Unhappy families : using tabletop games as a technology to understand play in education

Research in Learning Technology 2018. @ 2018 J. Lean et al. Research in Learning Technology is the journal of the Association for Learning Technology (ALT), a UK-based professional and scholarly society and membership organisation. ALT is registered charity number 1063519. http://www.alt.ac.uk/. This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license.


Introduction: tabletop games as playful technology
The inclusion of an article on tabletop games in a journal whose first editorial speaks of an interest in 'the future of computer-based learning technology' (Jacobs 1993, p. 3) might appear provocative.Against a backdrop of easily recognisable (digital) technologies, it is our contention that tabletop games provide an approach to playfulness in game play that usefully supplements studies based on their now more prevalent digital counterparts.Here, we propose tabletop games as a form of technology, both related to and distinct from digital technology, and also as a means through which to better understand the specific affordances of the digital.In other words, by exploring the use of analogue technologies such as tabletop games, we hope to prompt critical reflection on the pedagogical use of digital technologies (Selwyn 2010).In particular, by examining 3 (page number not for citation purpose) game, played in its analogue and digital forms, might lead to very different player experiences because of this.These aspects are of particular interest in an educational context because they allow us to focus on discussions around the link between play and games that might otherwise be overlooked when focusing exclusively on the digital.

What is play?
The relationship between games (as technology) and play (as it arises from that technology) is well illustrated by the language used to delineate the two concepts in Caillois' Man, Play and Games (2001): Man, Play, and Games was written in French, Caillois' native tongue.Many languages do not have separate words for 'game' and 'play'.In French, for example, game is 'jeu' and play is 'jouer', the verb form of the same word.The original title of his book is Les Jeux et les Hommes (Games/Play and Man).(Salen and Zimmerman 2004, p. 308) The overlaps and absences in this terminology capture neatly the question that impels this article, namely, the role of 'play' in 'game-play': 'the experience of the game set into motion through the participation of players' (Salen and Zimmerman 2004, pp. 309-310).Our concern is with the relation of 'play' and 'games'; an overlap manifested in this language issue.This is at the heart of Caillois' understanding of games as a continuum between ludus (highly regulated games) and paidia (unstructured, spontaneous playfulness).
A similar 'language issue' occurs when we turn attention to playfulness, a term that, in English, forges a connection between player, playing and playful.Disentangling these terms is one of the challenges of this article.In our definition of playfulness, we turn initially to the work of De Koven, who writes: Playfulness is a gift that grants you great power.It allows you to transform the very things that you take seriously into opportunities for shared laughter, the very things that make your heart heavy into things that make you rejoice, it turns junk into toys, toys into art, art into celebration.It turns walking into skipping, skipping into dance.It turns problems into puzzles, puzzles into invitations to wonder.(De Koven 2014, p. 31) Alongside the validatory nature of De Koven's definition (playfulness equates to positive experience), play is here defined in terms of social interaction (its laughter is 'shared') and, more notably, it offers a clear distinction between the technology employed ('things', 'junk' and 'toys') and the playful uses to which they are put.In this, De Koven's understanding of games is close to Sicart's description of toys as 'enablers/vehicles ' (2014, p. 42), in other words facilitating props, of play.De Koven's definition of playfulness becomes clearer when placed alongside his discussion of what playfulness is not, or where it is not found: Here's a thought to think about: most games don't encourage playfulness.[…] If by games you mean sports, or games with official rules, or traditional games like shuffleboard and chess and bridge; playfulness is not what you'd think of as happening very much.(2014, p. 32) Here, there is a distinction between games and playfulness (and perhaps between playing and playfulness), with any overlap only seeming to exist in games that lack a competitive element.This precludes playfulness from those games that are characterised by what Caillois calls agôn, that is, games in which an 'equality of chances is artificially created, in order that the adversaries should confront each other under ideal conditions, susceptible of giving precise and incontestable value to the winner's triumph ' (2001, p. 14).It is this aspect of playfulness that we set out to examine in our workshop.We intended to provoke discussion of the following questions: It is the theoretical tension between playfulness and games that we set out to examine in our workshop, exploring the question of whether playing tabletop games in particular enables a form of playfulness which foregrounds the distinction between play and games.In asking this question, we aim to reframe discussions of all games in education to understand whether the perceived benefits of playing games in educational settings match up with the experience of playing them.Ultimately, the workshop suggested several issues that were not related to the above questions, but to the broader theme of the role of tabletop games in discussions of digital technology.

