her show of pleasure, and her withdrawal, and an adult who is any good at the game is being led as much as leading. She can answer or decline to, and before long she can begin, by catching the adult’s eye and producing the sound that started it last time. The transfer of initiative that runs through the whole of childhood has started.
The first game has no winner, no objects, and no pretence. It has the structure that persists through everything that follows: the initiating of an opening, the receiving of an answer, the discovery that the other is willing to continue, and pleasure in the other’s responsiveness. Long before a child can win a game, the child can enjoy making another living being answer back.
IV. Making the Body and World Playable
Through the rest of infancy and the toddler years, with no sharp boundary of age, the child’s surroundings turn into a field of things that can be done. Crawling, walking, climbing, throwing, stacking, carrying, and hiding each bring coordinations within reach that were not available before. A staircase becomes something to climb, a slope something to run down, a box a hiding place or a container, and a puddle an event. Some of these were made and some were found, and a child at play in a place is nearly always coordinating with both at once, with gravity, weather, and terrain and with what people have built there. The world has not changed. The child’s capacities have differentiated it.
Through the first year the infant’s engagement with things in company grows steadily more differentiated. Adults arrange encounters around objects from the first months, infants respond to another person’s attention to a third thing by about three months (Striano and Stahl 2005), and a review of fifty-one studies concludes that triadic coordination is built up gradually from birth through adult mediation and does not appear suddenly at nine months (Mendoza-García and Moreno-Núñez 2023). In the second half of the year the infant takes a larger part, looking from the object to the adult and back, showing, giving, and pointing (Trevarthen and Hubley 1978; Tomasello 1999). Games of give-and-take, and of dropping a spoon to have it handed back, belong to this period. Made things were in her world from the start. What is new is that she engages them as counterparts with uses that others demonstrate and expect, and takes a hand in the coordination by which they are used.
Two mediations come forward here that the first game barely touched. Dwelling, the coordination with conditions that nobody made, shows itself in play chiefly as a matter of access and orientation: getting to the end of the garden, knowing the way back, finding a spot behind the sofa that is one’s own. The stakes are where a child can go, whether she can play there without interruption, and how far from the familiar she can move without losing her bearings. Multimateriality becomes a matter of what things afford and who controls them. Objects are counterparts that do not answer back. Blocks topple, balls roll away, a container empties and can be filled again, and the toddler discovers that she can cause an event, repeat it, and vary it, and that the thing will behave in the same way whatever her mood. A tower of blocks adds something further, since it persists after the building is over, and with persistence come new stakes, of possession, of preservation, and of who is entitled to knock it down.
The play of these years exercises all three types of counterpart together, and it is where they are told apart. The tower obeys gravity whatever its builder wishes, and infants are sensitive to the solidity and persistence of objects within their first months (Spelke and Kinzler 2007). The playmate, or the parent, answers back. And the child’s own excitement and frustration have to be managed if the game is to go on, a coordination with her own state in which the adult at first carries a large share. Children do not learn the physical world first and the social world afterwards. They sort the two concurrently, and young children readily treat as responsive whatever has a face or a name, so that the teddy bear is sad and the car is cross. Piaget (1929) called this animism and took it for a confusion, and later work has questioned both how pervasive it is and how much it owes to the way children are asked (Carey 1985). The reading proposed here is an interpretation and no established finding: such attributions apply the child’s oldest and most practised discernment, formed in the first months with those who cared for her, to counterparts that resemble agents.
The stake that these years bring most sharply into view is the doing of things oneself. Every parent knows the toddler who is building a tower and is distraught when an adult helpfully adds a brick. The tower is trivial. What was at stake was authorship of the event, and the play has shown, with great clarity, that this has recently become something that can be lost. It is no accident that the stake appears in the same months in which children come to recognise themselves in a mirror and to say I (Amsterdam 1972). A self that can be the author of an event has lately become available, and play is where it is tried out. It would be easy to label the stake competence, but the label is too coarse. The same child may care nothing for building well and a great deal for building unaided.
Some of this play looks, to adults, like the pursuit of danger. Sandseter’s (2007) observations in Norwegian preschools found children seeking out great heights, high speed, dangerous tools, dangerous elements such as water and fire, rough-and-tumble, and places where they could disappear from adult view. Such play shows that the stake of bodily security is not always met by avoiding disturbance. Children deliberately approach the limits of their balance and nerve, and the thrill lies in a felt disturbance that stays just within what they can manage. Two things need to be kept apart in describing it. One is the experienced challenge, which is the sought opening of the child’s coordination with her own body and with height or speed. The other is the actual exposure of the body to harm, which is a fact about the branch and the drop and may be greater or smaller than the child supposes. Play of this kind buffers its stakes imperfectly, which is part of what children learn from it, and adults disagree about how much buffering to supply.
