Frame Games: The Role of Context in Defining Play Behavior Fabio Paglieri Interdisciplinary Centre for Play Research (CIRG) & Doctorate in Cognitive Science, University of Siena, Piazza S. Francesco 8, 53100 Siena, Italy [email protected]

Abstract. This paper aims to clarify the role of context in defining play behavior, for both the individual and society. I will argue that (1) relevant features of play behavior are no matter of logical levels of abstraction, but (2) they concern time closure, space limitation and consequences restriction. I will (3) include such features in a cognitive definition of play behavior, which in turn (4) would be developed in a more refined account of social play, considering different kinds of player and detailing their typical play behavior. Finally, (5) I will briefly discuss my own strategy of describing contextual features in cognitive terms, as attempt to develop a single theoretical framework for both mind and context.

Keywords: Cognitive modeling - Representing context and contextual knowledge

1

Introduction

Play studies are a multidisciplinary domain with a long, sometimes confused tradition [1, 2, 3, 4, 5, 6, 7, 8, 9]: after a period of partial oblivion, recently they excited a new wave of interest [10, 11, 12, 13, 14, 15], mainly focused on cognitive and social aspects of play activities. It is widely aknowledged that the notion of context raises many issues relevant to play and game studies. Here I do not intend to summarize all of them. Instead, I will focus on two topics: (1) the role of context in cognitive definition of play; (2) the role of context in social definition of play. First topic is mainly developed in chapter 2 and 3 of this paper: I take Gregory Bateson’s works as starting point, arguing for a cognitive revolution in play definition and a new account of play context. In chapter 4 I deal with social definition of play activities, discussing the relation between enjoyment expectations and goals typology, and its relevance to social value of play. Finally, chapter 5 summarizes paper contents, and a conclusive remark is put forward about integrating mind and context in play studies.

2

Bateson’s Paradox: Logical Frames for Psychological Issues

Gregory Bateson [16, 17, 18] was the first scholar to deeply investigate context issues in connection with play activities. His best-known claim was that each play activity requires meta-communication, i.e. a statement which provides a context for its own interpretation. So in play behavior, each single action carries the message “This is play”, which is crucial to interpretate the action correctly. Usually the message is implicit; nonetheless, it is always supposed to work in defining a proper interpretative context for play behavior. On this basic assumption, Bateson developes a theory of play as paradox of abstraction, i.e. a situation which triggers beneficial confusion between logical levels. According to Bateson, the message “This is play” necessarily generates a paradox of the Russellian or Epimenides type – a negative statement containing an implicit negative meta-statement. He claims that in play situation, e.g. in monkey play-fighting, the implicit statement “This is play” can be rephrased like this: “These actions in which we now engage do not denote what those actions for which they stand would denote”. In turn, Bateson considers the expression “for which they stand” synonymous of “which they denote”, and further developes the message in: “These actions, in which we now engage, do not denote what would be denoted by those actions which these actions denote” [17]. This would lead to a kind of liar paradox, like the phrase “This sentence is false”: if the sentence is false, it is false that is false, so it is also true; on the other hand, if the sentence is true, it is true that is false, so it is also false. In both cases there is a contraddicition. Bateson does not limit his analysis to single play messages, but suggests paradox spreads over the whole play behavior. The message “This is play” would establish a paradoxical frame as the proper interpretative context for any action during play activity. Roughly speaking, that would allow to assume contraddictory position in play situation, at least temporary – Bateson would probably say it allows to experience contraddiction in one’s own behavior. In fact, Bateson stresses the beneficial nature of such opportunity, and the psychological value of any paradox of abstraction. I do not intend here to dispute beneficial effects of paradoxical experience, on which I partially agree. Instead, I will criticize some assumptions of Bateson’s analysis, in order to highlight some weaknesses in his account of play behavior, and to suggest the opportunity for radical reformulation of the issue. I will consider the example provided by Bateson himself, concerning playfighting among animals: if we take the single action of biting playfully the opponent and we call it a “nip”, Bateson’s analysis claims that the nip denotes the bite, but it does not denote what would be denoted by the bite, e.g. aggression [17]. Here arises the paradox, because usually bites denote aggression: so how can the nip at the same time stands for a bite and does not stand for aggression? My answer is quite simple: I suggest there is no need at all to consider nips as denoting bites, so the whole argument has false premises. If we look closely to the relation between playful nips and serious bites, what do we see? Clear enough, there is imitation: a nip is a behavioral feature strictly modelled on another behavioral feature – the bite. But imitation does not imply denotation, i.e. what is being imited does not necessarily serve as content for the playful behavior. Why should we suppose that the nip stands for the bite in the same way as the word “bite” stands for

