A Living Value Theory of Play: Fun, Frames, and the Worlds That Players Make (Playtime 1/3)

Abstract

Why is the same game magical one evening and flat the next, or a delight with one person and a chore with another? Classic theories of play ask what play is and leave fun to taste. This article proposes three claims. Play is a distinctive way in which living coordination is opened, varied, maintained, and enjoyed under reallocated stakes. Games are deposits that stabilise conditions for such coordination without guaranteeing its enjoyment. And playing can generate a temporary world in which relevance, frames, and stakes are continually reorganised. What makes playing wonderful cannot be preserved. It can only be achieved again.

I. The Same Game, Twice

Two friends sit down on a Friday evening to a board game they both know well. Within twenty minutes neither is aware of the time. Each move draws a groan or a laugh, a run of bad luck becomes a running joke, a rule is bent by agreement and the bending improves things, and at midnight they are surprised to find themselves still at the table. On Saturday the same two people set out the same board, and nothing happens. The moves are legal, the game proceeds, somebody wins. Both would say, if asked, that it was fine. Neither suggests playing again.

The puzzle generalises in three directions. The same person can find chess fascinating at fourteen, tedious at twenty-five, absorbing again at forty, and dull once more at sixty, while the rules stay what they were. The same game on the same evening can be a delight with one opponent and a chore with another of equal skill. And now and then two people discover, without either having planned it, that a particular kind of play works between them and with nobody else, and that for a season it is among the best things in their lives. Then, often, it stops, and neither can say when it was played for the last time.

Most theories of play have little to say about any of this. They ask what play is, what it is for, and how it differs from work, from ritual, or from games. Why a given kind of play is fun for these people now, and is no fun for others or for the same people later, has usually been left to taste, as though it lay beneath the reach of theory. This article takes the opposite view. The variability of fun is the best evidence there is about what play is, and the fact that play is so much better with some people than with others is among the best evidence about what being with others can do.

Three questions are kept apart throughout, because they receive different answers far more often than is usually noticed. The first is whether an activity is organised as play: whether there is a game, an invitation, a recognised frame. The second is whether those taking part find it worth continuing for its own sake, which is roughly what English means by fun. The third is whether it sustains or damages what is vital to them. A compulsory team-building exercise is organised as play and is rarely fun. A slot machine can hold a player for hours and is bad for her. Two colleagues doing the washing-up after a conference dinner may be having the best time of anyone in the building.

The territory is far larger than it first appears, and this article is a first survey of it. It reviews the classic theories of play and games, mainly from the social sciences, and asks whether a literature exists on the narrower questions of fun and of playmates. It then introduces a Living Value Theory account built on three claims. Play is a distinctive way in which living coordination is opened, varied, maintained, and enjoyed. Games are made arrangements that hold some of the conditions for such coordination steady without guaranteeing that anyone will enjoy taking them up. And actual playing can establish a temporary world of its own, within which what is relevant, what is at stake, and even whether this is play at all are continually being settled. The sections that follow develop each claim through cases, sort the kinds of play and game, explain why a single theory keeps failing, follow shared worlds of play to their end, and close with one stretch of the life course, the adolescent passage from playing to hanging out, which raises a question about stakes that is left for separate treatment.

II. Surplus, Practice, Development

Modern theorising about play begins with a compliment. In the fifteenth of his letters on aesthetic education, published in 1795, Friedrich Schiller (1967) proposed a play drive that reconciles sense and form, and declared that human beings are fully human only when they play. The nineteenth century turned the compliment into physiology. Herbert Spencer (1872) took play to be the discharge of surplus energy in animals whose needs are met too easily to exhaust them. Karl Groos (1898, 1901), observing that the young of the most capable species play the most, reversed the direction of explanation: play is practice for adult life, and animals do not play because they are young so much as have a youth in order to play. G. Stanley Hall (1904), in the book that gave adolescence its modern name, added the recapitulation thesis, on which the playing child rehearses the ancestral occupations of the species.

The twentieth century kept the developmental frame and changed its content. Sigmund Freud (1955) watched his grandson throw a cotton reel out of sight and haul it back, and read the game as the mastery of an absence the child could not otherwise control. Jean Piaget (1951) treated play as the predominance of assimilation over accommodation, the child bending the world to existing schemes, and ordered it into a sequence: sensorimotor practice play, then symbolic play, then games with rules, the last arriving at around seven and consolidating by eleven or twelve. His earlier study of boys playing marbles (Piaget 1932) showed younger children treating the rules as sacred and unalterable, and older ones treating them as agreements that players may change by mutual consent. Lev Vygotsky (1967) argued that play creates an imaginary situation governed by rules, that the stick ridden as a horse is the pivot by which meaning first separates from object, and that the child in play stands “a head taller than himself.” Erik Erikson (1950) saw the child using play to model and master experience, and later described adolescence as a psychosocial moratorium in which roles may be tried without being owed (Erikson 1968). D. W. Winnicott (1971) located playing in a potential space between mother and infant, neither wholly inside nor wholly outside, from which cultural experience as a whole later grows.

These accounts differ on almost everything except one assumption: play is explained by what it later yields. Brian Sutton-Smith (1997) called this the rhetoric of progress and noted how well it suits the adults whose business is to bring children up. It is also worth noticing where the stage theories stop. Piaget’s sequence ends with games with rules at about the age of twelve, and the developmental theory of play has little to say about anyone older.

III. Huizinga and Caillois

The founding text of the cultural study of play refuses the developmental question altogether. For Johan Huizinga (1949), writing in 1938, play is older than culture, since animals did not wait for human beings to teach them to play, and it cannot be reduced to any biological function without losing the fun that makes it what it is. He added that this fun resists analysis and logical interpretation, and that the English word has no exact equivalent in the neighbouring languages he knew. He characterised play as a free activity, standing outside ordinary life, absorbing the player utterly, connected with no material interest, proceeding within its own boundaries of time and space according to fixed rules, and promoting the formation of communities that surround themselves with secrecy. The playground, the card table, the stage, the court of justice, and the magic circle are all, on his account, temporary worlds within the ordinary one. Huizinga’s boldest thesis was that civilisation arises in and as play: law, war, poetry, and philosophy begin as contests and performances and retain a play element long after they have grown solemn. He also observed that the cheat, who pretends to keep the rules, is forgiven more easily than the spoilsport, who walks away and so reveals how fragile the play world was.

Roger Caillois (1961) admired the book and thought it incomplete. Huizinga had defined play by its spirit and ignored its kinds, and by excluding material interest he had no place for gambling, which involves a great deal of it. Caillois kept a formal definition (play is free, separate, uncertain, unproductive, rule-governed or make-believe) and added a classification. Games fall into four families: agon, competition under artificially equal conditions; alea, the surrender to chance; mimicry, simulation and the wearing of masks; and ilinx, the pursuit of vertigo in swings, spinning, and speed. Across these runs a second axis, from paidia, the turbulent and improvised, to ludus, the taste for arbitrary difficulty and binding convention.

In the same years, American anthropology produced a narrower and more testable typology. John Roberts, Malcolm Arth, and Robert Bush (1959) defined a game as organised play involving competition between sides, criteria for determining a winner, and agreed rules, sorted the games of some fifty societies into those of physical skill, strategy, and chance. The definition was deliberately tight, and its tightness had a cost: a great deal of what children everywhere do together fell outside it as mere amusement.

IV. Frames, Selves, Stakes

A second lineage treats play less as a kind of activity than as a relation between activities. George Herbert Mead (1934) distinguished play from the game as two stages in the making of a self. The playing child takes the role of particular others one at a time, mother, shopkeeper, policeman. The child in a game such as baseball must hold the attitudes of everyone on the field at once, and this organised set of attitudes, the generalised other, is the form in which a community enters the individual.

Gregory Bateson (1955) reached the most influential formulation by watching young monkeys at a zoo. Their play-fighting resembled combat closely enough that it could only proceed if the animals were somehow signalling to each other that it was not combat. Play therefore requires metacommunication, a message about messages, and the message is paradoxical: the playful nip stands for the bite without carrying what a bite would carry. Bateson called the device that sustains the paradox a frame. Erving Goffman turned the insight into a sociology. In his essay on fun in games (Goffman 1961), an encounter is surrounded by a membrane that filters the wider world through rules of irrelevance, so that the cost of the chess pieces or the rank of the players is not allowed to count, and fun arises when the tension inside the membrane is neither slack nor overwhelming. In his later work (Goffman 1974), play became the model for keying, the transformation by which a strip of activity already meaningful in one frame is performed as something else.

Georg Simmel (1949), earlier than either, had described sociability as the play-form of association: talk pursued for its own sake, in which the content matters only as the carrier of the pleasure of being together, with coquetry as the play-form of the erotic. Clifford Geertz (1972) took up the opposite extreme. In the Balinese cockfight, men wager sums so large that, by Bentham’s reasoning, it is irrational to play at all, and Geertz read this deep play as a dramatisation of status, a story the participants tell themselves about themselves. Claude Lévi-Strauss (1966) contrasted the two directly: a game begins from symmetry and engineers a difference between winners and losers, whereas ritual begins from asymmetry and conjoins, and he cited the Gahuku-Gama of New Guinea, who adopted football and played as many matches as were needed for both sides to finish level. Mihaly Csikszentmihalyi (1975) added the phenomenology, describing the absorbed state he called flow, in which challenge and skill are matched and the activity becomes its own reward.

Two further contributions bear directly on what happens between players. Hans-Georg Gadamer (1989) argued that the subject of play is the play itself rather than those who play it. Play is a to-and-fro movement with no goal that would bring it to an end, it renews itself in repetition, and it takes up those who enter it, so that all playing is also a being played. Hanne De Jaegher and Ezequiel Di Paolo (2007) gave a related thought an explicit form in their account of participatory sense-making. When two agents coordinate, the interaction can acquire a dynamics of its own, which neither controls and which alters what each of them can make sense of. Their example is the encounter in a narrow corridor in which two people trying to pass each other step repeatedly to the same side, held for a moment by a pattern that neither wants. Both accounts locate something essential in what goes on between the participants, and neither says much about why one such encounter is delightful and the next one dull.

What unites this lineage is the insight that the same movements can be play or not play depending on a framing that the participants accomplish between them. It is the lineage most useful for the questions pursued here, because sociability, teasing, and status are its subjects, and it is also the one most inclined to treat the frame as a message.

V. The Ambiguity Problem

By the middle of the twentieth century it was already suspected that these theories might not share an object. Ludwig Wittgenstein (1953) used games as his example of a concept without an essence. Board games, card games, ball games, and ring-a-ring-a-roses share no single feature, only a network of overlapping family resemblances, and anyone who insists there must be something common to them all is asked to look and see. Bernard Suits (1978) took up the challenge and produced the most elegant definition on offer: playing a game is “the voluntary attempt to overcome unnecessary obstacles,” undertaken in a lusory attitude that accepts the rules because they make the activity possible. The definition handles games well and play poorly, since a kitten with a ball of wool accepts no rule.

Sutton-Smith (1997) concluded that the disagreement lay in the theorists as much as in the phenomenon. Scholarship on play, he argued, is organised by seven rhetorics, each with its own constituency: play as progress, as fate, as power, as identity, as the imaginary, as the self, and as frivolity. Educators hear progress, mathematicians and gamblers hear fate, historians of sport hear power, folklorists hear identity, and each rhetoric finds the evidence it needs. On the biological side, Gordon Burghardt (2005) gave up on a definition by essence and offered five criteria to be met together: play is not fully functional in the context where it appears, is spontaneous or rewarding, differs from the serious version of the behaviour in form or timing, is repeated without being stereotyped, and occurs when the animal is fed, healthy, and free of acute stress. With these in hand he found play well beyond mammals and birds.