Playful Gloom
The 2017 Playful Learning Conference, a gathering of academics, game designers and education professionals with a particular interest in adult play, provided an audience with which to test our concept that tabletop games can be used as a technology to explore the idea of play as experience.Participants (mainly teachers, researchers and students, all of whom had an interest in games and play) were asked to play a tabletop game in order to explore the questions above.
The tabletop game that we selected for the session was Keith Baker's card game Gloom (Baker 2005).The game-play and underlying mechanics of this game of 'Unhappy Families' are described in an interview with Baker: You have a family of characters, and you want to construct the most tragic tale possible about your family -a tale that will eventually end in their demise.This is accomplished using a deck of transparent plastic cards.Each player places a family in front of them, and over the course of the game, you layer cards on top of a character to create an evolving story.Each card is something that could happen to a character; they might 'Break Many Bones' or be 'Pestered by Poltergeists'... or on the positive side, they could be 'Married Magnificently' or 'Delighted by Ducklings'.Each card modifies a character's score; because the cards are transparent, this creates a cumulative score.(Baker & Illingworth 2017) In addition to placing cards onto other cards to increase or decrease their scores, the game's instructions invite players to tell the story of their 'unhappy family'; however, this constructed narrative has no impact on the final score.This apparently extraneous instruction in the game's rulebook, an inherently 'playful' rule in that it has no connection to the game's outcome, institutes a degree of ambiguity as to what 5 (page number not for citation purpose) constitutes victory in Gloom -achieving the lowest happiness and winning (by the rules) or telling a good story.Baker expands on this: Like many role-playing games you could make the case that if you have fun creating the story you have 'won', even if technically you lost.To be clear: I think that the story is the most important piece when it comes to enjoying the game ... while the score is most important when it comes to winning it.Players are never judged on their stories.They don't have to tell stories, and they can win a game of Gloom without ever telling a tale.But in my opinion, if you play without story you're missing out on the best part of the experience that Gloom provides.(Baker & Illingworth 2017) This suggests that Baker's understanding of the relationship between play (in his words 'fun') and games falls somewhere between that of Caillois and De Koven.They are not the same thing, but likewise participating in a competitive game can be thought of as playful.According to this understanding of the game, games of competition (Caillois' agôn) need not preclude playfulness.
In the session at the Playful Learning conference, we invited the participants to reflect on what it means to 'win' a game of Gloom.Despite the invitation to play playfully the game's rules are also clear on what constitutes victory, so the experience of playing Gloom brings into question the role of the game's mechanics in relation to other competing ways of playing.Our 'Playful Gloom' session was thus initially presented in line with Caillois' understanding of games as a continuum between ludus and paidia, a debate we placed alongside the long-standing 'ludology-narratology' debate that sets games against stories (see Bogost 2017;Frasca 2003;Juul 2005).In order to do this, we asked the participants to play Gloom, to make a note of their experiences, and to report back to the rest of the group on their experiences of 'winning' or 'losing' at Gloom.
In observing the play session, we paid close attention to the three aspects of tabletop games that we had identified prior to the session (i.e.sociability, adaptability and tactility), observing each of them amongst the different groups of players.
First, we suspected that tabletop games would provide an immediately accessible shared social space for experience to take place.Single-player games, an uncommon subset of games in tabletop gaming, became a mainstay of early digital gaming, though, with the advent of shared technologies, 'the single phenomenon appears to have been a temporary abnormality' (Schell 2015, p. 394).While playing digital games is not necessarily a solitary activity, it does not always provide the framework through which social interactions can be provided (Jansz and Martens 2005).However, in playing tabletop games (other than those specifically designed to be played solo or 'solitaire'), players see the immediate reactions of fellow players, read their body language and can comment on their decisions.As Stewart Woods puts it: When a group of players sit down around a table to play a game, the social metagame begins.The very act of sitting together to engage in competitive play establishes a framework for social interaction that can never be entirely separated from the play of the game itself.(2012, p. 206) This was evidenced in that, alongside the competitive storytelling that Gloom encourages, there is a more empathetic side to the game.In Gloom, the player who has had the worst day goes first; this is decided by everyone discussing their days and then collectively voting as to who has suffered from the most misfortune.By encouraging players to talk about their recent lived experiences, Gloom provides both an accessible platform and a social mechanism for the players to interact, and in order for this interaction to occur all that is required is a copy of the game, between two and five players, and a table and some chairs, making it both an affordable and accessible platform.