the action of biting? Quite the reverse: denotation seems to be the wrong category to understand play activities and their relations with “serious” behavior. From the semiotic point of view, Bateson suggests there are in play two denotations chained together, but the first term of the first denotation is incompatible with the second term of the second denotation: A stands for B and B stands for C, but A does not stand for C. I propose instead there are (no less than) two parallel codes to interpretate behavior, one for play, one for “real life”: the former imitates some features of the latter, but they are semantically independent one from each other, e.g. A* stands for B* and A stands for B. Empirical findings in cognitive ethology strongly support such claim [13, 15, 19, 20, 21, 22]. Field studies have extensively documented relevant features of play actions which mark them as specialized signs with specialized meanings. Besides, even some “punctation signals” have been aknowledged [21], i.e. highly specialized signals which delimit the duration of play activity – when it begins and when it is going to end. A well-known example of such behavioral punctation is the so called “play bow” of canids, in which the animal crouches on its forepaws, elevates its hind and may wag its tail: this is an utterly unambiguous play marker, which implies no paradox at all. This view of play communication has a relevant consequence, which is also a further criticism to Bateson’s analysis: the playful nip could denote aggression, and often do. In other words, play actions have often many serious meanings, despite their playful nature – a point which have been already emphasized by many resarches in experimental and developmental psychology [23, 24, 25, 26, 27]. Regarding aggression, there is a huge difference between pretended aggression and playful aggression: in theatre, each time Othello strangles Desdemona we are aware there is no real aggression between the actors, but only the pretense of aggression; on the contrary, in chess we witness one of the most striking occurrence of total war between human beings, in which each one tries by any (legal) means to overcome the opponent, i.e. the aggression is definitely real. Sometimes we have pretense in play, e.g. in symbolic play among children, and that actually mixes up different categories [28, 29, 30, 31]: but also in children pretense there is often an element of competition, which implies true aggression. Speaking of “true aggression”, and more generally of serious meanings for play activities, I do not intend to suggest there are no differences between aggression in play and aggression in “real life”. But I do intend to suggest these differences does not concern “degree of reality”, nor logical level of abstraction, nor metacommunication. In next chapter, I will propose an alternative view on the boundaries between play and non-play. Quite obviously, the whole issue is deeply intertwined with the notion of context. Bateson suggests we must consider play activities characterized by a peculiar logical context – namely, the paradoxical frame created by the message “This is play”. I criticize this view, arguing that logical frames are misleading in defining play, and we should look for a different account of the phenomenon. Nonetheless, context remains a crucial issue, as far as we need some kind of frame to distinguish play from other typologies of behavior. As conclusive remark on the works of Gregory Bateson, it should not be forgotten the real nature of his interest in play behavior, which was mainly

concerned with the practical effects of play on human mind. His insistence on paradoxical features of play was due to the intuition that paradoxes of abstraction could prove to be beneficial experiences for the subject. To criticize his logical analysis does not imply to deny the value of such intuition. Here we questioned only its relevance in defining play behavior.

3

What Kind of Context for Play?

We have already seen play activities do not simply relate to a context, but they rather create a context, i.e. define a frame to guide the interpretation of single actions. In a developmental perspective, this property is rooted in the first occurrences of children play, when, according to Piaget, the assimilation of reality to imagination is overwhelming, while there is very little or no accomodation at all of mental concepts to real data [4]. Even in adulthood, with a much richer understanding of the relationship between reality and imagination, play still have the power of creating a context of its own. But now we need to investigate what kind of context is relevant for play behavior. Some authors [5, 6] have emphasized a key-feature of play context, quite unrelated to Bateson’s logical levels: the separation in time and space between play activities and other behavioral patterns. Play has always well-defined time limitation: there is a precise moment in which play begins, and at that time players already know when it is going to stop, either because there is a predefined time-span for playing (e.g. in soccer), or because the end-state of the activity is known (e.g. in chess). Even interruptions are usually planned in advance, as in many sports. More generally, play always has a definite time closure, and it does not support infinite duration. In many cases, this is paralleled by spatial limitation of the play area: play activities are to be performed in dedicated spaces, as playground, fields, chessboards, charts, clubs, and so on. Links between this “holy playground” and ritual value of play activities have often been stressed by anthropologists and social scientists [5, 6, 12, 32]. But in order to appreciate the capital importance of separation in play behavior, we need to consider its deontic meaning – that so far has often been understimated or utterly missed. Time closure and spatial limitation do not only provide physical boundaries for play activities: they also define basic normative features of play context. In fact, temporal and spatial separation applies not only to play actions, but also to their practical consequences. For “practical consequences” I intend here any practical effects of an action performed in play that still last after the end of play activity. Such pratical consequences are forbidden for play actions, i.e. actions with practical effects beyond play duration are not allowed in play. This is the point where Bateson was misleaded: instead of wondering how the nip differs from the bite, he should have investigated why they differ that way. The reason is simple enough: the bite has a lot of practical consequences the nip does not share (e.g. it harms who receives it, endangering the aggressor with the threat of further revenge), and that is why playing animals exchange nips instead of bites. In play we need to avoid practical consequences, and our actions are performed according to this purpose.

