Pretending to be a believer: on understanding religious fictionalism as a role-playing game

Abstract According to religious fictionalism, a non-believer can participate in religious life by playing a game of make-believe. Considering how games of make-believe build on imagination and pretence, I argue that religious fictionalism requires the non-believing participant to engage in role-playing. Turning to the literature on role-playing games, I demonstrate how religious fictionalism conforms to a qualified definition of such games. I also explore the theoretical consequences of adopting the role-playing perspective, by considering its impact on two key issues concerning religious fictionalism.


Introduction
Religious fictionalism is a belief-less approach to religion, usually interpreted as being fully compatible with both atheism and naturalism. 1 It rests on two core assumptions: that religious language can be fruitfully treated as fiction, and that one can participate in religious life by playing a game of make-believe.
In recent years, proponents of fictionalism have done much to defend and elaborate their views (Le Poidevin (1996);Idem (2003); Idem (2016); Idem (2019); Eshleman (2005); Idem (2010); Deng (2015); Robson (2015)). 2 However, while much has been said about the first core assumption, there is an unfortunate tendency in the literature to leave the notion of a game of make-believe on the intuitive, pre-theoretical level. The oversight is easily explained if we consider that religious fictionalism is modelled on more abstract types of fictionalism, such as scientific fictionalism. 3 Arguably, scientific fictionalism does not require the scientist to play games of make-believe. However, the fact that other major forms of fictionalism lack this feature only highlights the need to give it proper philosophical attention.
I explore the notion of make-believe in terms of imagination and pretence, emphasizing the difference between literary fiction and the kind of participation-based fiction in which a religious fictionalist partakes. Advanced forms of participation-based fiction usually demand that the subject takes on a role to play, and religious fictionalism should not be conceived of as any different. I argue that the religious fictionalist is rationally required to role-play and her role-character must be at least partly distinct from her ordinary self.
The study of role-playing games is a growing academic field, and I make use of the literature on role-playing to further our understanding of religious fictionalism. I demonstrate how religious fictionalism satisfies a qualified definition of role-playing games, a result with heavy theoretical impact. By way of example, I revisit two key issues concerning the approach (the idea that fictionalists need independent reasons for action and the instrumentalist assumption that fictionalists can receive the benefits of religion) and suggest how they should be re-addressed given the current perspective.
Before continuing, some preliminary remarks. First, I rely heavily on the notion of 'realist religion' in my discussion of fictionalism. By realist religion I mean religion practiced in the usual, truth-normed manner. In realist religion, central religious propositions are considered to convey important truths about a transcendent reality (insofar as human language can do so), and religious life is based upon (among other things) belief in these important truths. 4 Second, it is not always clear in the literature on fictionalism how extensive the game of make-believe is supposed to be. Some authors seem to suggest only that the fictionalist should pretend when performing specific religious actions, such as partaking in religious ceremonies. Others seem to envision the fictionalist pretending also in interaction with members of her religious community, as well as in other religiously relevant contexts. 5 For clarity, let us call the former 'limited fictionalism' and the latter 'extended fictionalism'. I think we should at least prima facie prefer extended fictionalism since it allows for a more substantial religiosity (although in the final section, I argue the importance of not extending it too much). Unlike limited fictionalism, it makes possible religious interaction with other people and invoking religion in one's personal life. 6 In what follows my concern is with extended fictionalism (if not explicitly stated otherwise). This introduction is followed by 'Religious language as a fictional discourse', a brief presentation of how fictionalism treats language and truth. In the section 'The religious game of make-believe' I explore the notion of make-believe in terms of imagination and pretence, and in 'The fictionalist as a role-player' I argue that religious fictionalism requires role-playing. In 'Religious fictionalism and role-playing games' I show that religious fictionalism can be fruitfully interpreted as a role-playing game, and in the final section 'The impact of the role-playing perspective' I explore some consequences of adopting this perspective.