Anthropologists of childhood complicated the picture from another side. Helen Schwartzman (1978) showed how far each school had seen in children’s play only what its own theory required. David Lancy (2008) and Suzanne Gaskins and colleagues (2007) documented how much the amount, the kind, and the adult valuation of play vary: parents in many societies neither play with children nor think play important, and in much of the world play and work are interleaved through the same day. David Graeber (2015) drew the sharpest version of a distinction that English makes and many languages do not. Play, in his sense, is free creative improvisation and can generate rules, while a game is governed by rules, and the appeal of games lies in their offering the one situation in which everyone knows exactly what the rules are.

The field thus presents a familiar condition. Each theory is persuasive about some play, none fits all of it, and the accumulation of cases has made a common definition less likely over time.

VI. Is There a Literature on Fun?

It would be surprising if nobody had asked why some play is fun and some is not, or why it matters so much who else is playing. Several literatures have asked, each from its own corner, and the corners have rarely been joined.

The most developed is the study of game design and of players. Raph Koster (2005) proposed that fun is the feeling of learning a pattern, so that a game stops being fun at the point where it has nothing left to teach, which is the fate of noughts and crosses. Nicole Lazzaro (2004) distinguished four routes to enjoyment, the hard fun of challenge, the easy fun of exploration, the serious fun of altered states, and the people fun of company, and Robin Hunicke, Marc LeBlanc, and Robert Zubek (2004) listed eight kinds of fun in an influential design framework. Richard Ryan, Scott Rigby, and Andrew Przybylski (2006) explained the pull of video games by the satisfaction of needs for competence, autonomy, and relatedness. Richard Bartle (1996) and Nick Yee (2006) sorted players by what they seek, whether achievement, exploration, company, or immersion. Jesper Juul (2013) examined the paradox that players seek out games in which they will mostly fail, John Sharp and David Thomas (2019) argued that fun deserves an aesthetics of its own, and C. Thi Nguyen (2020) described games as a medium whose material is agency itself. These accounts are rich on the properties of games and on the dispositions of types of player. Their unit is the game or the player, and they have much less to say about why a given game catches a given person in one season and leaves her alone in the next.

A second and much smaller literature is sociological. Goffman’s (1961) essay remains its starting point. Martha Wolfenstein (1951) noticed fun turning from a permission into an obligation in mid-century American advice on bringing up children. Walter Podilchak (1991) distinguished fun from enjoyment and from leisure, and tied it to interaction among people who treat one another for the moment as equals. Ben Fincham (2016) built a book-length sociology of fun on people’s own accounts of it, and stressed the oddity that something treated as unserious is also treated as essential to a good life. Gary Alan Fine and Ugo Corte (2017) argued that fun belongs to small groups: shared pleasure commits members to one another and gives the group a past worth retelling. Here the unit is the group, and fun is explained chiefly by what it does for group life.

A third literature is psychological and recent. Harry Reis and colleagues (2017) found that fun was more frequent and more intense in company, and in an experiment with the stacking game Jenga that playing with another person was more fun than playing alone, and more fun with a friend than with a stranger. They also remarked on how little research there was to build on. Charles Walker (2010) found flow shared with others more joyful than flow achieved alone. Reis, Annie Regan, and Sonja Lyubomirsky (2022) proposed a model of interpersonal chemistry, the sense that an interaction amounts to more than the sum of its parts, and Emma Templeton and colleagues (2022) showed that people who feel that they click answer one another faster in conversation, among strangers and among friends.

Developmental and comparative work supplies the remainder. Mildred Parten (1932) distinguished children who play beside one another from children who play with one another. John Gottman (1983) brought unacquainted children together in one another’s homes and found that the pairs who hit it off were those who managed to establish and sustain a shared activity. William Corsaro (1985) showed nursery children defending an established game against newcomers, on the sound intuition that the game was fragile. Students of animal play have put the partner question most directly. Erika Bauer and Barbara Smuts (2007) found that pairs of dogs went on playing together despite marked asymmetries between them, against the proposal that each must come out on top about half the time. Jackson Ham and Sergio Pellis (2024, 2025) found that juvenile rats, given a choice among several companions, play more with some than with others, that the favoured partners tend to be those who take turns and with whom play is more symmetrical, and that the preferences shift from one day to the next.

So there is a literature, and it supports three conclusions. Fun is real enough to study and to measure. It is strongly social, in species without symbols as well as in our own. And who the partner is matters, down to the individual. What is missing is an account that holds the pieces together. The design literature locates fun in the game or in the type of player, the sociology locates it in the group, and the psychology treats chemistry as a property of pairs, so that the three puzzles set out at the start are divided among fields that seldom cite one another: the game that catches at one time of life and at no other, the game that works with one person and not with the next, and the shared enthusiasm that ends. On the last of these almost nothing has been written. A claim of absence deserves caution, since no search is complete, but the pattern is clear enough to justify an attempt at joining the pieces.

VII. Play as Sought Opening

Living Value Theory begins from a single principle: life is recursive mediation in a mesocosm. Living beings coordinate continuously with their own bodies, with other living beings, and with the found conditions of the world, and human beings coordinate in addition with made things and with symbols. Most of this coordination is muted. It proceeds without becoming separately salient, which is the first level of recursivity, L1. When something in the texture of coordination changes, through disturbance, discrepancy, uncertainty, or attraction, the coordination becomes salient, which is L2, and may then be articulated in symbols, stabilised into portable categories, and reflected upon, which are L3, L4, and L5. The theory calls the movement by which a settled coordination becomes available to recursion again a recursive opening, and distinguishes openings that are disruptive, forced by a breakdown, from those that are induced by a party with standing to bring them about and those that are scheduled in advance.

Most opening in ordinary life is remedial. Something goes wrong, salience rises, repair is attempted, and the aim is to resettle into L1 with better coordination than before. Play is the case in which nothing went wrong. A child walking home leaves the pavement, which she can walk without attention, and climbs onto a low wall, where every step has to be found. A dog that has been lying peacefully beside another dog rises, drops its forequarters, and bounces away. Two friends who can talk to each other without effort begin trading insults of increasing absurdity. In each case a coordination that was working has been opened on purpose by those inside it, and the L2 that follows, the wobble, the chase, the uncertainty about what the other will say next, is the object of the exercise.

Play can accordingly be characterised, in the first instance, as the sought opening of working coordination. Three consequences follow. Play is neither a mediation nor a level of recursivity nor a type of counterpart, so that it has no fixed address in the architecture and is better understood as a way of moving through it; it can occur in any mediation, with any type of counterpart, at any level. Play is also an exercise of what the theory identifies as vitality, which consists in recursive fluidity, the capacity of settled coordination to stay muted where it works and to reopen when disturbance, curiosity, or attraction makes further recursion valuable, rather than in remaining at L1. And the salience that play seeks is positively valenced and held within a band: enough discrepancy to keep the coordination from sinking back into L1, and not so much that it tips into breakdown. Csikszentmihalyi’s balance of challenge and skill describes the same band from within, and boredom and fright mark its two edges.

The formula needs a qualification at once, because taken alone it suggests that play is a state of continuous heightened salience, and a great deal of play is nothing of the kind. Two musicians who have played a duet together for years, two children running a make-believe sequence for the hundredth time, and a doubles pair moving about a court without a word are all enjoying how well the coordination runs. They know what is coming, each entry arrives where it should, and much of the pleasure lies in the fit. An opening established the form, and openings remain possible inside it at every moment, a small delay, an exaggeration, a look, but the playing itself moves between stretches of fluent L1, local flares of L2, an occasional L3 negotiation over how the thing should go, and back again. Play is therefore better described as a mode of coordination in which a partly secured ground supports variations, repetitions, and possible openings that those involved find worth sustaining. The sought opening is what starts it and what stays available within it. This agrees with the theory’s account of vitality, which gives no priority to being permanently open, and it agrees with something every player knows, which is that a game that had to be exciting at every second would be exhausting.

The characterisation also recovers what was right in the oldest theories without their teleology. Felt disturbance draws on metabolic reserves and cannot be sustained indefinitely, which is why Burghardt’s relaxed field is a condition of play, why play is the first thing to vanish under hunger and threat, and why it concentrates in the provisioned young. Spencer’s surplus was a real observation about the cost of L2. Groos’s practice was a real observation about what repeated opening leaves behind: a living being that has opened its own coordination many times, in many directions, holds a wider repertoire of ways to recoordinate, which is the redundancy that protects any stake against local failure. Neither observation requires that play exist in order to produce these results. Play is what recursive fluidity looks like when it has reserves to spend.

VIII. What the Nip Does

Opening a working coordination is dangerous if the stakes of that coordination stay where they were. In Living Value Theory, stakes are mesocosmic facts about what would be lost if a coordination failed, as distinct from symbolic assessments of risk, and they are distributed among parties, usually unevenly. A chase is exciting to open only if being caught does not mean being eaten. Play therefore depends on a reallocation of stakes, and because stakes are facts, the reallocation has to be made in fact. Declaring it is not enough.

This is where Bateson’s formulation needs amending. The playful nip is a different coordination from a bite before it is a message about one: the jaw is inhibited, the stronger animal handicaps itself, roles reverse so that the chaser becomes the chased, and the bout is punctuated by bows and pauses that let either party withdraw (Bekoff and Byers 1998). All of this is accomplished within embodiment and being-with at L1 and L2. To describe the play bow as saying “this is play” is to read a sentence back into a coordination that contains none, the retrospective colonising of L2 by language that the theory warns against. The frame is first of all a set of real constraints on what bodies do to each other, and only among symbol-using players does it also become something that can be announced.

Human play extends the same reallocation through the poietic mediations. Multimateriality supplies blunted foils, padded gloves, counters in place of coins, and a ball in place of a spear. Multisymbolization supplies the score, which converts an outcome into a number that can be forgotten by evening. In each case the embodiment stake is buffered, and the opening can be explored further than it otherwise could.

Play does not thereby become free of stakes. The theory holds that the most fundamental stake of all is the capacity to continue coordinating, and play brings this stake to the surface, because once the stakes of the content have been lowered, what remains at risk is the play itself and the relation that carries it. The gravest sanction available to a child is the announcement that she is not playing with someone any more. Huizinga’s spoilsport is worse than his cheat for the same reason: the cheat threatens only the outcome, while the spoilsport withdraws the coordination. The being-with stake stays fully live throughout play, and playing is one of the principal ways in which it is met, so that friendships are made of play rather than merely expressed in it.

Reallocation also explains why the boundary between play and seriousness cannot be drawn once and for all. Stakes can be reattached, deliberately and to any degree. In Geertz’s cockfight, large sums and larger reputations are attached to an outcome that the participants take elaborate care to keep uncertain, by matching the birds, calibrating the spurs, and balancing the bets. The result reads less like a text than like a scheduled opening in which standing is put at risk on purpose. Professional sport reattaches livelihoods, gambling reattaches money, and a duel reattaches the body. Whether any of these is still play depends on how the stakes are allocated on a given occasion and for whom, which is why the same match can be play for the crowd and work for the players.