Within temporal and spatial boundaries of play context, a (meta)rule holds: there must be no practical consequences to the actions performed here, after the play is over. But this rule has a relevant corollary: if a play action has no practical consequences outside play context, then it is permitted while playing – no matter what are its nature and moral value elsewhere. We are forced to appreciate this aspect each time we witness instances of cheating and deceiving in play – that is, allowed cheating and deceving, according to game rules and not against them. E.g., bluffing is the most famous strategy in playing poker, and someone could even argue it is the only possible strategy for that game. Besides, various types of deceit have been highly specialized in card games and boardgames, while in many sports feints and counter-feints are usually employed to overcome the opponent. And so on. It should be noticed that all these actions, which in play activities are not only allowed, but even praised, would be instead prohibited and condemned outside play context. The same holds for the aggressive attitude which is the very essence of any competitive game: such an aggression will not be usually permitted in social realtionships, but it is given free rein in play context. The only restrictions are that (1) aggression must be limited to play context, without further consequences, and (2) it must be expressed according to game rules, whatever they are. So play behavior defines a peculiar normative context in which, provided there are no practical consequences outside play, many actions usually forbidden are allowed and sometimes even required and rewarded. This has nothing to do with the degree of reality of players’ attitudes (e.g. aggression), or with the level of abstraction required to perform play actions. I wish to insist that play behavior is absolutely “true”, i.e. it is as real and serious as any other activity. The difference is that consequences of play actions are strictly bounded to predefined temporal and spatial limitations, i.e. to the play context. Such context has been defined as “deontic” and “normative” to stress that acceptance of play context is always matter of free will and social negotiation. In this view, to play is to have certain beliefs regarding one’s own activity, and to play together is to share such beliefs with other players. This is a sort of “cognitive revolution” in play studies, since individual and social definition of play are described as a peculiar set of mental states (e.g. beliefs and goals), instead of a collection of behavioral traits [33]. Regarding context, it means that also contextual features of play will have to be expressed as particular beliefs – i.e. beliefs concerning play context. I suggest play context is a set of restrictions on practical consequences of play actions, which follow from the temporal and spatial boundaries associated to play behavior. However, it should now be remarked that such contextual description is not enough to define play. In other words, play context is only one feature of play definition – a necessary, but not sufficient feature. As tentative suggestion, I will propose here a more complete cognitive definition of play activity, that I have already discussed elsewhere [33]. So I will not review all its advantages and drawbacks; instead, I will insist on the role of context in such definition. My claim is that: X agrees that Y is playing iff X believes that:

1) Y is doing the action by her own free will, without any constraints and not for duty (belief of freedom); 2) Y is observing a set of rules Z which is arbitrary, not imposed by physical necessity or social pressures (belief of arbitrary regulation); 3) Y is pursuing motivation Q defined by the set of rules Z (belief of selfmotivation); 3.1) Y is pursuing motivation Q only for the lenght of current activity (subbelief of separation); 3.2) Y is pursuing motivation Q at her best (sub-belief of personal commitment); 4) Y knows in advance temporal limitations of current activity, i.e. she knows when it starts and when it is going to stop or to be interrupted (belief of limitiation); 5) Y does not forecast practical consequences to her action beyond temporal limitations of current activity (belief of restriction); 6) Y does not know in advance precise outcomes of her action (belief of uncertainty). In this set of beliefs, contextual features are represented by belief of limitation (temporal boundaries) and belief of restriction (normative boundaries), while beliefs concerning motivation suggest that play should also be self-motivated activity – a point that I will discuss further in the next chapter. Finally, I did not include in the list any belief dealing with spatial limitation, because I suppose time closure is already enough to discriminate between what is allowed and what is forbidden in play. There are now enough elements to summarize the role of context in cognitive definition of play. We can say that: (1) play depends on a context, i.e. contextual requirements of time closure and consequences restriction must be met, in order to classify an activity as “play” and an action as “playful”; (2) play creates a context, i.e. establish contextual rules that must be followed by players, and that contributes to the meaning of each play action.