Second, we thought that a tabletop game would be easily changed and adapted by players, especially during the process of play.This allows more flexible relation to the rules than might occur in digital games.Accordingly, while the rules of an analogue game might initiate play (regulated, competitive play), they also initiate the possibility of playing with (rather than within) the rules in a manner that is perhaps less straightforward in digital games.Juul's (2005) description of digital games illustrates this well, suggesting that 'rules are designed to be objective, obligatory, unambiguous, and generally above discussion' (p.121).Players of video games can and will play in ways that are contrary to the expectations of a game's designers, but this is the exception rather than the norm and without access to the game's underlying code even these forms of 'subversive' play usually remain within the parameters of the game as designed.Again this was evidenced by Gloom where, as shown in Figure 1, players were easily able to adapt and modify the game to suit their playing experiences.Had we instead elected to play a digital game with its own set of rules and gaming mechanics, such adaptation would have been much more involved, and would likely not have been possible during the time frame of the session.Furthermore, it would likely have isolated those participants who had the requisite skills to make such adaptations from those that did not, thereby further reducing the availability of a truly shared social space for experience to take place.
Third, we hypothesised that tabletop games provide a tactile experience, one which can be experienced with all of the senses.As Woods (2012) remarks, '[t]he elements of both graphical presentation and high-quality components contribute towards what Hunicke et al. term the "game as sense-pleasure"' (p.160).Wilson (2008) contrasts this aspect of tabletop games with digital games directly, noting the three-dimensional, tactile element as something which tabletop games have over digital games in terms of providing an interactive experience to players, although he acknowledges that tabletop play can be slower and less immediately gratifying.Gloom's high-quality transparent plastic cards are central to the experience of playing.As Sullivan and Salter (2017, p. 17) note, '[i]n the case of the Gloom series, the transparency of the cards adds to the play of the game, and is not merely a gimmick'.Touching the cards, and experiencing the unique stacking mechanism in which in-game information is covered and uncovered, is a key part of the game, introducing an element of play that has to be physically experienced to be appreciated.While video games, and most obviously VR, might offer an 'expansion of our physical and sensory powers' (Ryan 1999, p. 110), tabletop games enable us to experience our current physical and sensory powers in the moment, presenting us with an opportunity to fully engage our senses during play.
Beyond our observation of the play session, we were keen to hear from the players themselves on some of these issues.The participants spent the majority of the plenary discussion describing their experiences of the game, and of the narratives that they constructed.While they also reported who 'won' the games from a points-based perspective, this constituted only a brief part of the discussion, thereby supporting Baker's observation that you can 'win' Gloom even if you lose (numerically).Figure 1 shows one group of participants playing Gloom, having first re-arranged their cards so that they were facing the other members of their group.Traditionally, Gloom is played with the cards facing each of the players, but by turning them to face each other, the group made the collective decision that a collaborative approach to storytelling (better facilitated by the new arrangement of cards) was preferable to a purely points-based victory.This is perhaps what De Koven would classify as movement towards a well-played game, 'changing the game' (De Koven 2013, p. 47) so that we can continue to play well together.An alternative reading of this player activity is that the competitive nature of the game was effective in initiating playful behaviour, signalling a possibly productive merging of gaming and play as defined by Caillois.
Following an initial discussion of the relationship between ludus and paidia, and the possibly similar division that has been made in game studies between games and stories (discussed during the session in terms of structure and narrative), the participants were asked where they would place the game on a two-dimensional framework, describing the relation between 'story' and 'structure' (on the x-axis) and 'play' (on the y-axis).
For the purposes of this exercise, the framework asked the participants to consider the story and the structure of the game as diametrically opposed constructs, and to then consider the extent to which their experience had been playful.Following a 30-min discussion, consensus in the room seemed to be that Gloom was marginally more concerned with story than structure, and that it was playful in its nature.Its position in the framework can be seen in Figure 2. Following the placement of Gloom, the participants were asked to place their favourite tabletop games on the same framework; these had been written down earlier in the session and used as an initial ice-breaking exercise for the participants to discuss their interest in tabletop games.
