Ritualistic Games,Boundary Control,and Information Uncertainty

Liminality and Information

Starting from Van Gennep (1909), religious and magical rituals have been called liminal spaces. They are a step outside mundane reality, yet exist in continuity with it, and have the capacity to facilitate significant social and cognitive changes, such as life transitions and the transmission of secret knowledge. In liminal spaces, normal systems of logic or ethics may not apply, as they are replaced by temporary new ones pertaining to the ritual at hand. An example of this is tribal rites of passage, where the initiates may be seriously harmed by their elders, who would normally face severe penalties for such violence. To enter the sacred chamber is to step into a world of new rules. As Huizinga (1939) suggests, ritual and play seem to share a common origin and properties, being both part of the play element of culture. Through analysis of their information traits, they have been shown to be facets of the same phenomenon or at the very least to be using the same mechanics for influencing participant cognition (Harviainen & Lieberoth, in press; Lieberoth & Harviainen, in press). In liminal games, lusory attitude (the willingness to accept the limitations of game-play in order to make it more rewarding; Suits, 1978) takes the place of faith as one motivation to participate. This effectively means that many of the tools used to examine the ritual process also allow us to understand the way certain games work.

The special social contracts surrounding rituals and [pretense and/or pervasive games] establish meaning and relevance-criteria. This creates a barrier to the outside world, telling us that “what happens in the woods stays in the woods.” The formal delimination is supported by special acts and goals within the activity, which separate it from everyday experience. These factors, when combined, enable the emergence of cognitive frames that invite players to perceive a more malleable reality beyond everyday practice. Participants may be aware of real life as well, but make sense of the primary, liminal framework because of the invitation to share a fantasy. (Harviainen & Lieberoth, in press)

Not all games are liminal. A chess match may create significant engrossment, but does not take the participants to a new reality the way a live-action or massive online role-playing game does. Chess may come close nevertheless, especially at intense high-level games where the atmosphere is similar to a football match and the game arena itself becomes an extended space with its own rules. Games that are liminal, or at least approach liminality—being thus liminoid, as Turner (1982) puts it—can be analyzed as information systems. ² In practical terms, this shows the underlying patterns of information manipulation that enable player engrossment and the illusion of fictional reality. Addressing the players’ cognitive state itself is, however, beyond the scope of this article. The liminality of a game is built upon a combination of factors. The game environment becomes a resignification zone (as per Deleuze & Guattari, 1980), in which elements are given a new, in this case ludic, meaning. Within it is a recursive set of signs pointing to other signs, unto infinity. A definite point exists, however, where the chain of references is altered, and that is the key level of interest for game researchers. This is the level of socially agreed-upon rules. Liminality is always based on social contracts, open or implicit, that delimit the space of activity (for games, see Salen & Zimmerman, 2004). Sign meanings arise from minds and bodies first as personal signification spheres (Brier, 2008), but communicable sign systems are constructed by way of shared discourse, which requires a community to create and support it (Bankov, 2008). A temporary reality relies on such sign systems for its existence. Due to this, the contracts function as a set of (possibly nonexplicated) rules indicating what may and may not be done within the magic circle of play or ritual (as per Huizinga 1939; Salen & Zimmerman, 2004). This creates both certainty and uncertainty for the participants not in charge of the activity: They know they are taking part in a rite or a game, but do not know what to expect from it. The need for constant resignification and environmental observation strengthens the uncertainty. Liminal games, like rituals, prolong the anomalous state of knowledge (ASK; Belkin, Oddy, & Brooks, 1982), the situation where a person knows that he needs more information, but does not know exactly what kind of information. As uncertainty increases, the participants’ focus is more easily fixed on certain sources that offer clarity of purposes and needs. These sources include peer groups, cognitive authorities, and diegetic (within-game-reality) goals—that is, things that are inside the shared temporary reality. In addition, the possibility of accessing outside sources exists, but that is often considered a breach of game etiquette, possibly even cheating (Harviainen, 2006). The liminal space needs to stay relatively isolated or at least to project the illusion of being isolated (Harviainen & Lieberoth, in press). This is where boundary control, the social protection system for the shared experience, comes in.