4

Goals, Enjoyment, and Social Definition of Play Activities

Now it is time to widen the scope of our investigation, including social influence and social values in play context. So far, we have considered a sort of highly idealized situation, in which play occurs indipendently by the broader context of society, and players are the only relevant subjects for our research. But it should not be forgotten that play is also a social phenomenon, not only for players, but also for spectators and fans, if there are any. Before exploring social dimension of play, I wish to stress its link with the cognitive definition aforementioned. We are to consider such definition as a prototypical account for play activities, meant to define the core of this behavior. In society, however, play phenomena are rarely met in prototypical occurrences: they are usually mixed up with different behavioral patterns, involved in many processes at once, and embedded in larger social dynamics. As a result, social definition of play activities requires a certain degree of variation, in which different kinds of play activity, with different social meaning, are taken in account [11].

In order to capture such variation in our theoretical model, we will need to relax some features of the prototypical description: namely, belief of separation and belief of restriction – i.e. beliefs concerning context. Even so, this does not lead to dismiss contextual issues. Quite the reverse: relaxing some constraints of our model will give us a richer and deeper understanding of context dynamics, provided we still mantain a formal description of such features. 4.1

Goal Typologies

I mentioned in previous chapter the opportunity for a cognitive revolution in play studies: now we will see that such conceptual shift make for good social definition of play as well. But first I have to specify what kind of cognitive revolution I am hinting at. Here I refer to symbolic approaches to cognition, i.e. studies which use mental states (beliefs, goals, desires, emotions) to model cognitive processes [34, 35, 36, 37, 38, 39, 40]. Other approaches have been proposed in cognitive science, e.g. sub-symbolic models with neural networks [41, 42, 43, 44, 45], but I will not use them here. The main methodological reference of my model are the works of Cristiano Castelfranchi and his research team at the National Council of Research in Rome [46, 47, 48, 49, 50, 51, 52, 53]. They propose a cognitive architecture mainly based on two kinds of mental states: beliefs and goals. Each one of this category is further analysed and unfolded in sub-typologies. Concerning play, we already discussed what are the relevant beliefs in its definition. It is now time to consider also its characteristic goals. The notion of goal has one basic definition and various sub-typologies. Basically, a goal is defined as an anticipatory representation of some situation, which has the property of leading agent’s behavior towards such situation. Developing this claim, it is possible to describe many different teleological notions: terminal goals, sub-goals, meta-goals, functions, motivations, expected effects, and so on [50]. Table 1 summarizes some of them and their relevant features. Table 1. Goal typologies and their definitions

Goal typology

Description

Terminal goal

A goal which is the ultimate aim of an agent behavior, and it is not instrumental to any further goal (cf. sub-goal).

Sub-goal

A goal which is instrumental to some further goal, as useful step in a greater plan (cf. terminal goal).

Meta-goal

A goal which represents an operative principle in agent’s planning, influencing her decision at different stages of the activity (e.g. the goal of maximizing pleasure and minimizing pain).

4.2

Function

A goal which is not represented in the agent mind, but nevertheless influences her conduct or use in a specific direction, either by being represented in the mind of another agent (e.g. the chair has the function to bear me because its designer assigned it that goal) or even being represented in no mind at all (e.g. selective pressure in nature).

Motivation

The goal which makes the agent choose to perform a particular activity, i.e. without such goal the activity would not be performed at all (cf. expected effect).

Expected effect

Each effect of the activity which is forecasted by the agent. If such effect is beneficial for the agent, it could be one of her goals, but it is not the one who makes her act that way (cf. motivation).

Motivational Patterns and Play Context

In order to achieve a complete cognitive definition of play, we should be able to indicate some motivational pattern characteristic of such activity. Two strategies are available: either we try to find out and describe a peculiar goals architecture, i.e. a goals structure typical of play activities, or we look for a characteristic goal, i.e. a goal common to every play activity. First strategy leads to develop the idea that play is somehow “self-motivated”. But what does that mean? It could be play-goals have to be selected among goals set by play rules, without any further purpose – the same idea expressed by the belief of self-motivation (see chapter 3). But this is just a general rule, not a specific architecture. To have a structural explanation of play motivation, there should be some structural property which characterizes goals structure in players’ mind, e.g. convergence or divergence in planning. Evidence does not support any precise claim at this level, and goals structure seems to be matter of individual variation among players. Some subjects typically play in goal-oriented fashion, aiming precisely to specific goals (e.g. winning, having fun, or whatever), and they usually achieve high degree of planning convergence. At the same time, other players show a definite exploratory inclination, which leads them to change frequently their goals, i.e. presenting patterns of planning divergence. So structural explanations alone are not adequate to account for players’ motivation: we need also to single out some characteristic goal of play activities. Here I propose a kind of hybrid strategy, concerned both with structure and with contents of play-goals. I suggest play presents at least one characteristic goal, the goal of enjoyment (see further in this chapter for details on the notion of “enjoyment”). But this goal does not need to serve always as motivation for players: it can move through different goal categories, producing a wide variety of play configurations (see Table 2).