Other children enter slowly. Two toddlers with the same box of bricks are for a long time playing beside each other more than with each other. What arrives first is the wish to have a feat witnessed, in the look back at a parent after the jump has been landed. The stake in that look is recognition, an early form of something that will matter a great deal later, which is that an achievement counts for more when someone whose regard matters has seen it.
By the end of these years the mesocosm that preceded the child has been enlarged and reconfigured by her own action. She has places, things, and feats that are hers, and she can start, stop, and repeat a good many activities without help. Pretence, which begins within these same years, adds a possibility of a different kind.
V. The Invention of Other Worlds
During the second year a stick becomes a horse, an empty cup holds a drink, and a doll is fed and put to bed. Simple pretence of this kind grows more elaborate over the following years, as objects stand in for other objects, roles are taken, and other children are drawn in (Fein 1981). Vygotsky (1967) saw in the stick ridden as a horse the pivot by which meaning first comes apart from the object. In the present terms, the child has begun to organise activity around distinctions that do not hold literally in the surrounding world and that become locally operative through playing. The sofa is a ship for as long as those aboard treat it as one.
Pretence takes several forms, which do not arrive in a fixed order. It can be embodied, as when a toddler lifts an empty cup and drinks. It can be enacted with things and with other people, without a word of explanation. And it can be articulated in language, which is what allows a fiction to be proposed, shared, and corrected. Nor do children’s attributions divide neatly into the believed and the pretended. A child may treat her bear as literally sad, as sad in the game, as possibly sad, or as sad for as long as that is useful, and may pass among these within a minute. The question of interest is less whether she believes the attribution than how she coordinates with a counterpart under it, how the attribution is kept up, and how freely she moves between treating it as literal, pretended, uncertain, or convenient. Play is where a flexible relation among attribution, actual responsiveness, and symbolic framing is worked out, and that relation goes on developing well beyond early childhood.
How such a world is established, stabilised, and kept up is worth describing with some care. A shared fiction usually begins with an articulation made for the occasion: let’s say the sofa is the ship, and you be the captain. The articulation binds nobody beyond this afternoon. It has to be taken up by the other players to have any force, and a child who replies that it is only a sofa has ended the voyage before it began. Once taken up, it is sustained by action more than by words, as cushions become cargo and the carpet has to be swum across. The fiction also has to survive interruptions. Children leave a pretend world for lunch and re-enter it afterwards, and what allows them to do so is partly memory and partly material: the cushions are still where they were, and the arrangement of the room holds the world in place while its makers are away. If the game is played again the next day, and the day after, its articulations begin to settle into conventions that no longer need stating, and the ship has become a standing feature of that household’s play. Conventions of this home-made kind sit alongside portable ones that children are handed from the start, in nursery games, in the rules adults supply, and in the routines of older children.
The boundary of such a world is kept by the players themselves, and keeping it is a large part of what they do. Catherine Garvey (1977) noticed that children at pretend play move constantly between acting inside the fiction and stepping outside it to direct it, usually in a different voice: pretend you didn’t see me, no, you have to be asleep. A good share of any session of make-believe consists of this stage management. The children are playing in the world and, in the same breath, deciding what the world contains, and the second activity is no less absorbing to them than the first.
Three kinds of imagined world can be distinguished by what anchors them. Some are private, held by one child, and the paracosms studied by Taylor and her colleagues (2020), found in roughly one in six of the eight- to twelve-year-olds they interviewed, are the most elaborate of these, with histories, maps, and populations. Some are shared and improvised, like the ship. And some are anchored in a durable work that the children did not make. When my children were four and six they converged, for no reason anyone could identify, on Mozart’s The Magic Flute. For the better part of a year they listened to it again and again, sang the arias to each other, staged scenes with whatever props were lying about the house, and compared filmed productions closely enough to argue over the merits of different singers. They had not invented the opera. They had made it the ground of a world, with characters, situations, favourite moments, and a repertoire of things to do in it, which they could enter together, leave, and enter again.