Boundary Control

Similarities between play and rituals roughly centre around 1.) action context or sequence, 2.) exaggeration, repetition and other structural modifications, 3.) psychological or communicative motifs, rather than immediate practical ones, 4.) a special “zone” for action and understanding. We have shown that this is particularly important at the level of frame-preserving interaction and information effects. For instance, changed actions call attention to themselves as frame-relevant, and give participants an opportunity to take part in the information flow, even without full insight into meaning and efficacy. (Lieberoth & Harviainen, in press)

No information environment is a social vacuum and neither do participants enter it from one. This means that each player brings to the game both his preknowledge and any prevailing paradigms, which affect his information behavior (as per Ingwersen & Järvelin, 2005), personal signification systems included (as per Brier, 2008). Without that brought-with material, the game would not function (Crookall, Oxford, & Saunders, 1987). Such paradigms include experiences with other games, group memberships, and preplay encounters with game-related material produced by others. For example, checking out the official forum of a massive multiplayer online game, and finding out that the community of people using the forum is heavily against sharing quest spoilers, is very likely to affect a new player’s information seeking, at least as far as that game is concerned. At the same time, one does not want to look like a complete novice, as inexperience can be frowned upon. ³ Thus, one normal way of information seeking, that of asking directly, is socially prohibited by the community to a great extent. To solve this problem, players resort to a technique called berrypicking (Bates, 1989): They pick up bits and pieces from various sources, such as conversations, forums, and advice pages, instead of doing more holistic searches or readings on just one source. Unlike in similar practices, such as cherrypicking (the picking of what is seen as the best parts offered), berrypicking typically concentrates on appropriating that which is most conveniently available, which may or may not be the most informative, reliable, or accurate parts. This technique is, when necessary (in order to avoid serious game breaks, for instance), supplemented with direct questions. By the use of berrypicking, disturbances to the magic circle of play are kept to a minimum. This process is a part of what is called boundary control in the study of religion. Boundary control also exists in many other social contexts, and the term originates from systems theory. Yet it is particularly apparent in charismatic religions, due to their strong need to separate “us” from “them” (Galanter, 1999). In such cases, it refers normally to the way by which cults maintain their crucial belief in the difference between members and nonbelievers, but the same system also drives secular organizations to function as a whole—and creates engrossment at liminal play. Boundary control means that an imposed sense of new reality—a fictional one or a reinterpreted mundane one—is consistently sustained by the participants, for all the participants. Should it fail, for one or all, belief also fails. For boundary control to be successful, a liminal activity’s magic circle has to prevent unwelcome information from crossing it, either by blocking said information or by altering it to an acceptable form. The magic circle has to be nonporous (Harviainen & Lieberoth, in press). This is what sets liminal pretense games apart from their siblings such as bibliodrama, sadomasochist role-playing, and postmodern magic, which need a porous magic circle, because they have goals that exist outside of play and are thus allopoietic in nature (Harviainen, 2011a). Guided by authorities (e.g., priests, tribal elders, game masters) and sufficiently agreed-upon rules (dogma or the social contract of play, as per Harviainen & Lieberoth, in press), the participants preserve and sometimes also adjust the magic circle. Three key factors make up a group’s boundary control: The aforementioned, shared social need not to disturb the illusion (Letcher, 2001), shared conceptions of what objects and information elements are situationally relevant (P. Wilson, 1973), and a heightened reliance on cognitive authorities—persons or groups defined as holding significant expertise on the matter at hand, regardless of whether they actually have true expertise concerning it (Galanter, 1999; P. Wilson, 1983).This is extremely significant for game-play and not just religion. Successful boundary control not only enhances engrossment, but is also fundamental for what McGonigal (2003) calls the Pinocchio Effect, “the desire for a game to be transformed into real life, or conversely, for everyday life to be transformed into a ‘real little game.’” Should the control fail, immersion is disturbed, possibly to the point of an Emperor’s New Clothes effect—a situation where the participants’ sense of the fictional reality stops and they also realize no one else is taking the game seriously either (Harviainen & Lieberoth, in press). To avoid this, devoted players not only support the boundary control in general, but also personally work as blunters (as per Baker, 1996), rejecting information that could damage their own game experience. As game scale grows, boundary control becomes a layered process, due to differences in player preferences. For example, the online role-playing game The KINGDOM OF LOATHING (2005-) gives only very limited game-mechanics information on many effects within the game itself and offers only small clues for most in-game quests. The game’s official forum uses a policy of allowing the free distribution of almost all game information (including spoilers), as long as those are hidden within a (spoiler) background. In contrast to these, the game also has an official wiki, where all such elements can be freely displayed for all, and the publishers even allow the use of game interfaces, which give easy access to such information during play. It is thus up to each player to choose how much information he wants to discover by himself, whether to preserve the sense of liminality or not. The choice determines the player’s participation in social contracts regarding the game, as each option represents a different relationship to game challenge and policy of information access. This is obviously very different from a live-action role-playing game (LARP) based on secrets dealt to each player, where participants are only expected to ask for extra out-of-play information in cases of omissions in the material they were given, and where they may be physically distant from such information sources. An example of this is PRAYERS ON A PORCELAIN ALTAR (2007-2008), a LARP for 8 to 10 players specifically designed to function without an obvious preappointed cognitive authority of any kind. Its sole means of interaction is communication (physical and verbal), and it deals with applicants to a local theater academy who have a severe hangover and issues to solve (see Harviainen, 2011b, for details). The game master is present solely to clarify the provided material. The emergence of cognitive authorities, and the selection of who they are, takes place within the game’s fictional reality, from among characters that are initially on an equal footing, as they are given no leverage as related to available in-game information.

