Perhaps the oldest definition provided for what constitutes a game comes from Roger Caillois (1958), who was pursuing an essentially anthropological investigation into cross-cultural play in his book Les Jeux et Les Hommes (pictured left). It is worth noting that Caillois was French and thus used the term 'jeu', which means both 'play' and 'game' – this gave him a very wide perspective on what could be entailed by the term, and as a result his definition is quite inclusive. In brief, Caillois states that everything that constitutes play is free (non-obligatory), separated by limits specified in advance, uncertain, unproductive, governed by rules, and entails make-believe. The last two are the previously identified foundation conditions and can be set aside for now, and the separation condition can be understood as refinement of the rules-fiction dualism. This leaves three value judgements: that play (and hence games) must occur freely, without obligation; that play entails uncertainty; that play must be non-productive.

Unfortunately for Caillois, it is no longer clear that the first and third apply. As Edward Castronova, Thomas Malaby and others have noted, massively multiplayer games severely blur the lines as to what kinds of play are 'productive'. Similarly, it is unclear in the case of a gold farmer who also plays World of Warcraft for fun why their evening gaming should qualify as play but their daytime gaming should not. The boundary between work and play assumed by Caillois has become impossible to sustain in practice. Nonetheless, it is clear that there is a voluntary aesthetic that can be identified, since it can be found in Caillois, Suits and McGonigal, not to mention Elliott Avedon and Brian Sutton-Smith ("we can define a game as an exercise of voluntary control systems…"). This aesthetic seems solely to entail that the players of a game are, as Caillois suggests, under no obligation to play. Since the same game could be played in both a voluntary and a mandatory context – imagine being forced at gunpoint to play Russian Roulette – it's difficult to see why this would serve as a workable boundary condition, though. It almost seems as if what is intended by this value judgement is to say that games cease to be fun when they are mandatory. It's not entirely clear that this is a correct proposition, but it is a perfectly viable aesthetic of play.

There is one element of Caillois' definition left: the requirement for uncertainty. It is possible that this might be foundational – are there any clear cases of games that don't entail uncertainty? Either way, we can tentatively accept an uncertainty aesthetic as a position entailed in Caillois' view, and implied by Costikyan's (and possibly Koster's). It is not necessarily the case that Caillois is making a strong case of the aesthetic merits of uncertainty so much as he is reporting his view that those who enjoy play are deriving some proportion of their enjoyment from the uncertainty of the activity. This sits oddly with some of the activities Caillois includes within his category: theatre, for instance, is given as a very formalised kind of playing. While I can support this on the basis of Walton's make-believe theory of representation, it still seems to be problematic: if a person enjoys a stage play greatly and returns to watch it again, is uncertainty really their interest? It is not entirely plausible to suggest that the audience are hoping for things to go wrong! Similar concerns relate to players who master a particular videogame sequence perfectly, and return to execute that sequence again and again. It's not a wholly satisfying explanation for this behaviour that their motivation involves the possibility that they might fail. This seems to lend support to the idea of there being some kind of viable uncertainty aesthetic in games.

The uncertainty aesthetic finds its strongest expression in the work of anthropologist Thomas Malaby, who in his 2007 paper "Beyond Play: A New Approach to Games", decries the assumption that games and play should be seen as continuous. Malaby is particularly hostile towards Caillois' suggestions that games must not be productive, and in general wants to disentangle the concept of 'play' from the concept of 'game', and similarly to separate 'play' from 'work'. Malaby's defines games as follows:

A game is a semibounded and socially legitimate domain of contrived contingency that generates interpretable outcomes.

"Semibounded" is intended to take into account concerns about the 'leaky' Magic Circle (which Castronova drew attention to) while the requirement for social legitimacy is a fascinating and unique condition that offers an anthropological interpretation of games rather than an aesthetic value judgement, as such (although it might be taken to support the social aesthetic that Costikyan drew attention to). Both are perspectives on the rules-aspect of games, but from a vantage point so distant that it is barely recognisable as such. It is the "contrived contingency" and "interpretable outcomes" that are the core of Malaby's definition, here, and these can be seen as the uncertainty aesthetic crossing over into the fiction-aspect of games (the 'interpretability'). In many respects, Malaby's definition is the broadest, but of course this breadth also lets in many things that people wouldn't normally consider games. This is the nature of any widely construed definition, however – the trade-off in defining 'game' is always between being too inclusive, or being too exclusive. There seems to be a tendency for exponents of the victory and conflict aesthetics to vehemently prefer exclusivity, whereas exponents of uncertainty aesthetics like Caillois and Malaby prefer inclusivity.

In a paper the following year entitled "Anthropology and Play: The Contours of Playful Experience", Malaby expands his definitions to take onboard a new concept on play – one based on seeing it as "an attitude characterized by a readiness to improvise in the face of an ever-changing world…" Although he does not reference Suits, Malaby here is aligning 'play' with what Suits called 'the lusory attitude', although since Suits sees games as expressly goal-oriented (victory, conflict or reward aesthetics) – his intention is narrower than Malaby's. Nonetheless, both Malaby and Suits are gesturing at the special mental state that players enter into when participating with the uncertainty that lies deeply in the nature of games. As Malaby recognises, this has roots even older than Caillois' study of play, since Johan Huizinga's Homo Ludens (1938) – which inspired Caillois – is based on this idea. As Malaby writes:

For [Huizinga], the play-element – marked by an interest in uncertainty and the challenge to perform that arises in competition, by the legitimacy of improvisation and innovation that the premise of indeterminate circumstances encourages – is opposed above all to utilitarianism and the drive for efficiency. (Caillois likewise, despite his misleading claim that games are occasions of "pure waste," recognizes the centrality of contingency in games.) Huizinga felt that the play-element had been on the wane in Western civilization since the eighteenth century, threatened by the drive for efficiency and the routinization of experience it brought.

There is an aesthetic value judgement here – explicit in Huzinga and Malaby, tacit in Suits and McGonigal – that values the playful experiences over and above the pursuit of victory. This is definitely present in Suits' , albeit somewhat concealed by his overt focus on goal-oriented play, and it is inescapable in Malaby who has no trace of the victory, conflict or reward aesthetics at all in his concept of a game. This playful aesthetic (for want of a better phrase) is intimately connected with the uncertainty aesthetic – indeed, it is a refinement of it, in perhaps a parallel to the way the conflict aesthetic often appears intimately connected to the victory aesthetic. It is Caillois' paidia, Suits' lusory attitude, and Malaby's view of play as a disposition, and it directly opposes any aesthetic that would seek to make games solely about winning, battle or puzzles.

Next week, the final part: An Island of Play