That episode shows something general about the relation between a fixed deposit and living play. A recording does not respond to anyone, however often it is played. Because it stayed the same, two children could build on it, and each child’s anticipation of a scene could be recognised and joined by the other. A stable made thing supported an interrecursive coordination between two people that neither could have built from nothing at that age. Stories read aloud, television series, and the franchises of the toy industry serve children in the same way, as ready-made ground on which their own worlds are raised.
The stakes of imaginary worlds are easily misdescribed, because the fiction suggests that nothing real can be at issue. What is at stake for those inside is the continuity of the imagined situation, which a careless adult or an uncooperative sibling can destroy in a sentence; the chance to contribute to it, and to have a contribution accepted; recognition in the role one has proposed for oneself; and continued access to the people with whom the world is shared. Marilyn Fleer’s (2012) point applies here, that the motives at work in play are formed in the playing. Neither of my children had any prior need of Mozart. Within months the opera mattered to them a great deal, and what mattered was a world that they had made out of it together.
VI. Worlds Shared, Contested, and Defended
A shared playworld exists only as long as its participants keep it going, and keeping it going is work. Entry has to be negotiated, roles allocated, permissions granted and refused, modifications proposed and accepted, and disputes settled. Who may establish a rule, who may overrule whom, who decides whether a contribution counts, and who can declare the game over are questions that arise in every group of children at play. They are sometimes treated as interruptions of the play. They are better understood as among the coordinations through which collective play exists at all.
The ethnography of childhood has documented this in detail. Corsaro (1985) showed nursery children defending an established game against newcomers, on the sound intuition that the game was fragile and that a clumsy entrant could wreck it, and he described the oblique strategies by which experienced children get in anyway, by hovering, by doing alongside what the players are doing, by offering something the game can use. Corsaro and Evaldsson (1998) traced how routines of this kind, repeated day after day, amount to a peer culture with its own valued objects, shared knowledge, and expectations. Marjorie Harness Goodwin (2006), recording girls at hopscotch and jump rope, found the rules of the game inseparable from a running negotiation over who stood where in the group. Iona and Peter Opie (1969) showed how much of this apparatus children inherit from other children, in counting-out rhymes, truce terms, and ways of picking sides that have been passed down for generations without adult help. Vivian Paley (1992) found exclusion painful enough among five-year-olds to legislate against it in her classroom.
Rules are present in children’s play from very early, in the routines of peekaboo and ring games and in whatever adults lay down. Piaget (1951) placed games with rules proper at about seven, and his study of boys playing marbles remains instructive (Piaget 1932). The younger boys held the rules to be sacred and unalterable, handed down by fathers or by God, while applying them very loosely. They were meeting a made arrangement as though it were found ground, like the weather. Boys of eleven and twelve treated the same rules as agreements among players, which could be changed by mutual consent so long as the change was fair. The contrast is too sharp as a timetable. Children between four and seven already distinguish a rule that one child invented, which that child may change, from a rule agreed by a group and from a rule set by an adult (Zhao and Kushnir 2018). Several achievements are involved, and they need not share an age of onset: recognising where a rule came from, taking part in negotiating its authority, treating it as binding, and eventually reconsidering a whole system of rules as something made.
What is at stake when a child is kept out? Consider a six-year-old who wants to join three others playing at shops and is refused. An observer might conclude that she wants status, and the observer might be quite wrong. She may want to spend time with one particular friend among the three. She may want access to an absorbing imaginary world. She may want to work the toy till. She may want to be a customer and have no wish at all to run the shop. Or the refusal may threaten a position in the group that she had taken to be secure. The same visible event can disclose different stakes depending on the history of the child’s coordinations, and they cannot be read off the behaviour. What she does next narrows the possibilities without settling them. She may try to enter by another role, start a rival shop, seek out one of the three later, appeal to an adult, or withdraw. Each of these is open to several readings. A rival shop may preserve an enjoyable activity, retaliate, or bid for the same playmates, and an appeal to an adult may express a sense of unfairness or may simply be the most reliable way in. A response identifies possibilities for interpretation. Identifying the stake requires repeated observation, comparison across settings, the child’s own account where she can give one, and attention to what in fact happens when a particular coordination becomes unavailable to her.