In all 22 documented runs of PRAYERS emerged a clear cognitive authority, which effectively regulated the revelation of in-game secrets. At every run, that authority was at all times contested and maintained by a projection of supposed supreme expertise, matching P. Wilson’s (1983) descriptions: a person considered by (many of) those present to have expertise within the given subject field, regardless of actual expertise or even competence regarding it. Typically, only those who are deemed to “know what they are talking about” become cognitive authorities. That attribution may sometimes be baseless, as people rely on things they already consider “known,” simply because their chosen “authority” confirms their own expectations. As the subject field of supposed expertise, in the case of games, is temporal, cognitive authority usually dissipates at the end of play, or soon after.

What matters for a liminal game to function is that the players at least on principle share an agreement to respect the other players’ interpretations of the magic circle. This may take place as either a part of following the rules of the game, or of a nonexplicated social contract regarding the game. Not to respect the magic circle is to be a cheater or a spoilsport (Suits, 1978). The differences in the systems, underlying the agreement—the information processes fueling successful play—are, however, that which is of particular interest for the study of ritual phenomena as information systems. Before examining the differences, however, a look at certain commonalities of the liminal process in general is necessary.

The Game as a Self-Referential System

In rituals and liminal games, such as role-playing games of any type, the magic circle of play functions as the border of a self-referential social system—to commit to playing will make that commitment rewarding, as the activity’s positive elements intensify one another (Harviainen & Lieberoth, in press). The participants construct and continue re-reconstructing the social system (collective network) involved. They refer to themselves and fellow players as members of that system, together being in charge of framing the boundaries of their social space (Klabbers, 2009). The system is not totally self-referential, but participants act as if it were, even when subtly accessing information sources beyond it (Harviainen, 2007). Present in such environments are factors that limit possible actions and access to information: isolation through the social contract, possible physical isolation (including online play during which access to other sites is limited), game rules, and the participants’ own attitudes. These limitations function as the basis that enables the transformation into the fictional reality.