One further distinction prepares what follows. The allocation of stakes is one thing and the participants’ discernment of that allocation is another. Because stakes are facts, a player can be wrong about them, taking for buffered what is fully live or fearing for her standing in an episode that was staged from start to finish. A play frame can therefore be unevenly distributed among the people inside it, held by some and missed by others, and it can become the very thing they are coordinating over.

A fuller statement can now be given. Play is a mode of living coordination in which living beings make variations in their ongoing coordination available for positively valenced uptake, under an allocation of stakes that allows the activity to be sustained, repeated, and transformed. Its value need not lie in any result beyond the playing, although play regularly alters the capacities, the relations, and the deposits from which later coordination proceeds. What the statement leaves open is the matter of uptake: why one variation is taken up with delight and another falls flat.

IX. Is This Play?

In most games the players know that they are playing, and what is uncertain is what will happen inside the game. There is a second kind of uncertainty, about what sort of situation this is, and a whole class of play is built on it. Bateson (1955) had already noticed that some play is organised around the question of whether it is play, as opposed to the premise that it is, and he had hazing and initiation in mind. Two films turn the question into a plot. In David Fincher’s The Game (1997), a wealthy banker is given a mysterious entertainment as a birthday present, and from then on cannot tell which of the things happening to him belong to the entertainment and which threaten his money and his life. In John Francis Daley and Jonathan Goldstein’s Game Night (2018), a staged kidnapping at a murder-mystery party is overtaken by a real one, the guests go on treating armed criminals as hired actors, and the frame turns over more than once before the end. A chess player asks what her opponent will do next. The people in these films have to ask what they are taking part in.

The two uncertainties are independent of each other. A game can have perfectly clear rules and an open outcome, and an episode can consist of perfectly intelligible actions whose framing is in doubt. This yields a distinction between playing within a frame and playing with the frame, and the second covers a large and familiar territory: the practical joke, the prank, the hoax, the surprise party, the confidence trick, the bluff that may or may not be a bluff, and a good deal of flirting. In each of them the actual allocation of stakes and somebody’s discernment of it have been deliberately separated. The results can be delightful, when the frame is revealed and the stakes turn out to have been buffered all along, and they can be disastrous. What separates a good prank from a cruelty is whether the stakes of the person kept in the dark really were buffered, and whether she would, on learning the frame, have agreed to the opening.

Discerning frames is a skill, and exercising it is one of the keenest pleasures that play offers. A chess player sees that an innocent developing move is the first step of a trap. A poker player sees that a show of hesitation is calculated. A friend sees that an insult is affection. The pleasure in each case is that of cleverness, which can be characterised as the positively valenced detection of an unsuspected recursive relation that alters the apparent possibilities of a coordination. It is sharpest when the other party has not yet seen what one has seen, and it needs no words.

Julia Donaldson’s The Gruffalo (1999) is a primer in it. A mouse walking through a wood is invited to lunch by a fox, to tea by an owl, and to a feast by a snake. The mouse discerns what each invitation is and declines it by inventing a monstrous friend whose favourite food happens to be the animal doing the inviting. When the monster turns out to exist, the mouse manages him as well, by walking him past the three predators and letting him draw the wrong conclusion from their terror. The mouse is not playing, since for the mouse the stake is being eaten. The play belongs to the child who is being read to, and who enjoys, from a position of complete safety, both seeing through each invitation and seeing that the predators have not seen through the mouse. A good deal of children’s literature, and most games of bluff, exercise the same capacity.

It is tempting to say that discerning a frame requires a higher level of recursivity than acting within one, and the temptation should be resisted. The levels of recursivity are an architecture of how coordination becomes salient, articulated, stabilised, and reconsidered. They do not count how many perspectives a player can nest, and being one move ahead of an opponent does not place anyone one level up. A dog that checks, after too hard a bite, whether the other dog still takes this as play is discerning a frame entirely within L1 and L2. Among symbol users, frames can also be named and classified. The remark that something is only a game is an L4 classification, the discovery that it was no game is a second classification at the same level, and L5 is reached only when the grounds of classifying are themselves put in question, as when people ask whether a game of this kind ought to exist or who is entitled to declare what counts as one. The films mentioned above entertain by forcing reclassification again and again.

Bateson saw that the frame could be played with. What the present account adds is attention to the living coordinations and the actual stakes involved in producing an ambiguous frame, sustaining it, detecting it, and getting it wrong.

X. Why Play Catches

Nobody comes to a game empty. A player brings a sedimented field of anticipation: how a ball of this weight comes off a wall, what bodies of this size can do, how this opponent tends to behave under pressure, how a joke of this kind usually ends. The field operates at L1, as ground, and none of it has to be held in mind. It has layers of different depth. Some regularities are very general and widely shared, such as the expectation that a thing thrown will come down or that a steady rhythm will continue. Some have been sedimented over years within one practice, as in a footballer’s sense of where the ball is about to be. And some are local to a single game, set up within minutes and exploited soon afterwards: in this game of tag the tree is safe, and everybody knows it by the third round.

Against such a field a variation can register, and the concept needed for what happens next is recursive resonance: a positively valenced, sufficiently constrained felt misalignment, generated within a sedimented anticipatory field. Each term does work. The misalignment is felt, which places it at L2. It is constrained, which separates the feint that almost comes off from the fire alarm that goes off in the middle of a game, an event far more surprising and no fun at all. It is positively valenced as it arrives, without waiting to be judged. And it needs a field precise enough to be departed from in one place and not everywhere at once. The study of musical expectation has documented the general mechanism in detail (Huron 2006). Play adds a complication that music lacks, since the player produces half the events herself.

Several consequences follow. The first is that familiarity and surprise are allies. A newcomer to cricket sees a succession of similar events, and a lifelong follower sees a ball that did something balls of that kind almost never do. The expert has more to be surprised by, because her field is differentiated enough for a small deviation to register as an event. The play of infants shows the mechanism at its simplest. In peekaboo the delight depends on the face returning a little later, or from a slightly different place, than it did the last time (Bruner and Sherwood 1976), and Daniel Stern (1977) described the games of mothers and infants generally as a theme with variations, in which the adult repeats a sequence while altering its timing by small amounts and the infant’s excitement tracks the alterations. The child’s demand to do it again is a demand for the same thing arriving once more, and knowing exactly what will happen turns out to be quite different from living through its arrival.

The second consequence is that resonance is only one of the pleasures of play. A long rally kept going for its own sake, a skipping rope turning, a game of patience laid out on a tray, a team passing the ball around the back with nothing at stake: these please through continued alignment and absorption, with very little misalignment in them. Flow, in Csikszentmihalyi’s sense, covers part of this ground. The distinction matters because people want different things from play at different moments, and a theory that recognised only surprise would misdescribe half of what they do.

The third consequence is that capacity is distinct from resonance. Recursive capacity concerns which coordinations have become available to a player: what she can see, anticipate, and execute. Resonance concerns whether taking them up is felt as worth it. A strong chess player can be bored by chess, and a weak one can be captivated. Years of coaching regularly produce adults who never touch the sport again. Koster’s account of fun as the learning of patterns comes close to identifying fun with growth in capacity, which explains the abandoned game of noughts and crosses and cannot explain the master who has loved the same game for sixty years or the learner who is improving steadily and would rather be anywhere else.

The fourth is that tastes in play behave like strata. A new enthusiasm is laid down beside the old ones without cancelling them, each layer keeps its own history of acquisition, and at any time some layers are active while others lie dormant. The adult who has not played marbles for forty years can still feel the pull of a good shot. Because the layers were laid down by a particular sequence of encounters, they are singular to the person. Two players of equal skill, from the same street and the same school, meet the same game with different fields: one had a grandfather who taught her cards on wet afternoons, and the other remembers being picked last. The same uncertainty can be exquisite for the first and of a kind the second learned long ago to dislike.

All of this concerns what the player has accumulated. The other half of the answer concerns her present condition, since there is no enjoying a game apart from the living coordination in which the enjoying takes place. A demanding game of strategy that is a pleasure on a Saturday morning can be intolerable at the end of an exhausting day, when something repetitive and nearly effortless is exactly right. Felt salience has a metabolic cost, and a player whose reserves are spent is offered a different coordination by the same box of pieces. The effect extends beyond tiredness. A game full of uncertainty appeals when the rest of life has become too predictable and repels when life is already saturated with uncertainty nobody asked for. Rules of perfect clarity are attractive when relationships or work have become hard to read. An exuberant physical game is welcome after long confinement, and a solitary puzzle after a week of meetings. Psychology has approached this from two sides. Michael Apter (1991) described people as switching between a state in which arousal is unpleasant and goals are serious and a state in which the same arousal is enjoyed within what he called a protective frame, and Dolf Zillmann (1988) showed people choosing their entertainment so as to correct the mood they were in.

It would be a mistake, though, to turn this into a rule. Sometimes a game catches for no compensatory reason whatever. Its patterned opportunities for action happen to meet capacities and anticipations that a singular history has laid down, and that is all there is to say. A theory in which every enjoyment existed to correct a deficit would bring back the rhetoric of progress at the scale of a single afternoon. The general proposition is more modest and more useful: the value of a game is no fixed property of the deposit. It emerges in the changing relation among the deposit, its uptake, the player’s mesocosmic history, and the present allocation of stakes.

Resonance is an event of seconds, and a further concept is needed for the longer phenomenon. A game catches when episodes of resonance lead a player to seek further encounters, and when the further play builds the anticipatory conditions for more resonance. The sequence runs from mere availability through a first compelling uptake, repeated resonance, and the sedimentation of a more differentiated field, to rising recursive relevance, as more of a life comes to be organised around the game, and then to continuation, withdrawal, or a later restoration. Repetition is the engine of this process. Children who play the same chasing game every afternoon are, on a superficial description, doing the same thing again and again. Within the repetition they are learning one another’s feints, favourite hiding places, and habits of cheating, and a variation that a newcomer would not even notice becomes exquisitely funny to those who have been there all month. The game grows more predictable at one level and richer in possibilities at another.

Two familiar facts follow from this. Established players often cannot recreate their enjoyment with newcomers, because the rules can be explained in five minutes and the accumulated field of mutual anticipation cannot be explained at all. And a game can stop catching while remaining exactly what it was, because the coordinations that made its variations matter have lost their relevance. Three questions about any such history should be kept apart: why a given enthusiasm was possible, how it in fact developed, and whether it could have been predicted. The theory aims to answer the first two. Where recursive beings are concerned, an inventory of prior conditions does not yield the third.

XI. The Year of the Magic Flute

A case from my own household brings these elements together, and since it involves two players it also prepares the next step. When my children were four and six they converged, for no reason anyone could identify, on Mozart’s The Magic Flute. They asked for it night after night. They listened to it again and again, sang the arias to each other, performed scenes with props improvised from whatever was lying about the house, and watched several filmed productions closely enough to tell them apart and to argue over which Queen of the Night and which Papageno were the best. They once sat through an entire visit to the barber watching it on a tablet propped against the mirror. After something between nine months and a year, it stopped. Nobody decided that it should. Neither child became a lover of opera, and the two of them have never again converged so completely on anything.