Figure 2 shows the final position of the favourite games of the participants using the proposed framework, and while the spread of post-it notes seemingly suggested a preference for games that are more reliant on structure (rules) rather than story, it also displayed a seemingly random assignment of 'play'.
The discussion that followed this exercise also revealed that despite the participants consisting mainly of academics who research play, the term 'play' was itself 'at play'.Rather than arriving at a single definition or answering our initial questions definitively, participants discussed a number of aspects of play that were more or less present in their own definitions; indeed, some of these aspects were mutually exclusive.
There was further discussion around whether it is more playful to follow the rules of the game or to try to transgress these in order to improve or extend the experience of the game.This is, essentially, a discussion about the connection between freedom and play, and whether play can or should be constrained by rules.Most recently, two sides of this debate have been exemplified by Sicart (2014), who emphasises the transgressive, self-directive aspects of play, and Bogost (2016), who argues that play occurs within constraints, and that it is these constraints that make play what it is.This debate is at the heart of A Well-Played Game.For De Koven, rules can eventually be changed and abandoned as we learn to play better.In the workshop there was no real conclusion to this discussion; the room remained fairly evenly divided between those who preferred playing by the rules and those who tried to get around them, indicating that this is perhaps a fairly fundamental division between players.

(page number not for citation purpose)
This topic fed into the debate about whether a competitive or collaborative approach to the game is more playful.Caillois' agôn (competition) and mimicry ( imitation) were raised as contrasting approaches, whether by playing the game our participants were competing to win or cooperating on a collaborative storytelling task.De Koven's contribution to this is to introduce the idea of 'coliberation' (2013, p. 148), the idea that, regardless of whether they are cooperating or competing with other players, players strive to make the game they are playing better by engaging with and challenging each other.
The idea of play as 'fun' was a particularly controversial topic of discussion, with some questioning whether fun is really the goal of play.De Koven certainly places an emphasis on fun, in particular 'having fun together ' (2013, p. 15).For others, fun is relegated to being connected to child's play; Caillois refers to ilinx or 'vertigo' and links it to activities which 'destabilis[e] perception' such as children whirling (2001, p. 23).Other terms, including engagement and 'flow' (Csikszentmihalyi 1990), were raised by members of the group as alternatives to 'fun'.
Ultimately, these discussions reiterated Sutton-Smith's (1997) suggestion that the diversity of play experiences, players, and play scholars, means that there is ambiguity around the concept of play.Much of the rhetoric around play which Sutton-Smith identifies was evoked by participants' comments, in particular the rhetoric of play as 'power', 'imaginary' and 'frivolity'.More recently, Juul (2016) has taken a similar approach to the topic of playing games specifically, identifying literature which treats game-playing as 'submission', 'constrained freedom', 'subversion' and 'creation'.Our discussion with participants echoed many of these ideas, in particular the tension between playing by the rules and bending the rules that is outlined above.
Juul (2016) suggests a normative aspect to these distinctions; a focus on a particular aspect of game-play in theory will inevitably lead to different game experiences in practice.Thus, only by encompassing as many conceptions of freedom as possible will designers be able to enable 'games and players [to] fit together and mutually constitute each other ' (2016, p. 358).This returns to the idea that a straightforward definitional approach to play may in fact limit our abilities to accurately describe the nuances of play as experience, involving players, game and context.

Post-session evaluation
By asking our participants to play Gloom, we aimed to treat tabletop games as a technology through which to explore play as experience; so it is important to identify what the session demonstrated about tabletop games in particular.In parallel with this assessment of tabletop games, we can consider what this might also say about digital games, especially with relation to their role in educational settings.The game provided an accessible shared social space for the experience to take place; a format that was easily changed and adapted by the players; and an embodied and interactive experience that encourages us to consider the experience of game-play and how this links to educational settings.