Information is not simply added as new knowledge. It is appropriated so that it fits with the person’s expectations and emotional values. For a believer within a liminal environment, this creates a feedback loop where the available information supports the ritual—or game—experience. That, in turn, makes the participant more likely to get involved with the activity, appropriate information from within, and act as an information source for others in a manner that supports it. (Harviainen & Lieberoth, in press)

The fictional reality, being a resignification zone, undergoes constant redefining, most of which is subtle and takes place only inside the players’ minds. Based on information gained from cognitive authorities such as peer groups and persons seen as experts, berrypicked clues, and personal preknowledge, players constantly reread, reinterpret the game environment. This uncertainty is one of the key things that keep play interesting. Which parts are resignified depends heavily on player preferences. For example, a competitive player constantly reevaluates skills and gear, while explorer types keep looking for new content to discover (Bartle, 1996; Yee, 2005).Resignification, unless taken to levels where sign-interpretation conflicts between players are inevitable, is also one of the elements sustaining the play experience. Other factors contributing to this include a sense of shared experience and community (communitas; Turner, 1969), the Pinocchio Effect, and the players’ own levels of lusory attitude. The redefining and sustaining elements furthermore strengthen each other. This, in turn, makes it meaningful for the participants to accept the initial limitations—the hindrances, which constitute the activity within a game—in the first place (Harviainen & Lieberoth, in press). By taking part in the activity, they intensify the experience for themselves and the other participants. This leads to a stronger magic circle for all. It, in turn, again strengthens the intensity of the participant experience, as long as no serious breaks appear (minor breaks being inevitable). Successful boundary control keeps them from happening. The process is further assisted by the presence of cognitive authorities inside the magic circle, as their commitment guides that of the others.

Information and Authority

Many role-playing games, like rituals, also contain a special type of cognitive authority, a directive information source. Such sources are persons or groups, which are not only seen as experts, but also given the ability to overrule current representations in the game, even during play (Harviainen, 2007). They are authorities whose power extends from outside the magic to inside it, affecting it well beyond just setting systemic goals. For example, if the game master of a LARP says that a person is your character’s former lover, she really is, even if this contradicts other, earlier information. Or should the company in charge of an online game decide to implement changes to the game, they are by default allowed to do so. This power, like any other influence from cognitive authorities, can be contested, but given the authority’s position, to do so is to both introduce potential game-experience breaks and to risk having to leave the game. Contesting authority may also carry out-of-play repercussions, such as failing a course (in the case of learning simulation/games) or social ostracism.

Issues of scale, time, and power-distance between cognitive authority and participants (Hofstede, 2001) complicate the contact between directive sources and players. It is significantly more difficult to call into question a change implemented by Blizzard Entertainment to WORLD OF WARCRAFT (2004-) than it is to contest the ruling of the game master in a small LARP. On the other hand, the continuity of games such as WORLD OF WARCRAFT makes it possible for communities to develop, the purpose of which is to counter unwelcome changes to the game. ⁴ Tracking these ways of change resistance is a very effective tool for discovering which information elements are considered the essence of the play experience by each group. By following patterns of information seeking and use during play, one finds clear evidence of play preferences (for an example, see Harviainen, 2006). Sometimes it is also possible to discover new power structures: For example, Myers’s (2008) Twixt experiment of playing according to rules, but against player conventions shows that cognitive authority on what is seen as “proper play” can shift away from the company in charge of a game. The tricky part is not in the following of said patterns, but in understanding how to recognize such information behavior in each game type. Yet this approach offers something well beyond simple reframing of game studies concepts as information behavior. Through deconstructing the social system within the boundaries of the magic circle into its information-process components, it is possible to actually understand why certain games engage the mind the way they do. After all, engrossment in an arcade game and a LARP is very different, as the former lacks certain immersive elements (such as the illusion of isolation) that are fundamental to the latter, a liminal game type. From the viewpoint of tacit understanding among the players, however, no difference at all between these forms of play may necessarily exist. Individual play preferences have a significant influence on such tacit understanding (Harviainen, 2006).

Liminal spaces not just include certain (tacit) information processes, they are liminal because of those very processes. While they most certainly do not account for all of liminality, these information processes are nevertheless an integral prerequisite without which the necessary cognitive changes could not function (Lieberoth & Harviainen, in press). To track those processes, to understand their variance and its impact, is to come one step closer to understanding why games affect us the way they do. Tracking them, however, requires knowledge of the scale of the phenomenon being analyzed.