Table 2. Enjoyment and goal typology

Goal category (enjoyment as...): MOTIVATION Type of player: true player Typical play behavior: the player is playing mainly to have fun and any other goal is a sub-goal for enjoyment, so she presents a strong goal-oriented attitude Goal category (enjoyment as...): EXPECTED EFFECT Type of player: pseudo-player Typical play behavior: the player pursues some other goal, which differs from enjoyment and it is not included in the definition of play; so she is not playing at all, in a subjective sense Goal category (enjoyment as...): META-GOAL Type of player: hedonistic player Typical play behavior: the player adjusts her actions to the “rule of the greatest enjoyment”, so she is usually having fun, but can be poorly goal-oriented Goal category (enjoyment as...): TERMINAL GOAL Type of player: demanding player Typical play behavior: the player demands to enjoy the activity as one of her highest priority; she shows involvement without plan convergence, and she will often quit any game she does not enjoy. Goal category (enjoyment as...): SUB-GOAL Type of player: occasional player Typical play behavior: partecipation in the game is conditioned to have some assurance of enjoyment, so the player does not take part in every game, but only in those she already knows to be enjoyable. Goal category (enjoyment as...): FUNCTION Type of player: slave player Typical play behavior: enjoyment of the player is pursued and planned by some external agency, e.g. society, so the player is being manipulated in her play, often without even knowing it The basic idea is that enjoyment can have different roles in different play activities: these roles are well-described in terms of goal typologies and contribute greatly to behavioral variety of play occurrences. If we define enjoyment as a binary relationship E between an agent p and an activity α, the fact that p enjoy α can be written as E(p,α). When E(p,α) is the function of α, that means agent p does not necessarily intend to enjoy doing α, but this result is teleologically produced by factors external to the agent’s mind (e.g. society pressures). Here the enjoyment of agent p is planned and worked out by someone else. This could be the case with the ancient idea of panem et circensem (food and fun) as means of social control – and I

guess we are not far from that even nowadays. In another context, E(p,α) can be a sub-goal, i.e. a goal which is pursued only as a mean for further goals. Here enjoyment is a sort of precondition for taking part in the activity: this kind of players can be seen as “selective”, i.e. they choose to play only on certain occasions, or they take part only in some games and not in others. Still different is the case when enjoyment is a meta-goal for players, i.e. it rules their decisions at each stage of the game. Such players are always fun-driven, but they can be poorly goal-oriented: they have no problem in forsaking goals defined by play rules, whenever doing so could assure them to have more fun. I suggest to characterize them as hedonistic, since they always choose the path of greater enjoyment. However, this attitude needs to be distinguished by the case in which enjoyment is a terminal goal for players, i.e. players consider fun as one of the greatest aim in their live, and a true priority for their behavior. That means enjoyment is their main goal not only in play, but in every activity: whenever they are playing, they always demand to enjoy it – in this sense, I would call them demanding players. From a cognitive point of view, hedonistic players and demanding players are quite different: the former stick to the rule “whenever I am doing something, I should try to enjoy it”; the latter follow a different rule, which is “if something is enjoyable to me, I should do it”. As a result, they show different behaviors in play: hedonistic players can forsake play goals and linger with many diversions, but basically they still participate in play; on the contrary, demanding players do not accept to continue a game they are not enjoying, and often they will quit it altogether. We have another interesting occurrence when enjoyment is an expected effect of player’s activity, i.e. she knows that such activity is likely to give her enjoyment, but this is not the reason compelling her to do it. Here enjoyment is not a key-feature in player’s motivational pattern: if she regards enjoyment as a positive result, she could make some (secondary) efforts to achieve it; otherwise, she would not be concerned at all with enjoying play. I suggest this is exactly the case of many professional players, i.e. someone who is paid to play. Professional players have always been a nuisance in defining play, because they stay at the crossing between play and work – two categories usually contrasted one with each other. In my view, professional players in social context have to be considered as proper players, i.e. they are usually seen as people who play, or even as the idealized player’s type. However, considering the cognitive features of their behavior, someone could argue they are not playing at all, because (1) their motivation is not defined by game’s rules, and (2) their actions have a lot of practical consequences beyond play context. So I would propose to call them pseudo-players, to stress their somewhat fuzzy status. But our theoretical model allow us to say something more precise about the relation between being a professional player and truly playing. In fact, it should be remarked that our definition is concerned with mental states, i.e. individual convictions regarding states of the world and one’s own motifs. So, to have enjoyment as expected effect in play means to aim deliberately all the time at something else, e.g. money and fame. I do not believe that every professional player shows such a cynical attitude. In my view, any professional player could be (and often is) a true player, i.e. she could play having enjoyment as motivation, and other