The episode can be told as a story about musical taste. Here it matters as a case of play. A recording is a nonrecursive deposit, which does not alter however often it is played and answers to nobody’s excitement. On that fixed ground the children built coordinations in every mediation: their own voices and bodies in the singing, household objects pressed into service as props, roles and names taken from a libretto in a language neither of them yet spoke, and above all each other. The opera became a world they could enter together, leave, and re-enter, with characters, situations, favourite moments, and a repertoire of things to do in it. One child’s anticipation of a scene was recognised by the other, who joined the anticipation before the scene arrived. A fixed musical structure made possible an increasingly elaborate shared play, which is the general pattern in which a nonrecursive deposit serves interrecursive vitality.

Two processes were at work and each fed the other. Each child was resonating with the music against a field that concentrated, voluntary, daily repetition had sedimented in months. And the responses circulated between them, so that each child’s attachment became part of what the other was attached to. They did not first love Mozart and then, as a separate matter, play together. The playing deepened the loving, and the loving supplied the playing with material.

The household contributed access, good recordings, time, and permission. It could not have contributed the convergence, and no household could. This is the first appearance of a principle to which the argument returns: the conditions under which resonance may be acquired can be arranged, and resonance cannot be handed over.

The ending is as instructive as the beginning. One child’s interest may have weakened first and removed something the other had valued. Other pursuits may have become more relevant to both at about the same time. Or the shared repertoire may have sedimented so thoroughly that it stopped calling for reopening. I do not know which of these occurred, and a theory should be able to distinguish the possibilities without pretending to choose among them after the fact. What the case establishes is that a shared playworld can dissolve while its deposit remains in perfect condition. The recordings still exist, and the opera is as good as it was.

XII. When Play Becomes Interrecursive

The simplest playful pleasure there is cannot be had alone. People cannot tickle themselves, because the sensory consequences of one’s own movement are anticipated and damped (Blakemore, Wolpert, and Frith 1998). Robert Provine (2000) found laughter to be some thirty times more frequent in company than in solitude. Young rats that have been tickled by an experimenter emit ultrasonic chirps and come back to the hand for more (Panksepp and Burgdorf 2003). With a nonrecursive counterpart, variation comes from one’s own action meeting something indifferent: the stone skips across the water or it sinks. With an interrecursive counterpart, variation comes from a second source of anticipation, one that is reading the first.

Consider two friends at a table tennis table. At first each is trying to return the ball. If they are well matched, something else can begin. One places a ball in a way that invites an especially difficult return. The other recognises the invitation and answers with a shot that is hard to reach and just possible. The first is delighted and tries something more adventurous still. A rally develops that neither could have planned, and what makes it enjoyable can no longer be assigned to either player’s performance or to the rules of table tennis. The play is being generated between them, and each player’s enjoyment depends partly on discovering what the other makes possible.

What distinguishes this from other successful coordination is the use to which success is put. Ordinary cooperation tends to settle uncertainty. Colleagues who work well together can anticipate one another, less and less needs to be negotiated, and the coordination mutes. Play does something peculiar with the same achievement, using a capacity for coordination to introduce further variations that invite new responses. The better two players read each other, the more ways they have of surprising each other in ways that can be appreciated and need not be felt as disruption. The point can be given a general form. In interrecursive play, the coordination achieved between living beings becomes the condition for further variation in that very coordination. Each participant may alter what the others can anticipate, do, and enjoy, while preserving enough common ground for the interaction to continue.

A variation offered to another player is an invitation, and its uptake answers the invitation as well as the ball. Each player becomes responsive to how the other is responding to the playing itself: the laugh, the raised eyebrow, the escalation. The other’s delight is part of what each is enjoying, so that the pleasure runs round a loop and cannot be divided into two private portions and added up. This is what Reis and his colleagues mean by an interaction that exceeds the sum of its contributions, what the quickened replies in Templeton’s conversations register, and presumably why Walker found shared flow more joyful than the solitary kind.

The word “may” in the formulation matters, because much social play never reaches this point. Parten’s children playing side by side with the same bricks are engaged in the same activity without being engaged with each other. Two adults can complete a game of cards without once finding a rhythm. Taking part in the same game and achieving a jointly compelling form of play are different things, and only the second is at issue when people say that they love playing together.

The comparison with the Magic Flute sharpens what is special here. A recording is stable ground around which two people’s responses can begin to respond to each other. In social play the ground itself moves, since the players keep changing the very activity with which the others are resonating. Each is a source of the other’s constrained deviations and is at the same time resonating with deviations that come back. A good playmate therefore does something no deposit can do, which is to produce variations fitted in real time to this player’s anticipatory field, neither too wild nor too tame for her. A designed game can only anticipate a player in general. This is why play with others has the highest ceiling of any play, and also why it varies most, since the other’s variations can miss in either direction and an evening can go flat with nobody at fault.

Gadamer and the theorists of participatory sense-making saw the central fact: the play takes up the players, and the interaction acquires a dynamics that neither commands. The present account can say more about the conditions. It specifies the type of counterpart, which separates resonating beside someone to a recording from resonating with her moves. It specifies the mediations in which a given form of play runs, from embodied timing to equipment to a running commentary. It asks how the stakes are allocated and how the play moves among the levels, from the L1 of the flowing rally through the L2 flare of a near miss to the L3 moment in which one player asks whether the other meant that. And it asks about relevance, which is the question of why this encounter matters to these two now. The animal evidence fits the same picture. What Ham and Pellis’s rats prefer in a partner is turn-taking, and what Bauer and Smuts’s dogs tolerate is inequality, which suggests that the condition for continued play is reciprocity of opportunity and that equality of outcome is dispensable.

XIII. Why the Right People Matter

When two people play well together, one of three things may be going on, and they are easily confused. The two may independently resonate with the same game. They share a taste, which is no guarantee that they enjoy playing each other, and two keen chess players can have a thoroughly unpleasant match. Or the two may simply enjoy being together, and therefore enjoy a game that neither much values. The game is a vehicle for being-with and is almost incidental, as when a grandmother and a five-year-old play snakes and ladders. Or, thirdly, a form of play may emerge that neither could have enjoyed in quite that way without the other.

The third is the most revealing. Two friends discover that a particular kind of teasing works between them. One makes an absurd remark, the other understands at once and develops the absurdity, and the first recognises that the joke has been improved as well as understood. Before long a small world exists between them, made of references, voices, rules, and expectations that mean nothing to anyone else and are enormously funny to the two of them. Fine (1979) called the stock of jokes, nicknames, and customs peculiar to a small group its idioculture, and Leslie Baxter (1992) catalogued the corresponding repertoire in friendships and couples, the private idioms, mock quarrels, and bits of role-play, and found it associated with closeness. Neither friend possessed that world beforehand. It is an interrecursive achievement in the full sense.

Something further happens in such cases that theories of taste have found hard to accommodate. The participants discover their own capacities for enjoyment through the responses of another person. One does not simply find somebody who shares a preference one already had. Through the shared play one comes to have a preference that did not previously exist in that form, and may never exist in it with anyone else. The typologies of players in the design literature assume that tastes are possessed before any encounter and then matched to games and to company. A good part of what people most enjoy was produced in the playing.

This has a consequence for prediction. Whether two people will play well together looks like a fact about the pair that is waiting to be discovered, and it is no such thing. A mango does not know that it has been preferred to an apple, and a board game is not improved by somebody’s considering it. A playmate is another evaluator. Her enjoyment of playing with me enters into my enjoyment of playing with her, my sense of how she is finding it alters how I play, and she can tell. The fit is generated by a process that changes both parties as it goes, so that no inventory of their separate traits settles the matter beforehand. This is why the confident prediction that two people will get on because they both love chess fails so regularly, and why two children matched by their parents for age and interests can stand in a doorway with nothing to say. People can be right for each other on paper and flat at the table, and the reverse. The same structure governs the choice of lovers, and flirtation is the place where the two coincide, being play whose stake is the relation itself.

Fit does not require similarity. A confident child who loves inventing elaborate games may play wonderfully with a quieter one who is happy to accept a role and then adds a detail nobody expected. Two people may delight in competing because one attacks boldly and the other defends with invention. What matters is whether each participant’s way of playing makes possibilities available that the other can take up with pleasure. Stage improvisers have turned the condition into a craft rule, on which an offer is to be accepted and added to (Sawyer 2003). A blocked offer ends the scene, and an offer accepted without addition leaves the partner to do all the work. Gottman’s children who sustained a shared activity, and the rats who preferred partners that took turns, were meeting the same condition.

There is also the matter of how much mutual exposure a form of play demands. To play with someone is to be read while reading, and to know it, and this recursive load can be more than a given pair can carry. Rules, turns, roles, masks, and a board to look at instead of each other’s faces all reduce it. Helmuth Plessner’s (1999) defence of tact, role, and reserve against the ideal of total transparency states the general principle, which is that distance of the right kind makes closeness bearable. Two shy colleagues can spend a happy hour at table football who could not manage ten minutes of banter, and old friends may need no game at all. The right playmate is partly the person with whom the load is bearable at the level of exposure that this particular mode of play requires, which is a great deal for teasing and very little for chess.

It follows that the right playmate is never simply the right person. A friend with whom philosophical argument is a joy may be hopeless at silliness. A superb doubles partner may be an awkward companion in make-believe, and a much-loved sibling may be impossible at Monopoly. Different relationships enable different openings, so that the right playmate is the right person for a particular mode of coordination at a particular time. The same pair can lose the fit and occasionally find it again years later. Nor do playmates and friends coincide. Most people have friends they would trust with anything and would not choose to play with, and playmates, the regulars at the weekly five-a-side, with whom they play beautifully and exchange almost nothing else. People differ, besides, in how readily they turn a situation into play and in how much they need to win once they have, and discerning this about someone is part of coming to know them. Such dispositions are real and fairly lasting, and they are less fixed than talk of character suggests.

XIV. Winning, Balance, and the End of Fun

For years I played chess with my oldest childhood friend. We were at about the same level, and it was always worth setting the pieces up again, because either of us might win. I also played with my son, and that went differently. At first he lost every game and minded a great deal. Then, at twelve or thirteen, he spent two or three weeks studying openings, by what means I have never discovered, and came back with a repertoire of opening traps that I, who never learned openings properly, could not handle. I have not won a game against him since, and we no longer play.

Chess remained chess throughout. What changed was the distribution of capacities between two players, and with it the recursive possibilities of the encounter. While he was weak I could read his moves easily, he could do little that surprised me, and nothing I did was an invitation he could take up. Once he had his openings, the position was reversed and I was the one being read. In neither condition was there much room for a variation offered and returned. His success made him a far better chess player and ended chess as something the two of us did together. Winning a game and sustaining mutually enjoyable play are different achievements, and the first can destroy the second. The player who always loses stops wanting to play, and so, more slowly, do the opponents of the player who always wins. My daughter still loves the card game that became our family game during a stay in India, and she usually wins it, which shows that winning can keep a winner keen for years. It does less for everybody else at the table.

The opposite case occurred at a choir retreat in a large house in France, full of people of all ages. On a hot day a football match came about in which I, the only adult, and a sixteen-year-old who could really play took on about ten younger children. We were outnumbered five to one and far ahead in strength, in speed, and, in his case, in skill. For about an hour the game went back and forth with nothing between the sides. I have seldom sweated so much or enjoyed a game more.

The differences between the players had not been removed. They had been set against one another, numbers against strength and stamina against skill, and the balance was found in the playing, where nobody could have guaranteed it beforehand. This gives a distinction between the formal equality of rules and the practical balance of recursive possibilities. Two chess players under identical rules may have no contest at all, and two absurdly unequal sides may have a superb one. Handicap systems are attempts to deposit a practical balance in advance, and a balance stumbled on has a savour that an arranged one lacks. The animal evidence pointed the same way: what keeps partners playing is that each has openings the other can answer, and equal strength is neither necessary nor sufficient for that.