Frequency, Emotional Arousal, and Involvement

The cognitive study of rituals (i.e., the study of rituals from the perspective of the cognitive, neurological, and evolutionary sciences, a part of the cognitive science of religion; see McCauley & Lawson, 2002, for an example) speaks of two distinct types: the rare and powerful, such as a tribal initiation rite, and the frequent and repetitive, such as the Roman Catholic mass. Whitehouse (1996) calls the former imagistic and the latter dogmatic modus. McCauley and Lawson (2002), in turn, envision a scale where the former is considered low-frequency high-arousal, and the latter high-frequency low-arousal.

According to McCauley and Lawson (2002), these factors are variables: Frequency is calculated by adding up the instances of play, in either a “raw count” of actual runs or as a potential performance rate. Arousal, in turn, is analyzable through the level of sensory pageantry present, and in rituals it is usually inversely proportional to performance frequency. It is the sensation of tension, of being at the cusp of something essentially unfathomable, a neurophysiological state that functions as the mandatory prerequisite for emotional engagement and immersion in rituals and ritualistic games. Without sufficient arousal, participation in ritualistic activities is limited to social reasons. Both emotional arousal and social participation are, in turn, parts of the wider category of “involvement,” which includes all facets of attending an event the participant cares deeply about. People attend games and rituals for various reasons, but sufficient emotional arousal is required for fully enjoying the activity in question. Without it, certain key parts are not experienced.

Arousal is also accessible to some level via psychophysiological research methods (Kivikangas et al., 2010): Using currently existing technology (facial electromyography, electrodermal and cardiac activity measurement, etc.), its presence, if not precise intensity, can be noted. This does, however, limit several factors related to the boundaries of the magic circle, such as sufficient freedom of movement, meaning that the sense of liminality may be lost in such measured play. Furthermore, using such methods to study LARPs or similar phenomena is, with currently existing tools, next to impossible.

In discussing games on ritualistic terms, both variables are necessary (Harviainen & Lieberoth, in press). While rituals usually fall quite clearly in one of the two categories, liminal games contain various levels of frequency and emotional involvement. One end of the frequency scale are games of the “one-session game of narration” type (OSGON; Frasca, 2000), which are run once and where character death means end of game participation. One-shot LARPs exemplify that type, as do single-run training simulations. At the other end are continuous online role-playing games such as THE KINGDOM OF LOATHING and WORLD OF WARCRAFT, where events and content constantly evolve, regardless of whether a player is present or not. Between these extremes can be found games such as continual tabletop role-playing campaigns (a visible gaming counterpart to dogmatic rituals) and simulations/games that run through multiple sessions. Likewise, emotional arousal varies. At one end, OSGON-LARPs such as PROSOPOPEIA BARDO 2: MOMENTUM (2006) actually seek to emulate elements from imagistic rites (see Stenros, Montola, Waern, & Jonsson, 2007, and Stenros, 2008, for examples), and designers of massive multiplayer online games seek to produce as much arousal as possible. This is not easy, as rites of terror affect people much more strongly than do pleasant elements (Whitehouse, 1996). Role-based learning simulations, in contrasts, may seek low arousal, so as not to disturb the educational content (Henriksen, 2008). Arousal, as a ritual variable, is not the same as engagement with the game, but the emotional sense of “being on the edge.” This is why a low-arousal game can still be highly challenging, addictive, or both. If arousal rises too high, the memory-imprint mechanisms of the activity change (Whitehouse, 1995). An optimal learning game would thus seek maximal engagement (Klabbers, 2009), but only moderate emotional arousal. The higher a liminal game’s frequency, the higher the risk that boundary leakage takes place. Given time and opportunities, some players will end up using sources outside of the play-space. At the same time, however, damage from such leakages decreases because a frequently or continually run game can mitigate any problems later on—something impossible to OSGONs. Increases in arousal, in turn, intensify the game experience. High arousal makes players more likely to skip errors and omissions in the game material, as they are highly involved in the flow of playing the game (Csíkszentmihályi, 1990). It too carries a risk: A problem which does break the continuity experience will be seen as way more significant, if it happens during high-arousal play. Yet the true attraction of high arousal for game designers is simply in the fact that a positive arousal is extremely engaging and often even necessary to create immersive play (as per Douglas & Hargadon, 2001).