Laid out by mediation, the stakes that shared play brings into view are far more various than the usual list of competence and status allows. In embodiment there are the regulation of excitement and fatigue, security from pain and injury, freedom of movement, and control over how one’s body is handled by others. In being-with there are reliable responsiveness, being invited and included, continued access to particular playmates, trust, loyalty, mutual enjoyment, and the positions of leading and following. In dwelling there are access to places, somewhere to play undisturbed, the control of territory, hiding places, and the freedom to go beyond familiar bounds. In multimateriality there are access to toys and materials, possession and exchange, and the preservation of what has been built. In multisymbolization there are the shared understanding of who is pretending to be what, the authority to say what counts, the legitimacy of rules, and the ability to make a description stick. These are candidate domains and no finished taxonomy, and the earlier distinction applies throughout: much of what is listed will on a given occasion be an objective or a local commitment, and becomes a vital stake only where a wider coordination depends on it. Some are conditions for others, and standing in particular may be less a stake in itself than a means of securing continued being-with.
Two further distinctions come out of the ethnographic record. The stake of taking part in a particular game is different from the stake of sustaining the world in which games are played. A child may be perfectly content to lose this round and devastated to be left out of the next. And stakes become operative at different levels. Children care about what happens inside the fiction, about who has the right to decide what happens, and, among older children, about whether the whole arrangement is fair, and a child who shows little interest in running the pretend shop may care intensely about being the one who says who can.
The distribution of these stakes is unequal. Some children are far more invested in a given world than others, and those who are least invested have the most power over it, since they can afford to walk away. Some control the rules, and others accommodate themselves in order to stay in. Older siblings are required to let younger ones win, and younger ones accept poor roles for the sake of being included. A playful world is an arena of living value in its own right, with its own inequalities, and adults who describe it as a harmless copy of their own society, or as children only playing, are reporting how it looks from outside.
VII. Making Abilities Count
In later childhood, with much earlier beginnings, a great deal of play comes to turn on comparison. Who runs fastest, who climbs highest, who throws furthest, who knows the answer, who can trick whom, and who is picked first are settled on the playground many times a day. It is tempting to describe this as the stage at which status takes over from earlier concerns, and the description should be resisted. Nothing is superseded. What happens is that the child’s increasingly differentiated capacities become objects of comparison within shared worlds, and the instructive question is how a capacity comes to count.
A game is an arrangement for making a capacity count. The playground race makes speed consequential, chess does the same for a certain kind of anticipation, and a riddle for the knack of seeing an unexpected solution. Games do more than reveal capacities that were there already. They stabilise criteria under which particular capacities acquire relevance, so that a child can be highly able in one playworld and nearly helpless in another, the fastest runner in the class being nobody in particular at cards. Children accumulate reputations accordingly, and the reputations are of more than one kind. Being liked and being regarded as popular or powerful are related but distinct forms of standing, only moderately associated across two decades of studies (van den Berg, Lansu, and Cillessen 2020), and which of them a child is pursuing in a given game is, again, something to be reconstructed.
Two pleasures are involved, and they differ. One is the pleasure of mastery shown, of doing a difficult thing well before people who can appreciate it. The other is the pleasure of cleverness, of seeing something that another participant has not yet seen, or of seeing through what the other hoped to conceal. Card games, riddles, bluffing, and practical jokes all turn on it. A child who has discovered this pleasure has discovered that reading other people and being read by them is itself a contest. It should not be supposed that the clever child is operating at a higher level of recursivity than her opponent. She has discerned more within the same coordination. Successful coordination with other people also runs well ahead of the ability to explain it. Two-year-olds hide forbidden things, tease, and feign, years before they pass the standard verbal test of false belief at about four (Wimmer and Perner 1983). How much understanding of other minds should be credited to infants on the evidence of where they look (Onishi and Baillargeon 2005) is disputed, and attempts at replication have left the matter open (Poulin-Dubois et al. 2018). The present argument does not depend on the answer. It needs only the observation that children succeed in teasing, hiding, and tricking long before they can say how.
Competition has a peculiar relation to the play that carries it. I played chess with my son for some time, and at first he lost every game and minded a great deal. At twelve or thirteen he spent two or three weeks learning openings, by what means I have never discovered, and came back with a set of opening traps that I could not handle. I have not won a game against him since, and we no longer play. His success made him a far better player and ended chess as something we did together. Winning a game and sustaining mutually enjoyable play are different achievements, and overwhelming victory, repeated, destroys the conditions for the second.
The opposite case came about at a choir retreat in France, where on a hot day I, the only adult, and a sixteen-year-old who could really play football took on about ten younger children. We were hopelessly outnumbered and far ahead in strength and skill, and for about an hour the game went back and forth with nothing between the sides. The inequalities had been set against one another, and a practical balance emerged that no rule had arranged. What keeps a contest worth playing is that each side has openings the other can answer. Equal ability is one way of securing this, and children are inventive about others, giving the little ones a head start, splitting the two best players between the teams, and declaring a goal by the smallest child to count double. Their sense of an unfair advantage is correspondingly exact, and it attaches less to inequality than to anything that removes the other side’s chances.