Sometimes the formal game is not worth beginning. When I played tennis with my children, my son wanted a proper match, scored by the rules, before his game was ready for it. A competitive game needs a floor of shared capacity under it before its form can do any work. And an occasion like that game in France cannot be ordered up again, since what was fun once is not necessarily repeatable.

XV. Why Play Cannot Be Forced

A parent can organise a game, explain its rules, supply the equipment, invite other children, and insist that everybody join in. None of these operations guarantees that any child will enjoy it. The distinction on which everything turns is between coordinating behaviour and generating positively valenced coordination. People can be made to perform the required moves. They cannot be made to undergo the corresponding resonance, which is settled at L1 and L2, beneath the reach of instruction. Huizinga held that play to order is no longer play, and Suits wrote voluntariness into his definition. Here the point follows from the architecture, since a deposit cannot compel the manner of its own uptake, and an invitation is in this respect a deposit like any other.

Play among people adds a difficulty that solitary enjoyments do not have, because the person trying to bring about the enjoyment is part of the coordination. A parent who anxiously needs a child to enjoy a party game alters the stakes that would have allowed enjoyment to emerge. The invitation becomes an obligation. The game acquires a new risk, that of disappointing the parent, and what was to have been an opening under buffered stakes turns into a performance under stakes freshly attached. The play is undermined by the effort to secure it, and the question whether everyone is having fun yet has the same effect in a smaller dose. Wolfenstein’s fun morality describes the cultural form of the difficulty: once fun is owed, failing to have it becomes a fault, and the stake is reattached from the other direction. The organised jollity of workplaces, where Fincham (2016) found fun to be muted at best, is the institutional form.

The principle can be stated simply. Play can be invited, supported, scaffolded, and cultivated, and its positively valenced uptake cannot be guaranteed by the party arranging the conditions. Two qualifications keep it honest. Reluctant participation often turns into real enjoyment, as every child dragged to a party and collected three hours later in high spirits demonstrates. And play is seldom entirely voluntary, since social pressure and unequal power are ordinary features of it. Complying with an invitation, entering into a game, and finding the resulting coordination worthwhile are three separate achievements, and only the first can be commanded.

The distribution of these powers is a political matter on any scale from the family upward. It is worth asking, of any setting, who gets to invite play, who can refuse, who is allowed to alter the rules, and who is expected to accommodate everybody else’s enjoyment: the older sister required to let the little one win, the partner who watches the other play, the junior employee at the away-day. The burden of keeping play going is distributed as unevenly as other recursive burdens. Vivian Paley (1992) introduced into her kindergarten the rule that you can’t say you can’t play, so that no child could be turned away from a game. The rule secured entry, which matters a great deal, since exclusion is the gravest sanction children have. It could not secure being wanted. A rule is able to change the allocation of stakes, and it leaves resonance where it was.

XVI. Whose Fun?

Enjoyment is frequently asymmetric. What one participant experiences as affectionate teasing another experiences as humiliation. One player is exhilarated by fierce competition while another feels exposed. A group can be enjoying a game that one of its members badly wants to end. Under these conditions the declaration that something is only a game, or was only a joke, is itself an exercise of power, since it announces that the stakes are off on behalf of the person who is bearing them. Stakes are mesocosmic facts, and saying that there are none does not remove them. A sought opening for one participant can be a disruptive opening for another.

The theory therefore has to ask, of any episode of play, whose opening is being sought, whose resonance is occurring, and who bears the cost of sustaining it. Sutton-Smith and Diana Kelly-Byrne (1984) warned long ago against the idealisation of play, pointing out that children’s play includes cruelty, coercion, and exclusion along with everything adults like to see in it. Aaron Trammell (2023) has argued more recently that the definition of play as voluntary and pleasurable was written from the position of those who are free to leave, and that play can be something done to people. The three questions distinguished at the outset accommodate the point. An activity can be organised as play for everyone present, fun for some, and damaging to others.

The hunt is the limiting case. English uses one word for what is played and for the animals that are pursued, a usage on record since the Middle Ages, and the association is no accident. Tracking, anticipation, stratagem, and the difficulty of outwitting a counterpart that reads its pursuer are among the recognised pleasures of hunting, and the interrecursivity involved is real and crosses species. The frame, however, belongs to one side only. No stake has been reallocated for the quarry, which is coordinating for its life. A single encounter can be play for one party and nothing of the kind for the other, and the formula of a sought opening under reallocated stakes is incomplete until it says whose opening and whose stakes.

Whether play stays viable for all its participants depends largely on their ability to stop, to decline, to slow the pace, and to renegotiate. Animal play is full of devices to this end. Play signals cluster around the roughest moves, stronger animals hold back, roles are exchanged, and a youngster that plays too hard finds itself without partners (Bekoff and Byers 1998). The human equivalents are familiar. A strong player eases off against a beginner, and the handicap systems of golf and Go are that courtesy deposited as rule. Bernard De Koven (1978) built a whole philosophy of the well-played game on the willingness of players to change any rule that stops the game being good for everyone in it. Linda Hughes (1983), watching girls play foursquare, found that the rules actually enforced had little to do with the official ones and a great deal to do with being nice to one’s friends.

A last distinction separates the game that catches from the game that captures. An activity can hold attention for hours while undermining the body, the relationships, and the other coordinations of the person held. The machine gamblers studied by Natasha Dow Schüll (2012) describe a zone in which nothing opens and nothing is at risk of ending, produced by devices engineered for continued uptake. Engagement is therefore no measure of value. A game can be designed with great skill to secure repeated play without being good for the life of the player, and the test is the third question: whether, mediation by mediation, the vital stakes of those involved are sustained. Many things pass that test without being much fun, and some things are a great deal of fun and fail it.

XVII. Games as Poietic Scaffolds

Nothing in the account so far requires made things or symbols. Students of animal behaviour conventionally sort play into three kinds, locomotor, object, and social (Burghardt 2005), and the three correspond closely to the three zoetic mediations and to the three types of recursive counterpart: the animal that leaps and twists for no reason is opening its coordination with its own body, a selfrecursive counterpart; the animal batting a pebble or a leaf is opening its coordination with something that does not answer back; and the animals wrestling are opening their coordination with each other. Play is zoetically sufficient. A complete form of it exists wherever embodiment, being-with, and dwelling operate through L1 and L2 with reserves to spare.

Games are a different kind of thing. A game is an arrangement of constraints and affordances, usually rules together with equipment and a marked ground, that makes certain possibilities of coordination available to whoever takes it up. Three aspects of any game should be kept apart: its constraints, the degree to which those constraints have been sedimented, and its living uptake on a given occasion. The second varies enormously. Children invent a game on the spot, fix its rules by word of mouth (you have to touch the tree, and you can’t go on the grass), play it once, and never return to it. At that point the game is an L3 articulation, made for this afternoon. If it is played again tomorrow and taught to a cousin, it has begun to travel, and the playground games that have passed from child to child for generations are full L4 stabilisations. With written laws, referees, and federations a game becomes an institution. These are different degrees and different histories of stabilisation, and all of them are games.

It follows that a game typically deposits earlier play and need not deposit any earlier playing of itself. A designer can specify a new game before anyone has played it, which is an ordinary case of projective fabrication, a thing made for a coordination that has not yet occurred, on the model of earlier made things. Nor do all games specify winners or an end. Open-ended role-playing, cooperative storytelling games, and the sandbox genre of video game can go on indefinitely without a victory condition. What games have in common is narrower. They hold conditions of play steady, and they leave undetermined what happens when players take them up.

English obscures the distinction between the arrangement and its uptake by using one word for both. Chess as the game is nonrecursive without qualification. A game of chess is a recursive process between two people who are taking it up. This matches the general rule for poietic deposits, which never preserve the recursion that produced them and preserve only constraints, affordances, and orientations for new recursion. The rules of hopscotch hold no trace of the afternoon on which someone first drew squares on a pavement.

Caillois’s axis from paidia to ludus measures the degree to which play has been sedimented. Graeber’s distinction is the distinction between zoetic opening and poietic deposit, and his remark that play generates rules names the direction of deposition. Suits’s lusory attitude names the uptake of a game as play, and since a deposit cannot compel the manner of its own uptake, a game can be taken up without play: by the child made to join in, by the professional for whom the match is a shift, by the gambler at the machine. Play without a game and a game without play are both ordinary.

The same reasoning yields a point that matters a good deal for the question of fun. A game can be designed to be played without thereby being enjoyable. Its rules may function perfectly and generate nothing that anyone finds compelling. Conversely a clumsy or broken game can become marvellous because the right people take it up inventively, as with the family card game whose scoring nobody can explain and which nobody would give up. Resonance cannot be deposited. What a designer can deposit is a set of conditions under which resonance has been known to occur, which makes a well-designed game something like a portable ecology of acquisition: it sets up a local anticipatory field quickly, and it supplies constrained deviations through shuffled cards, dice, hidden information, and the moves of opponents. It does this for a player in general, and whether it catches this player is settled elsewhere.

The deposit nevertheless does things that play alone cannot. A game is a designed mesocosm in which an extraordinary number of conditions are fixed in advance, one ball of specified weight, a pitch of specified size, a set duration, so that none of them has to be negotiated while the play is under way and the whole of the players’ attention is free for reading one another. Nonrecursive stability, in a well-made game, serves the openness of what happens inside it. A game also lowers the load of mutual exposure discussed above, by giving the players something other than each other to attend to. Where it specifies an ending it supplies what free play conspicuously lacks, a mechanism of finality. Free play ends when someone is tired, or hurt, or called in for supper, and nobody has won. A game with a result says when it is over and who prevailed, which is the disjunction Lévi-Strauss identified, and the Gahuku-Gama refusal of it shows that the mechanism can be declined. And a game allows play among strangers. Two children who have never met can play draughts within a minute, because the coordination they would otherwise have to build has been deposited for them. At sufficient scale, with federations, referees, and fixture lists, the game becomes an institution in the theory’s sense: a stabilisation of interrecursivity among people who could never all be directly responsive to one another.

The effect of a deposit on stakes can be watched at the moment a game begins. Two people who have been hitting a tennis ball back and forth agree to start a match. The court, the rackets, the ball, and the strokes are what they were a minute earlier. A symbolic deposit that existed all along, the scoring system, has been activated, and the same forehand that was useful practice now wins or loses a point and feels different to hit. The reverse also happens, when players abandon the score in the middle of a set and go back to trying out shots. The ability to move between these arrangements is recursive fluidity in a concrete form, and amateurs have more of it than professionals, for whom the match is a livelihood.

Piaget’s marbles study can be reread in this light. The younger boys who took the rules to be eternal were encountering a made anchor as though it were found ground, like weather. The older boys had discerned that the rules were deposits, made by players and open to being remade by them. Learning to play games includes learning what kind of thing a rule is.

XVIII. The World a Game Makes

About nine 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 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 the year of Mozart they have never played like that with each other.

The children had done more than find a way of passing the time. They had established a field in which things were available and relevant that had no existence outside it. Particular objects were valuable there, an exchange had consequences, a bargain created an expectation, and reading another player’s intentions mattered a great deal. Most of the materials came from the wider world, since money, trade, and the weighing of one thing against another were not their invention. What they did was to select certain possibilities from that world, reorganise them, and make them the ground of a shared activity.