Handling the Player’s Anomalous State of Knowledge (ASK)

Information-wise, imagistic high-arousal rituals work by extending the ASK: Put in simple terms, they rely on keeping the participants on their proverbial toes at all times, in need of new information relating to the situation at hand, in order to make sense of it, but without enough clues as to what kind of information that should be. Essentially, they do not know how to proceed in order to “do right.” This also works for liminal games: As long as a player does not have a clear task at hand, or all of the details of a quest available beforehand, engaging surprises can be encountered. This, however, requires that the uncertainty, too, is made meaningful. Game designers utilize this fact for various ends, as do tribal elders.

By creating distressing situations, restricting action opportunities, overloading working memory, and dealing with matters of great cultural importance, imagistic rituals undermine calm, reflected information seeking. In stressful situations information appropriation becomes more erratic, as novelty becomes distressing, the scope of cognition narrows to few working-memory items, and secondary process thoughts are usually eliminated from consciousness. Reliance on immediate sources, supported by emotional salience, including feelings of trust and belonging, increases. Such reliance even extends to shared symbols and narratives, which are usually already present in the foreground of rituals and games. (Harviainen & Lieberoth, in press)

The trick for successful ASK extension is in successful boundary control: Liminality is sustained by way of making sure the players do not have a ready schema and performative scripts at hand, and have to rely on in-game sources such as cognitive authorities. This way when they encounter new material, even insignificant material, it will be engaging and possibly immersive (Balzer, 2010; Douglas & Hargadon, 2001). ASK cannot be extended indefinitely, and will at some point be replaced by a situational script. At that point, play habits change, and this will lead some players to abandon the game, while others will concentrate on other facets of it (such as grinding for gear) or seek to reacquire their ASK state. It also bears noting that some players (see Bartle, 1996, and Harviainen, 2006, for examples) are not interested in the liminal aspect of these games at all (even if they do accept their reference systems and resignification) and will because of this not enter the extended state of ASK at any point. ASK manipulation can be measured to a certain extent, by analyzing boundary control measures. The key is in how much is kept as a secret, and when: For example, PROSOPOPEIA BARDO 2: MOMENTUM aimed for a level where the players had no idea what or who exactly was part of the game. This is an obvious extension of ASK, seeking to bring the game as close as possible to an imagistic rite. ⁵ Likewise, a successful marketing campaign for a new expansion of WORLD OF WARCRAFT will simultaneously generate both very clear information needs (leading to active searches for what it may include) as well as an extended ASK state, in the form of a desire to see the new content oneself.

Anomalous State of Knowledge (ASK) as a Distraction

At the same time as some liminal games seek to create imagistic experiences, a distinct group of designers is going the opposite way. In order to facilitate game-based learning of many kinds, they replace ASK manipulation with a clearly suggested certainty of needs, in accordance with Kuhlthau (2004). Learning simulations, which have a long history of utilizing role-play in a manner similar to traditional role-playing games, are evolving away from being based on a gaming approach (see Henriksen, 2009, for a thoroughly analyzed example). This is, from the ritual-theory perspective, a move toward the dogmatic modus. ASK cannot, however, be totally avoided in even highly controlled knowledge domains. It can only be minimized as much as possible (Belkin et al., 1982). Lainema (2008), for instance, recommends that the play environment of business simulations be transparent and the influence of the organizer be lessened, both of which are definite examples of working toward reduced ASK while retaining all of the potential liminality of such play. The same principle applies to educational role-play, as such changes facilitate a smoother recognition of information needs, also those not directly related to the goals of the game.

Schick (2008), however, well notes that such moves, if too drastic, may result in players breaking the game frame, that is, dropping out of character and ruining the illusion of liminality. Given that, as Balzer (2011) observes, the illusion and immersive play appear to be necessary prerequisites for a role-playing scenario’s didactic use; such breaks can mean that next to nothing, from the perspective of the intended outside goal, is learned. An excellent example of this can be found from Laakso (2004), who studied trainee UN military observers partaking in simulations with role-play in them: One interviewee clearly states that as soon as he realized the organizers had designated a correct solution to each scenario, he dropped all attempts at role-playing and situational immersion, and instead concentrated on just finding the option predefined as the right one. This, naturally, stands in contradiction to everything a training simulation or game is supposed to accomplish on the educational level. Hyltoft (2010), in his analysis of a Danish boarding school, which teaches solely by way of role-playing, suggests that a balance between learning and entertaining can be maintained, as long as the fun part does not make the game any less challenging. This is important because while entertainment is only one incentive for play, the sense of not enjoying a game can turn the participant hostile toward the intended learning goals or make the game and its contents seem like a pointless waste of time (Lappi, 2011).