There were, of course, limitations to our approach which must be acknowledged.Firstly, this case study represents only one group of people, all of whom had an academic interest and expertise in play, and who had decided to attend a session that explicitly mentioned the use of tabletop games in its description.If further evidence of the self-selecting nature of the participants was needed, then the favourite games that were selected (see Figure 2) point to an audience who were familiar with a wide variety of tabletop games including what might be termed 'hobby games' rather than the better-known mainstream titles such as chess and Monopoly.While this self-selection was actually beneficial in terms of addressing their thoughts on play, it is also true that the session would likely have been different if run with an audience who were largely inexperienced in playing tabletop games.
In addition to the audience in our sample group, it is also the choice of game itself that needs to be considered.It would also be reasonable to ask whether the same group of people would have experienced play in the same manner had a game other than Gloom been chosen.For example, what would have happened if a more complex and longer game had been chosen, one which had a clearly competitive element and was more concerned with game-play mechanics than story?The degree to which a tabletop game can be considered social, modifiable and tactile is dependent not only on the design of the game but also on the people playing the game and the environment in which it is played (Garfield 2000;Woods 2012).In using tabletop games to address the idea of play as an experience, it is therefore vital to account for all of these variables, with future studies presenting the opportunity to further test the hypothesis that as with digital games (see Squire 2011), it is not just the game itself but rather the participants and the environment that need to be considered.In this way, play needs to be considered a situated activity; its meaning can depend to a large extent on things outside of the game being played.
Indeed, this is perhaps the key point from the session which needs to be considered when examining digital games in education, as even these are similarly situated activities which will not have the same outcomes in different settings.Even though they may offer different affordances in terms of social interaction, modifiability and tactility than tabletop games, learners will still try to engage with these aspects to a greater or lesser extent depending on their prior experience of games.Some learners will follow the rules of the game in the same way that they follow the rules of the class; others will find ways to collaborate or compete with each other, make changes to the game or will focus on aspects of the physical experience of playing that may not have been considered by the educators who designed it.

Conclusions
The Playful Learning session suggested that tabletop games should be considered as a complementary technology to digital games in terms of exploring the idea of play as experience.Based on the session it is worth: (1) revisiting play as outlined by De Koven and examining further the implications of his concept of playfulness to better understand the implications about this session's outcomes and (2) revisiting the link between tabletop and digital games in light of this session, to understand how we can utilise this conception of play in our use of technology in the classroom.De Koven's concept of play is predicated on the idea that play, as a purposeless act, is the means through which we can build community and move closer to living better lives.He ultimately moves away from the idea of playing games and towards a purer idea of play beyond games, play as mastery over nothing in particular (De Koven 2013).For De Koven, games are at best a means to an end, a way to encourage an initial sense of playfulness; at worst, they are a controlling aspect over play, something which can corrupt play by enabling cheating, competition and overreliance on reified rules.It seems that 'true' play occurs when a game's rules are changed to the point that the game itself is no longer required.This vision of play could perhaps be seen as a utopian ideal.It is clear from our experiences of the Playful Learning session and of playing games as adults that, though some adults are more playful in their approach to life than others, many people need encouragement to play a game (and to play with a game once they are playing it).Play does not just happen, regardless of whether it should, and games provide an impetus for play to occur.We agree with De Koven that, once established, play is a free (and freeing) activity, but again this leads to issues; play means different things to different people, and even within a group of players there may be very different conceptions of what it means to play.De Koven acknowledges that it can be frustrating to deal with players who are more focused on the rules than you, or those who want to abandon the rules and indulge in flights of fancy.Part of the process of playing, then, is negotiating what play even means; games at least provide a structure around which this negotiation can take place.The implication of this for education is that even the most passionate 'gamers' will need persuading to work in a different context and with different people; indeed, this thought can be extended for educators using games in their classes, who need to consider not just the utopian ideal of a class that plays but also the potential reality of one that will not.
An alternative way of examining this issue is suggested by acknowledging the difference in the way that individuals experience play and games.McCarthy and Wright (2004) criticise instrumental 'user-experience'-focused approaches to technology, and instead offer an understanding of technology 'as simultaneously prosaic and aesthetic experience, contributing to the liveliness of our experience' (p.188).What may have been previously understood as a relatively simple interaction between a tool and its user is instead understood with reference to 'openness', the idea that our experiences with technology may be emotional, intellectual and sensual as well as practical.Using these terms, it could perhaps be argued that experiencing a game purely through its rules might be understood as a 'prosaic' experience, and that experiencing it through its narrative as an 'aesthetic' experience.In practice, any experience of a game will combine these two aspects (and more); play effectively becomes the complex way in which individuals and groups combine these modes of experience.