A provisional term is useful for what they made. A playful submesocosm is a temporarily organised field of living coordination, established within an existing mesocosm, in which particular counterparts, constraints, symbolic distinctions, and stakes acquire a locally operative relevance through the continuing uptake of its participants. Such a field is not self-sufficient. The players still get hungry and tired, depend on adults, and can be called away by a meal, a journey, or a parental decision, and a visit of two and a half weeks set a limit to this one. Its boundaries are permeable in both directions, so that a grievance over a bargain can be carried to the dinner table and a bad mood from the dinner table can be carried back. Inside those boundaries the coordination can nevertheless be extraordinarily dense.

The term marks a distinction that the word game hides. A game is an arrangement of constraints, and a playful submesocosm is the field generated when people take such an arrangement up. The same deposit supports very different fields. Chess with my friend and chess with my son were two fields on one deposit. The card game we played in India was, as a deposit, a small commercial pack available anywhere, and what it made in our case was a way for a whole family, across ages and abilities, to be absorbed in one thing together. It was the one game on which everyone could agree. A field can also arise with scarcely any durable deposit behind it, as in New Zealand, where, so far as the adults could tell, the rules existed only among the three who played.

Huizinga’s magic circle said where such a field is. The present terms say what it is made of, which is continuing uptake; what holds it together, which is the resonance of those inside it; what crosses its boundary, which is materials, energy, stakes, and interruptions; and why it can come to matter so much. Jaakko Stenros (2014) has argued that the circle needs to be taken apart into the psychological experience of playing, the social contract among the players, and the cultural form of the game. The three correspond closely to resonance, to an allocation of stakes agreed among participants, and to the deposit.

A playful submesocosm has its own distribution across the levels of recursivity. Some of what goes on in it runs at L1, in fluent handling and habitual recognition. Some is salient at L2: a suspicious offer, a thrilling gain, a feeling that something is not right. Some is articulated at L3, as when a rule is made up to deal with a case nobody foresaw. Some has the standing of L4 within the field, the established rules, the categories of valuables, what counts as winning. And now and then the players reach L5 and ask whether the whole arrangement is fair and should be reorganised. Csikszentmihalyi’s flow resembles L1, since in it complex coordination proceeds without being separately monitored, and the two should still not be identified, because the excitement and felt challenge of an absorbing game involve recurrent L2. An absorbing game is better described as a particularly successful distribution of coordination across the levels. In some contests, keeping fluent coordination intact under pressure is what decides the result, and the penalty taker or darts player who begins to monitor a movement that used to run by itself has, in the sporting phrase, choked.

Set beside one another, the three episodes from my own family resist reduction to a single mechanism of fun. In New Zealand the children were enjoying a world they had built jointly. With Mozart, a finished work of art and the reciprocal amplification of two attachments were central. With the card game, a portable deposit made participation across the generations possible. These are different configurations of resonance, and a theory of play should expect such differences.

What the cases have in common can be put as a proposition. In its living uptake a game does more than supply rules. It temporarily reorganises what is relevant, what actions count as, which stakes matter, and how the participants read one another. The game is the arrangement, and the world is what the players manage to make together by playing.

XIX. Kinds of Playful Coordination

Once play is treated as a mode of coordination, its varieties can be sorted by what is opened and how, and the sorting quickly shows how little the varieties share.

The first dimension is the type of counterpart. Play with nonrecursive counterparts includes skimming stones, stacking blocks, flying a kite, solving a jigsaw, and surfing, in which the wave is fully indifferent to the surfer and only partly predictable. Play with a selfrecursive counterpart, the player’s own body, includes spinning until dizzy, swinging, holding one’s breath, and running for the pleasure of it. Play with interrecursive counterparts includes chasing, wrestling, shared make-believe, teasing, and every team sport. These differ in kind. The stone will do the same thing however it is regarded, and skill at skimming consists in modelling it ever more closely. A playmate’s next move depends on her reading of one’s own, so that skill consists as much in managing one’s own legibility as in reading hers, and no amount of modelling settles the matter in advance.

Caillois’s four families can be read against this dimension, with the proviso that they correspond to the types of counterpart without being derivable from them. Agon is typically interrecursive play under engineered symmetry of starting conditions, although a runner can compete against her own best time. Alea is play with a nonrecursive counterpart that has been manufactured to be stable and durable while resisting both prediction and control, the specification of a die. Chance games show why recursive category must be kept apart from predictability, since the die is as nonrecursive as the moon and far less predictable than most friends. The gambler’s conviction that luck is running her way attributes responsiveness to a counterpart that has none, and the attribution survives because nothing in a short run of throws forces its revision. Ilinx is commonly selfrecursive, the deliberate unsettling of the body’s own regulation of balance, and is just as often sought in company, on a roundabout or a dance floor. Mimicry plays with attribution itself. The child who rides a stick or feeds a doll engages a counterpart under an attribution that everyone present knows does not hold, sometimes announced (let us say that this is the horse) and sometimes simply enacted. The four families are better treated as analytic dimensions that intersect with the types of counterpart, and most actual games mix them: poker combines a shuffled deck with the reading of opponents and the deliberate falsification of one’s own legibility.

The second dimension is the level of recursivity. Rough-and-tumble runs at L1 and L2 and needs no words. Pretend play lives largely at L3, in articulations that hold for this interaction and lapse when it ends: the floor is lava, you be the shopkeeper. When such articulations become portable and are passed from child to child across years and cities, they have become L4 stabilisations, and the Opies (1959, 1969) documented a vast body of them, rhymes, counting-out formulas, and truce terms such as barley and fainites, transmitted for generations among children without adult help and without writing. Codified sport is L4 with institutions attached. L5 is reached whenever players turn on the rules themselves: house rules, the argument about whether that was fair, the invention of a new game.

The third dimension is mediation. Play in dwelling is play with puddles, snow, slopes, trees, tides, and darkness, counterparts no one made. Play in multimateriality is play with made things, and it matters whether the thing was made for play. A toy or a playground is anticipated play deposited in plastic and steel, usually by adults, while a cardboard box or a flight of municipal steps is taken up for play against the purpose sedimented in it. Play in multisymbolization includes puns, riddles, nonsense, and the score. Further dimensions cut across these three. Openings may be induced on the spot by an invitation, scheduled, as with break time or a league fixture, or disruptive, as with a practical joke. Games differ in how they treat legibility, from hide-and-seek, which is about nothing else, to team sports in which each side must be readable to itself and unreadable to its opponents, and in how frequently the outcome-deciding event occurs. They differ as well in the density of the deposit. A single-player video game is a deposit of unusual intricacy, nonrecursive in itself, which players routinely and pleasurably engage under an interrecursive attribution, as an opponent out to get them, while a multiplayer game puts interrecursive counterparts back on the other side of the same deposit.

These dimensions say where a piece of play is located. They say little about why the kinds feel so different or why different people are drawn to them, and for that a complementary sorting is needed, by what becomes valuable in the playing. At least eight modes can be distinguished. In embodied exploratory play the value lies in discovering what one’s own body can do, in swinging, balancing, spinning, and tumbling. In material exploratory play it lies in discovering what nonrecursive counterparts afford: stones, water, sand, a box of bricks. In agonistic play it lies in the reciprocal production of obstacles and resistance. In cooperative play it lies in sustaining a common activity in which each participant enables the contributions of the others. In improvisational play it lies in jointly generating possibilities that no participant could have specified in advance. In mimetic play it lies in making a temporary alternative world habitable through roles, attributions, and material supports. In playful sociability it lies in sustaining being-with through humour, teasing, and shared attention. And in rule-making play it lies in negotiating, testing, and transforming the conditions of play itself, which is why children often spend as long arguing about a game as playing it and do not seem to regard the time as wasted.

The list is neither exclusive nor final. It names ways in which playing can become worth doing, and its use is to show how far the official description of a game can stand from what its players value in it. A single game moves among the modes. Children playing football on a patch of grass may within half an hour be competing fiercely, inventing a rule to deal with the tree, enjoying the plain feeling of kicking a ball, and using the whole business as a way of being together. A child may love football without caring whether her side wins. A chess player may care more for the beauty of a position than for the result. Two friends may love a board game chiefly because its predictable structure gives them a way of spending four hours in each other’s company. Beginning from the living coordination, and not from the formal definition of the game, makes all of this visible, and it also explains a common cause of the flat evening, which is that two people can be playing the same game in different modes, one of them to win and the other to talk. The player types of the design literature are these modes stabilised into kinds of person, a portable classification of what is in fact a shifting distribution.

XX. Why There Is No Single Theory

Set against these dimensions, the failure of the single definition becomes intelligible. What ordinary speech calls play is an L4 classification of a cluster, in the same way that a forest or a reef is, and to ask what play is, as though it were one counterpart with one character, is to put the question to the classification. A toddler stacking cups, two dogs wrestling, a chess final, a round of banter, and a night at the roulette table differ in mediation, in type of counterpart, in level, in stake allocation, in the density of deposit, in the kind of opening involved, and in what those taking part value in it. Wittgenstein’s family resemblance was an accurate report of what the classification bundles.

The classic theories can then be read as low-bandwidth accounts, each precise about one region of the cluster. Groos and Spencer saw embodiment and its metabolic economy. Piaget and Vygotsky saw the L3 articulation by which a stick becomes a horse. Mead saw being-with on its way from particular others to the L4 generalised other. Bateson and Goffman saw the reallocation of stakes, and described it as a message. Huizinga saw the deposit, the bounded and rule-governed world, and took its boundaries to be the essence of play. Caillois saw the types of counterpart without having names for them. Geertz saw the reattachment of stakes and assigned it to multisymbolization. Csikszentmihalyi saw the band of positively valenced salience from the position of a single player. Suits defined the uptake of a game and concluded, reasonably, that he had defined games. Gadamer saw the to-and-fro and left its conditions unspecified. The literatures on fun divide in the same way, with the designers attending to the deposit, the sociologists to the group, and the psychologists to the pair. Each account fits its region, and each produces mediational or recursivity misfit when extended over the whole.

What Living Value Theory contributes, then, is three interlocking claims and a set of dimensions, in place of a definition by essence. Play is a distinctive way in which living coordination is opened, varied, maintained, and enjoyed, under an allocation of stakes that makes this viable. Games are poietic arrangements that stabilise some of the conditions for such coordination without guaranteeing that anyone will take them up with pleasure. And actual playing can establish a temporarily differentiated field within a mesocosm, in which relevance, frames, stakes, and levels of recursivity are continually organised and reorganised, and in which resonance arises, is sustained, and dissolves. Any particular case is described by saying which coordination is opened, with what type of counterpart, at what level, with which stakes reallocated to whom, with what deposit in support, and what in it is found worth continuing, by whom, and for how long. The characterisation is deliberately wide. It makes kin of play, wonder, learning, and a good deal of art and inquiry, and it supplies no sharp line between them, because the architecture gives no reason to expect one.

It should also say what would count against it. Three kinds of case would do so: enjoyable play in which careful description finds no opening at all, reciprocal play that stays pleasant with very little positive misalignment in it, and highly absorbing activity that those engaged in it regard as work. Such cases should be used to refine the account and not classified away. The first kind has already forced one refinement, the recognition that much of the pleasure of play lies in fluent coordination with openings merely available, and the others may force more.