VIII. The Layering of Playworlds: Why Some Catch and Others Disappear
The trading game in New Zealand and the year of The Magic Flute were both extraordinarily absorbing, and neither remained so. Understanding why requires asking what accumulates in a world of play that is returned to day after day.
Three conditions can be distinguished in the catching of a playworld. The first is availability. The participants must have access to a configuration they are able to take up, and a game that calls for capacities, materials, or answering moves that are absent cannot take hold in that form. The second is resonance. Something in the coordination has to produce a positively valenced salience against what the players have come to anticipate, a variation that delights where it might have confused or bored. The third is recursive amplification. Repeated participation builds a more differentiated field of anticipation, shared references, routines, and knowledge of one another’s characteristic moves, which makes ever more specific variations enjoyable and increases the place of the activity in the players’ days. Repetition is therefore no enemy of enjoyment. Children who play the same game every afternoon are, on a superficial description, doing the same thing again and again, and within the repetition a variation that a newcomer would not notice becomes exquisitely funny or exciting to those who have been there all week. It also follows that established players cannot easily share their enjoyment with a newcomer, since the rules can be explained in five minutes and the accumulated history cannot be explained at all.
The two episodes differ in one respect that bears on their intensity. With the opera, the children were resonating with a deposit that stayed exactly as it was while their relation to it changed. With the trading game, the players could alter the game itself, so that the world with which each was resonating was continually being remade by the other two. In both cases the pleasure lay in the uptake, and it could not have been deposited in a recording or in a set of rules.
The same history can undo the world it built. The field may be exhausted, with nothing further to find in it. The players’ capacities may change, together or, more disruptively, at different rates, as when one child outgrows a game that the other still loves. A third child may arrive, or one of the three may go home. Other possibilities may become more relevant. With the opera, I do not know which of these occurred. One child’s interest may have waned first and taken something from the other, or both may have moved on at once. The recordings remained in perfect condition, and nothing about Mozart had changed.
Across the whole of childhood, at least four processes are at work, and they are worth separating. The first is accumulation, in which new possibilities are added and old ones stay. A five-year-old capable of elaborate pretence still enjoys the exchange of nonsense sounds that delighted her as a baby. The second is reconfiguration, in which an earlier form of play acquires new participants, rules, or stakes. Chasing begins as a game between a parent and a crawling infant, and by eight it is a group affair with bases, alliances, and tactics. The third is dormancy, in which a form of play stops being sought while the capacity to enjoy it remains, so that a different companion or a changed situation can make it attractive again. A ten-year-old who would not be seen with building blocks may spend a happy afternoon with them when a much younger cousin supplies the occasion. The fourth is loss, which is different from the other three. A place is built over, a friend moves away, a body grows too big for the den, and the game cannot be recreated however much it is wanted. To describe every ending of play as a waning of interest is to miss the cases in which interest survived and the conditions did not.
Layering is compatible with development having a direction. Capacities for joint engagement, for pretence, for taking up rules, and for reconsidering them build on one another, although their trajectories overlap and none waits for another to be complete. What becomes important follows no sequence at all, and what became available earlier is not left behind. A ten-year-old’s play contains the infant’s exchange of sounds, the toddler’s climbing, the four-year-old’s pretence, and the seven-year-old’s rules, each active or dormant according to the company and the day.
This layered picture is the article’s main argument against the idea that children simply grow out of earlier kinds of play. In many cases nothing is outgrown. The capacities remain, and what changes is their organisation within a developing mesocosm, in which some coordinations are active, others dormant, and others no longer possible. A form of play can vanish from a child’s days without vanishing from her repertoire of possible enjoyments, a proposition that can be tested by watching what children do when the usual audience is absent.
For the method described earlier, the layered picture has a practical consequence. What a child is seen to play on a given day is a sample of the active layer only. One caution therefore governs every inference: no single episode establishes what is vital to a child, and distress, enthusiasm, or competitiveness observed on one afternoon have to be set against the child’s wider history before anything is concluded from them.
IX. When Playing War No Longer Works
I was about twelve, and my friends and I were playing war. We had been playing war for as long as I could remember. We were soldiers, and we had our weapons. On one particular day, which I can still recall, it felt wrong. I believe all of us felt it on the same day, and that phase was over.