results (richness, social admiration, someone’s love, etc.) as expected effects – exactly the reverse of pseudo-player’s motivational pattern. Finally, there is the case when enjoyment works as motivation for players, i.e. they are playing in order to enjoy it, not because enjoyment is a general priority of their lives (cf. demanding players), but because it is the characteristic goal of this specific activity. Here enjoyment is linked to play-goals, i.e. players measure their enjoyment according to goals defined by play rules. As a result, we have that particular type of players we first mentioned, the most goal-oriented and dedicated ones: they use to play just for enjoying it, and that also means they can end without a single bit of enjoyment, if they miss their goals. This is not a drawback of our definition, but a natural consequence of linking player’s enjoyment to goals’ satisfaction. By the way, disappointment in play, expecially in competitive games, is quite common, and should not be regarded as a masochistic feature. On the contrary, it results from a very basic distinction: the difference between intending to enjoy something (motivation for playing), and actually enjoying it (possible outcome of play behavior). In Table 2, players that have enjoyment as motivation are defined true player, but only because they fulfill our prototypical definition of play behavior. I do not intend to imply that other player’s typologies are somewhat less important, but I wish to suggest they are less typical, at least regarding cognitive features of play behavior. In social instances of play, we often meet untypical players, i.e. people who play without fulfilling some requirements of our prototypical definition. Besides, we could have different opinions regarding features truly essential to play definition, and that would lead to accept different player’s typologies as proper model of play behavior. This is exactly why we need many related account of play behavior in social context, as I try to provide in this chapter. By way of suggestion for future studies, I strongly advocate further development, modification, and even criticism of my own model. Regarding what is summarized in Table 2, I welcome the idea of treating player’s typologies as dynamic concepts, which can sometimes overlap and interact with each other. That means enjoyment could be assigned to more goal typologies at once – with some obvious limitations: from definition of goal typologies in Table 1, it follows that enjoyment cannot be at the same time motivation and expected effect, nor terminal goal and sub-goal. But it could be motivation and terminal goal at once, sketching a new, interesting player’s typology: the addicted player, i.e. someone who is urged to play by an enormous desire for enjoyment, whom she is not able to satisfy otherwise. Such a type is well-known to anyone who deals with compulsive play, a well-studied pathology, unfortunately too often considered outside the broader context of play behavior [54, 55, 56, 57, 58]. A cognitive and social model of play activities could provide us with the necessary framework for exploring such phenomena, including their pathological developments. Of course, the account presented here meets a major drawback: we still miss a clear definition of what enjoyment is. There is also need of more refined conceptual analyses on the relationship between enjoyment and other similar notions, such as “amusement” and “fun”. Each of them is deeply related to play behavior, but their role has still to be studied. At present stage of research, there are no easy answers on these issues, and we can only stress the importance of further studies on the notion of

enjoyment, both in play and in other activities, both in conceptual analyses and in empirical researches. However, the interplay between enjoyment and goal typology give us stimulating suggestions for studying motivational patterns in play, which hold even without a formal definition of enjoyment.

5

Conclusion

The model presented here aims to combine context issues with cognitive and social definition of play. In chapter 3, we discussed how play context is embedded in cognitive definition of play, via belief of limitation and belief of restriction. Then we noticed that only relaxing such beliefs we could deal with social variation of play phenomena, that usually is ascribed to context. Finally, we developed a formal account of such variation, in which context issues are again internalized as different goal typology for the purpose of enjoyment. This strategy of dealing with contextual features is somewhat unusual, but I consider its results quite promising. For instance, it gives us new insight on the status of professional player, which is a typical context-dependent issue: e.g. soccer players are either playing or working, depending on context. But relevant context could be described as goal typology for enjoyment: whether professional players have enjoyment as motivation, they are actually playing, despite any practical consequences of their activity; but if enjoyment is only an expected effect of their behavior, while some practical consequence is their true aim, then they are definitely working, if money is their goal, or performing some other serious activity disguised as play (e.g. trying to positively impress potential partners who witness their play, as in males play-fighting among some animals [13, 27, 59]). But how much is useful to internalize contextual features in play definition? This is an open question, which should be answered both with empirical observation (e.g. testing cognitive and social definition of play behavior in psychological experiments [60] and sociological surveys) and with further developments of the model, e.g. inquiring the true nature of enjoyment. However, I forecast a general advantage of such theoretical strategy: it will allow and demand us to compare two foundamental features of human behavior – mind and context. Scholars in cognitive science already believe that such topics should be studied together [48, 49, 50, 53]. The same holds for play, whose definition will necessarily require both cognitive features and contextual elements.