XXI. When Play Lets Go

Playful coordination exists on several timescales at once, and its endings differ accordingly. A joke catches for a few seconds. A game catches for an evening. A group of friends may spend several years absorbed in the same online game, and two children may live inside an opera for nine months. A single successful move, an enjoyable session, an enduring shared practice, and a playworld that is central to a period of life are different achievements.

The longer forms can be described as a history of recursive relevance. At first the game is one available possibility among many. Repeated enjoyable uptake confers relevance on it. The players build routines around it, elaborate its internal distinctions, accumulate stories about memorable occasions, and begin to anticipate the next session, and Fine and Corte’s observation that fun gives a group a past worth retelling belongs here. In time a considerable portion of their being-with may come to be organised around this one practice. Kelly-Byrne (1989), who played regularly with a single child over the course of a year, recorded a shared playworld being built up in this way and then giving place to something else as the relationship changed. When such a practice ends, what alters is part of the organisation of a shared life, which is a larger matter than a preference among leisure activities.

Losing interest in a game is therefore different from losing the coordination that the game made possible, and at least three outcomes need to be told apart. Sometimes the friends simply move their friendship into another activity, which shows that the being-with stake was protected by redundancy all along. Sometimes they discover that the game was the main thing holding the friendship together, and the friendship goes with it. And sometimes they remain close and never again achieve that particular pleasure in each other’s company. The end of play involves different redistributions of relevance in each case, with different consequences for the stake underneath.

As to why play lets go, the answer has the same form as the answer to why it caught. The value of the play lay in a relation among a deposit, the histories of the players, their present condition, and the allocation of stakes, and any of these can shift. Children grow, and where there is an age gap one grows away faster. A third person arrives. A timetable changes. The field is exhausted and the game has nothing further in it, which is Koster’s case and one case among several. Or the stakes are reattached, as when a kickabout turns into a league with a table and the people who came for the kickabout drift off. Or the practical balance goes, as when one player improves out of the other’s reach.

Such endings are peculiar in that they are rarely events. There is a last time that a parent swings a child round by the arms, a last round of a bedtime game, a last evening on which a group of friends logs on together, and almost never does anyone present know it. Relevance is withdrawn without leaving a mark, since a coordination that has stopped being sought produces no disturbance and so nothing to notice. The realisation comes later, at L2 or L3, usually prompted by an object: the box at the back of the cupboard, the recording.

What remains is a dormant layer. Two people who shared an intense period of play may remember it with great affection and feel no wish to resume it, so that the historical value of the coordination survives its present attraction. Because the field was sedimented and has not been erased, restoration is sometimes possible, and old friends at a reunion can be back in the old voices within minutes. Nothing guarantees that this will happen.

XXII. From Playing to Hanging Out

Play changes its forms across the whole of a life. A child drops a game she adored, an adult is seized by a hobby at fifty, two colleagues discover late in their acquaintance that they can make each other helpless with laughter. One such passage has a linguistic marker, and it is treated here in outline only, because it raises a question about stakes that needs a study of its own.

In many English-speaking households, somewhere between the ages of eleven and thirteen, a word goes missing. A child who for years has asked whether a friend can come over to play begins to ask only whether the friend can come over, and soon announces that the two of them are going to hang out. This is a tendency of particular places and decades, and the parallels in other languages are suggestive only: German children who used to spielen come to abhängen or chillen, French adolescents traînent, and in Spain they quedan. The shift is selective. The verb survives wherever it takes a named object, so that teenagers go on playing football, playing Fortnite, playing the guitar. What lapses is the intransitive use, play as the name for what friends do together when nothing in particular has been specified.

I remember the day it happened to me. I was twelve, and my friends and I were playing war, as we had done for as long as I could remember, and it felt wrong. I believe all of us felt it on the same afternoon, and that was the end of it. Nothing about the game had changed.

The word had been doing symbolic work. In the settings where the shift is marked, “play” is the L4 category through which adults organise childhood: playtime, playground, playgroup, playdate, go and play. The category carries an allocation of stakes, for to say that children are only playing is to say that what happens among them does not count. That description was never accurate. Exclusion, embarrassment, quarrels over fairness, and the loss of a friend matter intensely to six-year-olds, as Paley’s and Corsaro’s observations make plain. What changes at around twelve is the organisation and interpretation of the stakes. Sensitivity to the regard of peers rises sharply (Blakemore 2018), more of life is organised around friends, and the young person begins to refuse a description that adults have been applying on her behalf. “Play” persists where its object is a deposit, a sport, a video game, an instrument, because the deposit vouches for the seriousness of the uptake, and adults in English likewise play golf or bridge and play intransitively only with children and dogs.

The activity continues under the new name. Much of what adolescents do when hanging out meets every classical criterion of play: teasing, banter, mock insult, daring, impersonation, flirting, kicking a ball about. What alters is the structure of the frame. Children’s play is typically framed by declaration, announced before it begins and suspended by a truce word. Adolescent play tends to put the move first, an insult, a shove, a compliment, and to settle afterwards, by how the move is taken up, whether it was play. “Just kidding” is a frame claimed in retrospect. This is play with the frame in the sense set out earlier, conducted with fewer supports than any game provides, and it requires each participant to manage her own legibility as she goes.

Hanging out is the setting this kind of play needs. Paul Corrigan (1976), asking boys on the street corners of an English industrial town what they did there, was told that they did nothing, and found that doing nothing was a full occupation of talking, joking, waiting for something to happen, and now and then making it happen. Hanging out is being-with held at readiness, a temporal and relational field in which playful resonance can arise without having been specified in advance. A playdate or an organised activity fixes the opening beforehand, and hanging out leaves its content undecided. Scheduled play is none the worse for being scheduled. The difference is that the temporal organisation of an activity determines what kind of spontaneity is possible within it, and the rarest kind of fit, the play that two people could only have found with each other, has to be stumbled on in unscheduled time.

What needs explaining is why the frame goes underground at this age and no other, and the likeliest answer concerns stakes. In childhood play, the stakes of the body are buffered and the stakes of standing among peers are live. With puberty a further kind of stake enters, sexual and romantic, and it is added to the other two without replacing them. The capacities that childhood play exercised, imagination, the taking of roles, mutual responsiveness, the discernment of frames, are redeployed in flirtation and courtship, where the frame is kept deliberately undecided because what is at stake is the relation itself. Courtship is heavily scripted, as John Gagnon and William Simon (1973) showed, and its English vocabulary is the vocabulary of games: a player, playing the field, playing hard to get. None of this means that childhood play turns into sexuality. The relation is a redistribution of recursive relevance, in which make-believe war stops being compelling as other forms of being-with begin to be. A declared frame, which switches the stakes off in advance, is of no use to people for whom the question of whether something counted is the point. It is also the case that sexual play makes the opposition of play and seriousness untenable, since an encounter can be playful, pleasurable, consequential, and bound by strict requirements of consent all at once, the capacity to stop or to revise being part of what keeps the frame viable there as everywhere else. How bodily, social, and sexual stakes combine in adolescent play is the question that the lapse of a word opens, and it is left open here.

Two cautions apply. Adolescence as a marked stage is widespread (Schlegel and Barry 1991), but its length, content, and storm are variable, as Margaret Mead (1928) argued against Hall, and where children work alongside adults from an early age the category of play is thinner and its withdrawal less dramatic. Nor is hanging out something that adult life everywhere leaves behind. The Bengali adda, the long, aimless, argumentative conversation among friends at a tea stall or on a neighbourhood porch, is hanging out sedimented into a named and valued adult practice with its own places, hours, and etiquette (Chakrabarty 2000).

XXIII. The Value of Play

Much of Living Value Theory concerns how life achieves viable coordination, how that coordination becomes muted, and how it is reopened when something goes wrong. Play shows that living beings do not seek stability only in order to be rid of disturbance. Sometimes they secure a coordination so that they can enjoy making it variable again. The harness of a roller coaster exists so that the drop can be thrilling, and a small child shrieks with pleasure at being tickled because she is certain that the adult will stop when she has had enough. The security is built for the sake of the disturbance.

In play with others the value lies in a further discovery, that another living being can meet one’s variations with responses that make the activity more compelling than it could ever be alone. This is a distinctive achievement of being-with, and it gives a fuller account of why other people matter than accounts built on need. People are valued for helping to meet wants that existed before them, for protection, and for making daily coordination easier. They are also valued because they make available forms of life that would otherwise not exist. A good playmate makes a possible world available. Sometimes the world lasts an afternoon and sometimes for years, sometimes it becomes a permanent part of a relationship and sometimes it leaves only a memory, and its emergence is controlled by nobody involved.

The principle of invisible value applies with particular force. Value is whatever sustains, repairs, intensifies, or improves coordination, whether or not it is ever measured or recognised, and the coordinations that meet a stake most fully are the likeliest to drop from view. Play in its vivid forms is an unmuting undertaken on purpose, and the classic theories attended almost entirely to those forms. The muted side of the same capacity, the easy company from which openings arise and into which they resettle, looks like nothing because it is working.

A good deal has been left for further work, and the account should be judged partly by whether it extends to the harder cases. Some contests are rigged, in the sense that the announced conditions and the operative ones diverge, and the difference between a player who does well inside a frame and a party able to alter the frame is a difference of power. The advice to play the game, as it is given in workplaces, refers to conditions of this unannounced kind. Some encounters are play for one side only, of which hunting is the clearest instance. War contains exhilaration, joking, and rivalry without becoming a game, and its re-enactment as a hobby attaches solemn obligations of commemoration to activities whose pleasures have then to be publicly played down. Acting and role-play have barely been touched. And sexuality, where embodied stakes of great weight meet the most intricate interrecursive coordination there is, plainly needs treatment of its own.

The account does not predict. It cannot say on Friday whether Saturday’s game will take, and it would be a worse theory if it claimed to. What it supplies is a set of questions that can be put to contrasting cases. When the same pair plays the same game on two evenings and only one of them comes alive, what differed in the reserves each brought, in the first few variations offered, in whether they were taken up, and in the stakes attached to the evening? When two siblings who shared a game drift apart in their interests, did the shared engagement decline before the individual enthusiasm or after it? For the two friends with whom this article began, nothing about the game changed between Friday and Saturday. A coordination that had been achieved on one evening had to be achieved again on the next, and on that occasion it was not.

That is the general condition. The rules of a game can be preserved, and so can the pieces, the recordings, and the memories. The living coordination through which the playing became wonderful cannot be kept in that way. It can only be achieved again, and nothing guarantees that it will be.

Selected References

Apter, Michael J. 1991. “A Structural Phenomenology of Play.” In Adult Play: A Reversal Theory Approach, eds. John H. Kerr and Michael J. Apter. Amsterdam: Swets and Zeitlinger.

Bartle, Richard. 1996. “Hearts, Clubs, Diamonds, Spades: Players Who Suit MUDs.” Journal of MUD Research 1(1).

Bateson, Gregory. 1955. “A Theory of Play and Fantasy.” Psychiatric Research Reports 2: 39–51.

Bauer, Erika B., and Barbara B. Smuts. 2007. “Cooperation and Competition during Dyadic Play in Domestic Dogs, Canis familiaris.” Animal Behaviour 73(3): 489–499.

Baxter, Leslie A. 1992. “Forms and Functions of Intimate Play in Personal Relationships.” Human Communication Research 18(3): 336–363.