For years that world had done a great deal for us. It had its territory, its equipment, its roles, and a long history, and it gave occasion for courage, cunning, loyalty, and tactical invention. Nothing about its rules had changed, the friends were the same friends, and we were presumably better at it than we had been the year before. A durable and thoroughly enjoyable world of play lost its hold, more or less overnight, without any alteration in its form.
What that world did not do is as instructive as what it did, and the point bears on one of the most influential accounts of children’s games in the social sciences. Among the Kabyle of northern Algeria, Pierre Bourdieu (1990) described a game called qochra, played by boys in the first days of spring. The players struggle over a cork ball, which stands for a woman, as daughter or as wife. One boy, chosen by lot, is the father. He has to defend his daughter against abduction by the others, and he also has to get rid of her, by passing her on to another player under the rules. A father who fails is shamed, and in the extreme case he is put out of the game with the ball tied under his shirt, so that he looks like a girl who has been got pregnant. Bourdieu counted games of this kind among the structural exercises through which a group hands on its practical mastery from one generation to the next. In them the sense of honour is inculcated from the earliest days of life and reinforced by calls to order from the group (Bourdieu 1977), and the girls, who do not play, learn their place from being represented in the game by an object. The dispositions so acquired are, on his account, durable and transposable, a habitus that will generate the conduct of the adult man of honour.
Bourdieu gets some of this right. No game is innocent of the world around it, and qochra draws on the Kabyle arrangements of honour and marriage as surely as the trading game drew on money. Games are deposits, and they carry what has been sedimented in them. Children’s bodies learn in play. Nor is Bourdieu’s theory of practice a mechanical one, since he insists that the conduct of adults is regulated improvisation, open and uncertain as it unfolds. The determinism lies in a narrower place, in the relation he assumes between a world of play and the shared world that surrounds it. To call a game a structural exercise is to suppose that what is practised in the game is what is deposited in the player, so that the playworld is a scale model of the mesocosm and playing it installs the corresponding disposition. The homology between the game and the society is established by the analyst. That it is transferred into each boy who plays is assumed and is nowhere shown.
My own case is evidence against the assumption. I played soldier for years, with great commitment, and no military habitus was inculcated in me. There is no one-to-one transfer from what is played to the skills, habits, and goals of life outside the game, and the distinctions drawn in this article explain why there could not be. The objectives of a game are different from the commitments its players form and from the vital stakes that ride on it. What mattered to us in playing war may have been the friends, the territory, and the invention of plans, with soldiering as the fiction that carried them. A deposit does not determine its uptake, and the same game supports different playworlds for different players, so that one boy in the role of the father may be absorbed in honour, another in the dodging and running, and a third in not being left out. Pretence is coordination under an attribution that the players can hold as literal, as pretended, or as merely convenient, and what it exercises is first of all the handling of pretence. And what play leaves behind is a layer, which later life may keep active, let fall dormant, or lose.
Where the continuity that Bourdieu describes is real, it has therefore to be explained by something more than the game. In Kabylia, on his own evidence, the game of honour among adult men took up daily what the boys’ game rehearsed, and the calls to order came from the whole group. The carrying over was done by a surrounding mesocosm that met the same disposition in many coordinations at once, a redundancy that would hold the disposition in place whatever was played in spring. Where the surrounding world does not take up what a game rehearsed, the layer falls dormant, which is the likeliest account of my own case. Play is, on this view, a poor predictor of the adult taken alone, and the old idea that play is practice for adult life, of which the structural exercise is a sociological variant, mistakes one possible career of a playworld for its function. Whether anything built in play is carried into the shared world, and what, is a matter for reconstruction case by case.
The preceding sections supply terms in which to ask what happened, and they do not supply the answer. The world may have been exhausted and incapable of producing anything new. Other possibilities may have become more relevant. Changing relationships among the boys may have altered what taking part meant. The category of make-believe may have become embarrassing, a classification taken over from older children as much as felt. Children of that age are also increasingly able to stand back from conventions they have long used, and the category of play, under which adults had filed everything we did, may have been one of those reconsidered. The bodily changes of puberty also begin at about that age. A memory cannot decide among these, and several may have been true together. Nor can it say whether what occurred was dormancy or loss.
At about the same time our vocabulary changed. For years we had said that we were going out to play. From then on we said that we were going out, or meeting friends, which is what English-speaking teenagers call hanging out. The intransitive verb lapses in many households at around this age, while play that takes an object, football, a video game, an instrument, carries on undisturbed.