Examining the act of playing tabletop games from this perspective, it becomes apparent that De Koven may be too dismissive of the game as the technology being experienced and as the context in which people play.Play, when treated as a mode of experiencing games, allows us to understand that 'playing by the rules' and 'playing playfully' are not mutually exclusive ways of playing but merely part of the process of experiencing the game in different ways.A more playful mode of experience requires something to be played with, even if this object becomes less of a focal point for the experience.When play is seen in this way, it seems almost inevitable that it will occur within the context of games, but what actually occurs and is experienced becomes more difficult to describe because of the diversity of experience.Frustratingly, this communal aspect of play, which makes play such a fascinating topic for study, and such a potentially powerful tool for learning, is also what makes it so difficult to categorise and understand.The only 'given' is that players will (by definition) play; beyond this, what they do within (and with) the space of story and rules is unpredictable.Ultimately, it needs to be acknowledged that this unpredictability implies that games cannot be seen as a guaranteed solution to educational problems such as engagement, although they can be incredibly powerful when players make them work.Instead, when trying to understand the role of digital and tabletop games in education, we should consider the different ways in which the experience of play may affect the social interactions that take place in the classroom.While the Playful Learning session allowed us to observe, and discuss, the affordances of tabletop play in an educational environment, the outcomes of the session were also useful in formulating questions that might be asked of digital technologies.Two areas in particular are helpful in developing our understanding and use of digital technology: first, the 'social' aspect afforded by the material and ergonomic properties of tabletop games, and second, the relationship that pertains between players and the games they play when players are free, and able, to adapt the technology.
The social aspect of tabletop games, in that they often require player interaction (be it collaborative or competitive) was certainly helpful in generating conversation and debate.Here, the relative 'weakness' of tabletop games (that they lack the obvious sensory immersion, which is perhaps the appeal of their digital counterparts) might be perceived as a strength.Of course, this doesn't alter the fact that many learners will be familiar with, and attracted to, digital environments, but it seems reasonable to suggest that when players are tasked with 'running' the games that they play, the group dynamics and responsibilities are likely to shift in interesting, and productive, ways.
The ability to modify tabletop games rapidly to suit either pedagogic aims or the needs and interests of players has an immediate practical value in the classroom in terms of resource (both financial and in terms of time [Whitton and Moseley 2012]).Putting these material concerns aside, a related benefit derived from the materiality of tabletop games lies in the possibility of empowering learners to adapt their own environments in a way that is perhaps less straightforward when digital technologies are employed.The challenge in achieving this with digital technology is, of course, not insurmountable, and clear benefits accrue if learners can be equipped with the skills with which to design and adapt their own digital resources.
To conclude, we have argued that by considering tabletop games as a form of technology, we can expand our understanding of the role of games, play and playfulness in education and revivify the discussion of digital technologies in these contexts.At the heart of this argument is the idea that players experience play differently depending on the extent to which they engage with particular aspects of games, which were certainly present in Gloom, but which feature in all games to a greater or lesser extent.Innovative educators, then, should move beyond measuring the 'effectiveness' of technology in the classroom and towards examining the ways in which different play experiences contribute to (and potentially detract from) learning.This involves understanding the ways in which games enable social interaction, physical manipulation and play outside of the rules, regardless of whether they are tabletop or digital.We hope that this serves as an inspiration to future studies, and that tabletop games can be used by other scholars to further develop understanding of play and playfulness in education.

Figure 1 .
Figure1.Keith Baker's Gloom, played with each set of cards facing the other players, which was suggested for enabling a more playful experience.Note also the circular placement of the cards, appearing to create a large, combined family.

Figure 2 .
Figure2.The placement of the favourite tabletop games of the participants.As can be seen, Gloom (indicated by the circle) is seen as being marginally more concentrated on story rather than structure, and as being relatively playful.

•
What impact do high levels of regulation in games (both digital and tabletop) have on playfulness?• To what extent is participating in a competitive game, or a game which has competitive elements, playing?• To what extent does the experience of playing tabletop games align with the dominant definition of play proposed by De Koven?