Bekoff, Marc, and John A. Byers, eds. 1998. Animal Play: Evolutionary, Comparative, and Ecological Perspectives. Cambridge: Cambridge University Press.

Blakemore, Sarah-Jayne, Daniel M. Wolpert, and Chris D. Frith. 1998. “Central Cancellation of Self-Produced Tickle Sensation.” Nature Neuroscience 1(7): 635–640.

Blakemore, Sarah-Jayne. 2018. Inventing Ourselves: The Secret Life of the Teenage Brain. London: Doubleday.

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.

Burghardt, Gordon M. 2005. The Genesis of Animal Play: Testing the Limits. Cambridge, MA: MIT Press.

Caillois, Roger. 1961 [1958]. Man, Play and Games. Translated by Meyer Barash. New York: Free Press.

Chakrabarty, Dipesh. 2000. Provincializing Europe: Postcolonial Thought and Historical Difference. Princeton: Princeton University Press.

Corrigan, Paul. 1976. “Doing Nothing.” In Resistance through Rituals: Youth Subcultures in Post-War Britain, eds. Stuart Hall and Tony Jefferson. London: Hutchinson.

Corsaro, William A. 1985. Friendship and Peer Culture in the Early Years. Norwood, NJ: Ablex.

Csikszentmihalyi, Mihaly. 1975. Beyond Boredom and Anxiety. San Francisco: Jossey-Bass.

Daley, John Francis, and Jonathan Goldstein, dirs. 2018. Game Night. Burbank, CA: New Line Cinema.

De Jaegher, Hanne, and Ezequiel Di Paolo. 2007. “Participatory Sense-Making: An Enactive Approach to Social Cognition.” Phenomenology and the Cognitive Sciences 6(4): 485–507.

De Koven, Bernard. 1978. The Well-Played Game: A Player’s Philosophy. Garden City, NY: Anchor Press.

Donaldson, Julia. 1999. The Gruffalo. Illustrated by Axel Scheffler. London: Macmillan Children’s Books.

Erikson, Erik H. 1950. Childhood and Society. New York: Norton.

Erikson, Erik H. 1968. Identity: Youth and Crisis. New York: Norton.

Fincham, Ben. 2016. The Sociology of Fun. London: Palgrave Macmillan.

Fincher, David, dir. 1997. The Game. Los Angeles: PolyGram Filmed Entertainment.

Fine, Gary Alan. 1979. “Small Groups and Culture Creation: The Idioculture of Little League Baseball Teams.” American Sociological Review 44(5): 733–745.

Fine, Gary Alan, and Ugo Corte. 2017. “Group Pleasures: Collaborative Commitments, Shared Narrative, and the Sociology of Fun.” Sociological Theory 35(1): 64–86.

Freud, Sigmund. 1955 [1920]. Beyond the Pleasure Principle. Standard Edition, vol. 18. London: Hogarth Press.

Gadamer, Hans-Georg. 1989 [1960]. Truth and Method. 2nd rev. ed. Translated by Joel Weinsheimer and Donald G. Marshall. London: Sheed and Ward.

Gagnon, John H., and William Simon. 1973. Sexual Conduct: The Social Sources of Human Sexuality. Chicago: Aldine.

Gaskins, Suzanne, Wendy Haight, and David F. Lancy. 2007. “The Cultural Construction of Play.” In Play and Development: Evolutionary, Sociocultural, and Functional Perspectives, eds. Artin Göncü and Suzanne Gaskins. Mahwah, NJ: Lawrence Erlbaum.

Geertz, Clifford. 1972. “Deep Play: Notes on the Balinese Cockfight.” Daedalus 101(1): 1–37.

Goffman, Erving. 1961. Encounters: Two Studies in the Sociology of Interaction. Indianapolis: Bobbs-Merrill.

Goffman, Erving. 1974. Frame Analysis: An Essay on the Organization of Experience. New York: Harper and Row.

Gottman, John M. 1983. “How Children Become Friends.” Monographs of the Society for Research in Child Development 48(3).

Graeber, David. 2015. The Utopia of Rules: On Technology, Stupidity, and the Secret Joys of Bureaucracy. Brooklyn: Melville House.

Groos, Karl. 1898 [1896]. The Play of Animals. Translated by Elizabeth L. Baldwin. New York: Appleton.

Groos, Karl. 1901 [1899]. The Play of Man. Translated by Elizabeth L. Baldwin. New York: Appleton.

Hall, G. Stanley. 1904. Adolescence. 2 vols. New York: Appleton.

Ham, Jackson R., and Sergio M. Pellis. 2024. “Play Partner Preferences among Groups of Unfamiliar Juvenile Male Rats.” Scientific Reports 14: 16056.

Ham, Jackson R., and Sergio M. Pellis. 2025. “Groups of Familiar Male Rats Form Unstable Partner Preferences When Play Fighting during the Juvenile Period.” iScience 28: 112562.

Hughes, Linda A. 1983. “Beyond the Rules of the Game: Why Are Rooie Rules Nice?” In The World of Play, ed. Frank E. Manning. West Point, NY: Leisure Press.

Huizinga, Johan. 1949 [1938]. Homo Ludens: A Study of the Play-Element in Culture. London: Routledge and Kegan Paul.

Hunicke, Robin, Marc LeBlanc, and Robert Zubek. 2004. “MDA: A Formal Approach to Game Design and Game Research.” In Proceedings of the AAAI Workshop on Challenges in Game AI. San Jose, CA: AAAI Press.

Huron, David. 2006. Sweet Anticipation: Music and the Psychology of Expectation. Cambridge, MA: MIT Press.

Juul, Jesper. 2013. The Art of Failure: An Essay on the Pain of Playing Video Games. Cambridge, MA: MIT Press.

Kelly-Byrne, Diana. 1989. A Child’s Play Life: An Ethnographic Study. New York: Teachers College Press.

Koster, Raph. 2005. A Theory of Fun for Game Design. Scottsdale, AZ: Paraglyph Press.

Lancy, David F. 2008. The Anthropology of Childhood: Cherubs, Chattel, Changelings. Cambridge: Cambridge University Press.

Lazzaro, Nicole. 2004. “Why We Play Games: Four Keys to More Emotion without Story.” Oakland, CA: XEODesign.

Lévi-Strauss, Claude. 1966 [1962]. The Savage Mind. London: Weidenfeld and Nicolson.

Mead, George Herbert. 1934. Mind, Self, and Society. Chicago: University of Chicago Press.

Mead, Margaret. 1928. Coming of Age in Samoa. New York: William Morrow.

Nguyen, C. Thi. 2020. Games: Agency as Art. New York: Oxford University Press.

Opie, Iona, and Peter Opie. 1959. The Lore and Language of Schoolchildren. Oxford: Clarendon Press.

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.

Panksepp, Jaak, and Jeff Burgdorf. 2003. “’Laughing’ Rats and the Evolutionary Antecedents of Human Joy?” Physiology and Behavior 79(3): 533–547.

Parten, Mildred B. 1932. “Social Participation among Pre-School Children.” Journal of Abnormal and Social Psychology 27(3): 243–269.

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.

Plessner, Helmuth. 1999 [1924]. The Limits of Community: A Critique of Social Radicalism. Translated by Andrew Wallace. Amherst, NY: Humanity Books.

Podilchak, Walter. 1991. “Distinctions of Fun, Enjoyment and Leisure.” Leisure Studies 10(2): 133–148.

Provine, Robert R. 2000. Laughter: A Scientific Investigation. New York: Viking.

Reis, Harry T., Annie Regan, and Sonja Lyubomirsky. 2022. “Interpersonal Chemistry: What Is It, How Does It Emerge, and How Does It Operate?” Perspectives on Psychological Science 17(2): 530–558.

Reis, Harry T., Stephanie D. O’Keefe, and Richard D. Lane. 2017. “Fun Is More Fun When Others Are Involved.” Journal of Positive Psychology 12(6): 547–557.

Roberts, John M., Malcolm J. Arth, and Robert R. Bush. 1959. “Games in Culture.” American Anthropologist 61(4): 597–605.

Ryan, Richard M., C. Scott Rigby, and Andrew Przybylski. 2006. “The Motivational Pull of Video Games: A Self-Determination Theory Approach.” Motivation and Emotion 30(4): 344–360.

Sawyer, R. Keith. 2003. Group Creativity: Music, Theater, Collaboration. Mahwah, NJ: Lawrence Erlbaum.

Schiller, Friedrich. 1967 [1795]. On the Aesthetic Education of Man, in a Series of Letters. Edited and translated by Elizabeth M. Wilkinson and L. A. Willoughby. Oxford: Clarendon Press.

Schlegel, Alice, and Herbert Barry III. 1991. Adolescence: An Anthropological Inquiry. New York: Free Press.

Schüll, Natasha Dow. 2012. Addiction by Design: Machine Gambling in Las Vegas. Princeton: Princeton University Press.

Schwartzman, Helen B. 1978. Transformations: The Anthropology of Children’s Play. New York: Plenum.

Sharp, John, and David Thomas. 2019. Fun, Taste, and Games: An Aesthetics of the Idle, Unproductive, and Otherwise Playful. Cambridge, MA: MIT Press.

Simmel, Georg. 1949 [1910]. “The Sociology of Sociability.” Translated by Everett C. Hughes. American Journal of Sociology 55(3): 254–261.

Spencer, Herbert. 1872. The Principles of Psychology. 2nd ed., vol. 2. London: Williams and Norgate.

Stenros, Jaakko. 2014. “In Defence of a Magic Circle: The Social, Mental and Cultural Boundaries of Play.” Transactions of the Digital Games Research Association 1(2): 147–185.

Stern, Daniel N. 1977. The First Relationship: Infant and Mother. Cambridge, MA: Harvard University Press.

Suits, Bernard. 1978. The Grasshopper: Games, Life and Utopia. Toronto: University of Toronto Press.

Sutton-Smith, Brian. 1997. The Ambiguity of Play. Cambridge, MA: Harvard University Press.

Sutton-Smith, Brian, and Diana Kelly-Byrne. 1984. “The Idealization of Play.” In Play in Animals and Humans, ed. Peter K. Smith. Oxford: Basil Blackwell.

Templeton, Emma M., Luke J. Chang, Elizabeth A. Reynolds, Marie D. Cone LeBeaumont, and Thalia Wheatley. 2022. “Fast Response Times Signal Social Connection in Conversation.” Proceedings of the National Academy of Sciences 119(4): e2116915119.

Trammell, Aaron. 2023. Repairing Play: A Black Phenomenology. Cambridge, MA: MIT Press.

Vygotsky, Lev S. 1967 [1933]. “Play and Its Role in the Mental Development of the Child.” Soviet Psychology 5(3): 6–18.

Walker, Charles J. 2010. “Experiencing Flow: Is Doing It Together Better than Doing It Alone?” Journal of Positive Psychology 5(1): 3–11.

Winnicott, D. W. 1971. Playing and Reality. London: Tavistock.

Wittgenstein, Ludwig. 1953. Philosophical Investigations. Oxford: Blackwell.

Wolfenstein, Martha. 1951. “The Emergence of Fun Morality.” Journal of Social Issues 7(4): 15–25.

Yee, Nick. 2006. “Motivations for Play in Online Games.” CyberPsychology and Behavior 9(6): 772–775.

Zillmann, Dolf. 1988. “Mood Management through Communication Choices.” American Behavioral Scientist 31(3): 327–340.