The article stops at this threshold. Why some of the worlds of childhood suddenly cease to be compelling as adolescence begins, what new configurations of being-with become possible in their place, and what new stakes enter with them are questions for a separate inquiry. What can be said here is that the transition, whatever drives it, acts upon something that is already complex. The twelve-year-old who stops playing war has behind him a decade of worlds built, shared, contested, layered, and left, and he does not arrive at adolescence from a simple place.
X. Conclusion: Play as Mesocosmography
The trading game in New Zealand was more than a pastime, and it was something other than a rehearsal of economic skills for later use. It was a living achievement, in which three children established a locally meaningful world by coordinating with one another, and for much of a visit of two and a half weeks that world absorbed them completely.
This article has traced how the capacity for such achievements is built up, from an infant’s part in an exchange of sounds, through the toddler’s authorship of small events and the young child’s fictions, to the negotiated worlds and contests of later childhood. Throughout, new forms are added to old ones, earlier forms are reconfigured, and some fall dormant or are lost.
What develops along this line is poorly described as a succession of stakes, bodily security giving way to competence and competence to status. Stakes in several mediations are in play from the first months, and what changes is an expanding, overlapping, and continually reorganised arrangement of many kinds of coordination. Three achievements can be distinguished within it. Children discover worlds through play, meeting possibilities that were already present in their bodies, their relationships, their surroundings, and the arrangements of their culture. They construct worlds, reorganising those possibilities into arrangements with rules, boundaries, and histories of their own. And they make worlds matter. Through continued uptake, shared enjoyment, and sedimentation, an arrangement acquires relevance and can become consequential for coordinations that reach well beyond the game. The third achievement explains why the trading game mattered so much to three children when the adults could not say what its significance was, why the destruction of an elaborate construction is felt as far more than the loss of some arranged objects, and why losing access to a shared game can mean losing something in a friendship. None of this licenses an inference from what children play to what they will become. A world that is made to matter in play is not thereby installed in the player, and what is carried from a playworld into the shared world depends on what the rest of a life takes up.
The article’s central claim can now be summarised: Children’s play does more than make existing vital stakes observable. Through playful coordination children take part in constructing the arrangements within which particular things come to matter, become vulnerable, and sometimes become indispensable. Lists of what is at stake for children, whether bodily security, competence, recognition, or standing, name possible dimensions and cannot say in advance what matters in any particular child’s world. That has to be reconstructed, which is the work of the method proposed here. The familiar developmental question is how children acquire capacities. The question pursued in this article is how they come to inhabit, and to take part in constructing, the worlds in which those capacities and the stakes attached to them matter.
Selected References
Amsterdam, Beulah. 1972. “Mirror Self-Image Reactions before Age Two.” Developmental Psychobiology 5(4): 297–305.
Bourdieu, Pierre. 1977. Outline of a Theory of Practice. Translated by Richard Nice. Cambridge: Cambridge University Press.
Bourdieu, Pierre. 1990. The Logic of Practice. Translated by Richard Nice. Stanford: Stanford University Press.
Bruner, Jerome S., and V. Sherwood. 1976. “Peekaboo and the Learning of Rule Structures.” In Play: Its Role in Development and Evolution, eds. Jerome S. Bruner, Alison Jolly, and Kathy Sylva. New York: Basic Books.
Carey, Susan. 1985. Conceptual Change in Childhood. Cambridge, MA: MIT Press.
Corsaro, William A. 1985. Friendship and Peer Culture in the Early Years. Norwood, NJ: Ablex.
Corsaro, William A. 1992. “Interpretive Reproduction in Children’s Peer Cultures.” Social Psychology Quarterly 55(2): 160–177.
Corsaro, William A., and Ann-Carita Evaldsson. 1998. “Play and Games in the Peer Cultures of Preschool and Preadolescent Children: An Interpretative Approach.” Childhood 5(4): 377–402.
Fein, Greta G. 1981. “Pretend Play in Childhood: An Integrative Review.” Child Development 52(4): 1095–1118.
Fleer, Marilyn. 2012. “The Development of Motives in Children’s Play.” In Motives in Children’s Development: Cultural-Historical Approaches, eds. Mariane Hedegaard, Anne Edwards, and Marilyn Fleer, 79–96. Cambridge: Cambridge University Press.
Garvey, Catherine. 1977. Play. Cambridge, MA: Harvard University Press.