References 1. 2. 3. 4. 5.

Groos, K.: The Play of Animals. Chapman and Hall, London (1898) Groos, K.: The Play of Man. Appleton, New York (1901) Piaget, J.: Le jugement moral chez l’enfant. F. Alcan, Paris (1932) Piaget, J.: La formation du symbole chez l’enfant. Delachaux et Niestlé, Neuchâtel (1945) Huizinga, J.: Homo ludens: proeve eener bepaling van het spel-element der cultur. H. D. Tjeenk Willink, Haarlem (1938)

6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

18. 19. 20. 21. 22.

23. 24. 25. 25.

Caillois, R.: Les jeux et les hommes. Gallimard, Paris (1967) Winnicott, D.: Playing and Reality. Tavistock Publications, London (1971) Bruner, J.S., Jolly, A., Sylva, K. (eds.): Play: Its Role in Development and Evolution. Basic Books, New York (1976) Erikson, E.: Toys and Reasons. Norton, New York (1977) Pellegrini, A.D.: Ecological Context of Childrens Play. Greenwood Publishing Group, Westport CT (1989) Bencivenga, E.: Giocare per forza: critica della societa del divertimento. Mondadori, Milano (1995) Sutton-Smith, B.: The Ambiguity of Play. Harvard University Press, Cambridge MA (1997) Bekoff, M., Byers, J.A. (eds.): Animal Play: Evolutionary, Comparative and Ecological Approaches. Cambridge University Press, New York (1998) Pellegrini, A.D., Smith, P.K.: Physical Activity Play: the Nature and Function of a Neglected Aspect of Play. Child Development, 69 (1998) 577-598 Burghardt, G.M.: The Genesis of Animal Play. Chapman and Hall, New York (1999) Bateson, G.: Metalogue: About Games and Being Serious. ETC.: A Review of General Semantics, 10, 3 (1953) 213-217 Bateson, G.: A Theory of Play and Fantasy: A Report on Theoretical Aspects of the Project for the Study of the Role of Paradoxes of Abstractions in Communication. American Psychiatric Association, Psychiatric Research Reports, 2 (1955) 39-51 Bateson, G.: The Message “This Is Play”. Josia Macy Jr. Foundation, Princeton NJ (1956) Fagen, R.M.: Animal Play Behavior. Oxford University Press, New York (1981) Bekoff, M.: The Communication of Play Intention: Are Play Signals Functional? Semiotica, 15 (1975) 231-239 Bekoff, M.: Play Signals as Punctuation: The Structure of Social Play in Canids. Behaviour, 132 (1995) 419-429 Bekoff, M., Allen, C.: The Evolution of Social Play: Interdisciplinary Analyses of Cognitive Processes. In: Bekoff, M., Allen, C., Burghardt G.M. (eds.): The Cognitive Animal: Empirical and Theoretical Perspectives on Animal Cognition. The MIT Press, Cambridge MA (2002) Klein, M., Heimann, P., Money-Kyrle, R.: New Directions in PsychoAnalysis. The Significance of Infant Conflict in the Pattern of Adult Behaviour. Tavistock Publicatons, London (1955) Blurton Jones, N.: Ethological Studies of Child Behavior. Cambridge University Press, Cambridge MA (1972) Humpreys, A.P., Smith, P.K.: Rough and Tumble, Friendship, and Dominance in Schoolchildren: Evidence for Continuity and Change with Age. Child Development, 58 (1987) 201-212 Boulton, M.J.: Concurrent and Longitudinal Relations between Children’s Playground Behavior and Social Preference, Victimization, and Bullying. Child Development, 70 (1999) 944-954

26. 27.

28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47.