From this perspective, the division set by K. A. Wilson et al. (2009) between fantasy and representation as affecting learning appears to be faulty, in the case of liminal games. Such a scale does exist, but it is between levels of engrossment and immersion—players need to be sufficiently aroused and set apart from reality for immersion to emerge, but not so much as to ignore the intended learning content. What matters for this is not the level of representative content within the magic circle (except for personal preferences and the subject matter being taught, naturally), but rather the strength and maintenance of that circle, that is, boundary control. That, in turn, arises from the interplay between form, content, and context, which define meaningful play (Klabbers, 2009), and make respecting the imaginary magic circle something the players want to support.

In externally directed (allotelic) rigid-rule games, the game organizer is in primary charge of boundary control measures (Klabbers, 2009, p. 12). In more free-form games, such as LARPs, the players do most of the control, as a key part of their own reason for playing, the illusory reality, depends upon it (Harviainen & Lieberoth, in press). This is a difference that has to be observed when designing such artifacts, as they require completely different support systems.

Furthermore, the acquired information needs to be anchored, or it will be lost when taken out of the magic circle (Lieberoth & Harviainen, in press). Debriefings are a start, but they are not sufficient alone. Retaining skills directly related to one’s work is easy, but more holistic changes require a community that reinforces them. Such anchors—whether they are group meetings, documents, or other phenomena or objects—are again a measurable, visible part of the information environment.

For the design of liminal simulation/games for learning, these factors mean that it is necessary to look beyond didactic issues when implementing structural choices, such as deciding how complete an illusory reality the game will use. As Henriksen (2008) notes, the same thing that draws players to play the game—emotional arousal—may also adversely affect externally set learning outcomes, if taken too far. Likewise, some powerful game experiences function through ASK extension, a state apparently nonconductive to learning and information seeking, as it is by default a state where one does not even realize information needs, let alone solutions to them. ⁶ These problems increase as liminal simulations/games advance toward heightened realism, of the kind described by Klabbers (2000) as “mode III,” that is, “learning environments in which the learners are given the opportunity to interactively build their own system of resources and rules [, and which provide] conditions for the interactive self-reproduction of social systems” (p. 400). Simply reducing arousal and ASK is not the answer: As the firemen interviewed by Lloyd (2007) stated, if a training simulation is not sufficiently “real,” it is seen as just theory learning, no matter how well done it is.

A balance has to be maintained at all times between simulation/games being effective as experiences and as learning environments, especially if something beyond practical skills is being taught. The tool to achieving this is in the proper understanding of boundary control, including knowing when it is necessary to open those boundaries, or to make them porous, turning the activity into something other than liminal.

Further studying the topic is certainly necessary, as we are currently just extrapolating from case examples, and do not have a commonly shared theory for understanding game-based learning (Klabbers, 2003), let alone the impact of liminality upon it.

Discussion

One may ask, if liminal games are so close to rituals, seeming practically identical and using the same exact mechanisms, what (if anything), then, sets them apart? Lehrich (2004), for instance, claims role-playing is ritual, but “a distinct and specific kind of ritual, one with no exact equivalent in other ritual spheres.” In Goffman’s (1955) terms, all human interaction is ritual. So, following that paradigm, liminal games indeed are without doubt rituals. From such a point of view, it does not matter that instead of a faith component, participants use a lusory attitude to make them meaningful, as long as the form and the social reasons for entering the magic circle are alike.