Goodwin, Marjorie Harness. 2006. The Hidden Life of Girls: Games of Stance, Status, and Exclusion. Malden, MA: Blackwell.
Gratier, Maya, Emmanuel Devouche, Bahia Guellai, Rubia Infanti, Ebru Yilmaz, and Erika Parlato-Oliveira. 2015. “Early Development of Turn-Taking in Vocal Interaction between Mothers and Infants.” Frontiers in Psychology 6: 1167.
Lindqvist, Gunilla. 1995. The Aesthetics of Play: A Didactic Study of Play and Culture in Preschools. Uppsala: Acta Universitatis Upsaliensis.
Mendoza-García, Ana, and Ana Moreno-Núñez. 2023. “Early Triadic Interactions in the First Year of Life: A Systematic Review on Object-Mediated Shared Encounters.” Frontiers in Psychology 14: 1205973.
Onishi, Kristine H., and Renée Baillargeon. 2005. “Do 15-Month-Old Infants Understand False Beliefs?” Science 308(5719): 255–258.
Opie, Iona, and Peter Opie. 1969. Children’s Games in Street and Playground. Oxford: Clarendon Press.
Paley, Vivian Gussin. 1992. You Can’t Say You Can’t Play. Cambridge, MA: Harvard University Press.
Piaget, Jean. 1929. The Child’s Conception of the World. London: Kegan Paul, Trench, Trubner.
Piaget, Jean. 1932. The Moral Judgment of the Child. London: Kegan Paul, Trench, Trubner.
Piaget, Jean. 1951 [1945]. Play, Dreams and Imitation in Childhood. London: Heinemann.
Poulin-Dubois, Diane, et al. 2018. “Do Infants Understand False Beliefs? We Don’t Know Yet: A Commentary on Baillargeon, Buttelmann and Southgate’s Commentary.” Cognitive Development 48: 302–315.
Sandseter, Ellen Beate Hansen. 2007. “Categorising Risky Play: How Can We Identify Risk-Taking in Children’s Play?” European Early Childhood Education Research Journal 15(2): 237–252.
Spelke, Elizabeth S., and Katherine D. Kinzler. 2007. “Core Knowledge.” Developmental Science 10(1): 89–96.
Stern, Daniel N. 1977. The First Relationship: Infant and Mother. Cambridge, MA: Harvard University Press.
Striano, Tricia, and Daniel Stahl. 2005. “Sensitivity to Triadic Attention in Early Infancy.” Developmental Science 8(4): 333–343.
Taylor, Marjorie, Candice M. Mottweiler, Naomi R. Aguiar, Emilee R. Naylor, and Jacob G. Levernier. 2020. “Paracosms: The Imaginary Worlds of Middle Childhood.” Child Development 91(1): e164–e178.
Tomasello, Michael. 1999. The Cultural Origins of Human Cognition. Cambridge, MA: Harvard University Press.
Trevarthen, Colwyn. 1979. “Communication and Cooperation in Early Infancy: A Description of Primary Intersubjectivity.” In Before Speech: The Beginning of Interpersonal Communication, ed. Margaret Bullowa. Cambridge: Cambridge University Press.
Trevarthen, Colwyn, and Penelope Hubley. 1978. “Secondary Intersubjectivity: Confidence, Confiding and Acts of Meaning in the First Year.” In Action, Gesture and Symbol: The Emergence of Language, ed. Andrew Lock. London: Academic Press.
van den Berg, Yvonne H. M., Tessa A. M. Lansu, and Antonius H. N. Cillessen. 2020. “Preference and Popularity as Distinct Forms of Status: A Meta-Analytic Review of 20 Years of Research.” Journal of Adolescence 84: 78–95.
Vygotsky, Lev S. 1967 [1933]. “Play and Its Role in the Mental Development of the Child.” Soviet Psychology 5(3): 6–18.
Wimmer, Heinz, and Josef Perner. 1983. “Beliefs about Beliefs: Representation and Constraining Function of Wrong Beliefs in Young Children’s Understanding of Deception.” Cognition 13(1): 103–128.
Winnicott, D. W. 1960. “The Theory of the Parent-Infant Relationship.” International Journal of Psycho-Analysis 41: 585–595.
Zhao, Xin, and Tamar Kushnir. 2018. “Young Children Consider Individual Authority and Collective Agreement When Deciding Who Can Change Rules.” Journal of Experimental Child Psychology 165: 101–116.