Dunn, J., Hughes, C.: “I Got Some Swords and You’re Dead!”: Violent Fantasy, Antisocial Behavior, Friendship, and Moral Sensibility in Young Children. Child Development, 72 (2001) 491-505 Pellis, S.M.: Keeping in Touch: Play Fighting and Social Knowledge. In: Bekoff, M., Allen, C., Burghardt G.M. (eds.): The Cognitive Animal: Empirical and Theoretical Perspectives on Animal Cognition. The MIT Press, Cambridge MA (2002) Nachmanovitch Miller, S.: The Playful, the Crazy and the Nature of Pretense. Rice University Studies, 60 (1974) 31-51 Garvey, C.: Play. Harvard University Press, Cambridge MA (1977) Fein, G.: Pretend Play in Childhood: An Integrative Review. Child Development, 52 (1981) 1095-1118 Howes, C.: Sharing Fantasy: Social Pretend Play in Toddlers. Child Development, 56 (1985) 1253-1258 Geertz, C.: The Interpretation of Cultures. Basic Books, New York (1973) Paglieri, F.: Credendo di giocare. Verso un’interpretazione cognitivista dei processi ludici. Sistemi Intelligenti, 14 (2002) 371-415 Miller, G., Galanter, E., Pribram, K.H.: Plans and the Structure of Behavior. Holt, Rinehart & Winston, New York (1960) Fodor, J.: The Language of Thought. Crowell, New York (1975) Putnam, H.: Mind, Language, and Reality. Cambridge University Press, Cambridge MA (1975) Schank, R.C., Abelson, R.P.: Scripts, Plans, Goals, and Understanding: An Inquiry into Human Knowledge Structures. Erlbaum, Hillsdale NJ (1977) Johnson-Laird, P.: Mental Models: Toward a Cognitive Science of Language, Inference and Consciuosness. Cambridge University Press, Cambridge MA (1983) Minsky, M.: The Society of Mind. Simon & Schuster, New York (1986) Searle, J.: The Rediscovery of the Mind. The MIT Press, Cambridge MA (1992) Hinton, G., Anderson, J. (eds.): Parallel Models of Associative Memory. Erlbaum, Hillsdale NJ (1981) Pylyshyn, Z.: Computation and Cognition: Toward a Foundation for Cognitive Science. The MIT Press, Cambridge MA (1984) Rumelhart, D.E., McClelland, J.L., PDP Research Group: Parallel Distributed Processing: Explorations in the Microstructures of Cognition. 2 voll., The MIT Press, Cambridge MA (1986) Churchland, P.M.: A Neurocomputational Perspective. The MIT Press, Cambridge MA (1989) Churchland, P.S., Sejnowski, T.: The Computational Brain. The MIT Press, Cambridge MA (1992) Miceli, M., Castelfranchi, C.: A Cognitive Approach to Values. Journal for the Theory of Social Behaviour, 2 (1989) 169-194 Castelfranchi, C., Miceli, M., Cesta, A.: Dependence Relations among Autonomous Agents. In: Y. Demazeu, E. Werner (eds.): Decentralized AI. Elsevier, Amsterdam (1992) 215-231

48. 49.

50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60.

Castelfranchi, C., Conte, R.: Emergent Functionalitiy Among Intelligent Systems: Cooperation Within and Without Minds. AI & Society, 6 (1992) 7893 Conte, R., Castelfranchi, C.: Mind Is Not Enough. Precognitive Bases of Social Action. In: J. Doran, N. Gilbert (eds.): Simulating Societies: The Computer Simulation of Social Processes. UCL Press, London (1994) 267286 Conte, R., Castelfranchi, C.: Cognitive and Social Action. University College, London (1995) Castelfranchi, C.: Emergence and Cognition: Towards a Synthetic Paradigm in AI and Cognitive Science. In: H. Coelho (ed.): Progress in Artificial Intelligence. Springer, Berlin (1998) 13-26 Castelfranchi, C.: Through the Agents’ Minds: Cognitive Mediators of Social Action. Mind & Society, 1 (2000) 109-140 Castelfranchi, C.: The Theory of Social Functions. Challenges for MultiAgent-Based Social Simulation and Multi-Agent Learning. Cognitive Systems, 2 (2001) 5-38 Wykes, A.: Gambling. Aldus Books, London (1964) Halliday, J., Fuller, P.: The Psychology of Gambling. Harper & Rowe, London (1974) Dickerson, M.G.: Compulsive Gamblers. Longman, London and New York (1984) Dickson, L.M., Derevensky, J.L., Gupta, R.: The Prevention of Gambling Problems in Youth: A Conceptual Framework. Journal of Gambling Studies, 18 (2002) 97-159 Bechara, A.: Risky Business: Emotion, Decision-Making, and Addiction. Journal of Gambling Studies, 19 (2003) 23-51 Smith, P.K. (ed.): Play in Animals and Humans. Blackwell, New York (1984) Smith, P.K., Vollstedt, R.: On Defining Play: An Empirical Study of the Relationship between Play and Various Play Criteria. Child Development, 56 (1985) 1042-1050