Analyzed from the perspective of cognitive studies of ritual, things do exist which set the phenomena somewhat apart. For instance, the cost structures (see Bulbulia, 2009, for an example of the framework) of liminal games and rituals are usually somewhat different: Both incur similar penalties for failure to participate properly (Lieberoth & Harviainen, in press), but games rarely contain any offerings to gods or their like as mandatory requirements. Then again, neither do all religious rituals.

Given that both activities have a socially cohesive function, a similar structure, and so on, one can only speak of differences based on tendencies such as “in general, liminal games tend to be less scripted in form than religious rituals.” In effect, ritualistic activities seem to exist on scales, on which in the middle are events that are obviously both game-like and “traditional rituals” (such as Mardi Gras, a black mass performed just for fun, or a wedding with many ludic elements) at the same time. The lines of departure, in turn, are defined by whatever factor is chosen as crucial, be it the amount of belief in that which cannot be seen (“acting as if it were real”), the costs involved, or the amount of scripting.

Furthermore, if we believe Suits (1978) on games being about accepting unnecessary limitations in order to heighten pleasure through challenge, we can, however, see a crucial difference: Rituals, regardless of their type, are about efficiency and about achieving direct results that are taken outside the magic circle, even if they appear to be a very convoluted way of doing so (McCauley & Lawson, 2002). Liminal games are usually not about efficiency. Liminal learning games, however, are. As absurd as it seems, they can be considered to be inefficient for the sake of efficiency—just like rituals.

More important than such typologies, to game designers and analysts at least, is nevertheless to find out which rituals resemble which ritualistic games, in terms of frequency, emotional arousal, and information uncertainty (in the form of extended ASK). Those are the connections most useful for actual design work—educational or recreational—as well as research and should be seen as such. The ritual approach has been implemented into design already, in games such as PROSOPOPEIA BARDO 2 (see also Ericsson, 2009, for more examples). They were black-box experiments, in that their creators knew that ritual forms do work, but not the mechanisms by which they work. Given that the experiments were by all accounts high successes (see Nordgren, 2010, for an example), the question which naturally follows is, “How much further can we go, if we do know why and how they work?”

Researching ritualistic game attributes and variables can be done by using the tools of other game-related research: Frequency may be calculated; arousal assessed by observing sensory pageantry and player engrossment, as well as measuring them against frequency; and ASK manipulation seen by both analyzing the game-artifact’s design specs and by observing player behavior during play. All of these can furthermore be found again in player feedback, should it be looked at with a keen eye. Players who have either strongly liked or disliked a game can often be noticed reporting such things on various game-related fora.

Conclusion

This approach to the study of liminal games offers advantages that other approaches do not provide, because they concentrate primarily on structures or player experiences, while ignoring their fundamental prerequisites, described above. It reveals underlying structures and methods of control within game environments, including contested power dialectics. It also shows the roots of several game phenomena usually researched through other means, such as immersion, the Pinocchio Effect—and liminality itself. At their base is a set of information processes, which together enable their emergence, a potential of cognitive changes. The illusion of new reality needs to be kept up, and it is the presence of these information processes that does the sustaining of the magic circle: isolation from hostile information, a shared sense of what is significant in the current situation, and heightened reliance on cognitive authorities. This knowledge makes it possible for the other research approaches to dig deeper than before. For example, it gives educational scientists some more influencing factors to consider, factors which may actually be completely independent from any didactic design choices they have made, such as the inclusion of certain tasks that are expected to appear logical within the game’s reality, yet actually serve external learning goals. Likewise, it provides game studies scholars with alternative ways of explaining player-reported game experiences. Furthermore, a study of the information environment of a liminal game gives access to understanding how the game manipulates uncertainty, in the form of extending or limiting the ASK state. Researching these things is not difficult: It can be done with the same observations, forum analyses, and player interviews, as in other simulation- or game-related research. What is important, however, is that the researcher pays attention to the elements described here, that they be listed among things worth noting. They are fundamental components of the way in which magic circles function and must be recognized as such.

For liminal learning games and simulations, this means that in both designing and analyzing them and their effects on participants, attention should be paid to their boundary control measures and the way the games extend or avoid information uncertainty. This will not only make said simulation/games more effective, but also more enjoyable to participants and more effective as knowledge anchors in the long run.