Religious language as a fictional discourse
When we seriously doubt the ability of a certain discourse to put us in touch with reality, we usually abandon it. However, some such discourses come with major benefits of which we do not wish to be deprived. If that is the case, reinterpreting the discourse as a useful fiction can be a way of legitimizing its continued usage without flouting the demands of reason. This is the main idea behind the kind of instrumentalist fictionalism to which religious fictionalism belongs (other prominent examples include scientific, moral, and mathematical fictionalism). The religious benefits its proponents wish to preserve are commonly identified as spiritual development and moral awareness (Le Poidevin (1996);Idem (2003);Eshleman (2005)), although the details vary between accounts. 7 The interpretation of religious language as fiction is commonly taken to imply that religious language is truth-apt but not truth-normed. That a fictional proposition is truth-apt means that it has a truth-value, but since it is not truth-normed the truth-value is not relevant for its use (Robson (2015), 384). For example, when we read J. K. Rowling's work it is irrelevant that propositions like 'Harry Potter is a wizard' are strictly speaking false.
However, there is a notion of truth which is important for fictionalism, and that is fictional truth, namely, what counts as true within the fiction. In a fictional work, some fictional truths are explicitly stated (e.g. 'Harry Potter is a wizard') while others are implicit and can be assumed from context (e.g. 'Harry Potter has an ordinary human body'). It should also be noted that many propositions will lack a truth-value in a fiction, because a fiction is almost always 'open' in some regard (if a story tells us nothing about the weather, it is neither fictionally true nor false that it is raining).
How should we understand fictional truth? Robin Le Poidevin (2003) has suggested that fictional truth is determined by the fictional author. The fictional author is the author implied in the work, not to be conflated with the real author or the fictional narrator. 8 The idea is that whatever the fictional author can be taken to know is true in the fiction. While well suited for literary fiction, this idea seems inadequate for fiction without a fixed narrative, like religious fictionalism. When, within the context of the game, I pray to God, it becomes fictionally true that God hears my prayer, and the idea of a fictional author seems unable to account for how actions in a game can generate new fictional truths.
A broader notion of fictional truth more suitable for present purposes has been presented by Kendall Walton. According to Walton, whose work on fiction is explicitly centred on the notion of make-believe, fictional truth is determined by what one is 'mandated' to imagine in a fictional context: 'To call a proposition fictional amounts to saying only that it is "true in some fictional world or other." . . . a fictional truth consists in there being a prescription or mandate in some context to imagine something' (Walton (1990), 35-39). Walton's complete account is extensive, so I will rest content with the basic notion of fictional truth suggested here. The idea that fictional truth depends on what is suitable or 'mandated' to imagine when immersed in a fictional world fits well with religious fictionalism.
In religious fictionalism, the prescription concerning what to imagine comes from realist religion. To play the religious game of make-believe, the fictionalist needs to treat as fictionally true what a believer in the same position could consider literally true. In most situations, there are many possible ways in which a believer might respond, giving the fictionalist a range of possibilities rather than a single option concerning what 'truths' to imagine. It is important to stress that the situation itself might be entirely fictional, and that we are considering how a hypothetical believer might respond.

The religious game of make-believe
To participate in religious activities, the fictionalist must play a game of make-believe: To engage in religious practice is, on this account, to engage in a game of makebelieve. We make-believe that there is a God, by reciting, in the context of the game, a statement of belief. We listen to what make-believedly are accounts of the activities of God and his people, and we pretend to worship and address prayers to that God. (Le Poidevin (1996), 119) In this section, I consider the imagination and pretence involved in religious games of make-believe, as well as the use of props.
That a game of make-believe must involve a substantial use of the imagination is hardly controversial. Few would contest that imagination is the proper attitude towards fictional content, or that fiction requires at least some use of the imagination (Walton (1990), 40-41). Ingrid Malm Lindberg distinguishes between sensory imagination (visualization), propositional imagination that p and experiential imaginationthat is, imagining 'what it is like for a specific person to undergo a particular experience' (Malm Lindberg (2021), 31). 9 Since religious fictionalism is ultimately a view of language, it seems plausible to suggest that propositional imagination should be considered its basic attitude. However, it also seems obvious that the game of make-believe involved depends heavily on experiential imagination, since it basically consists in imagining undergoing different experiences as a religious believer.
According to Walton, the relation between imagination and fiction mirrors the relation between belief and truth: 'The relation between fictionality and imagining parallels that between truth and belief. Imagining aims at the fictional as belief aims at the true. What is true is to be believed; what is fictional is to be imagined' (Walton (1990), 41). This fits well with our understanding of religious fictional truth as dependent on realist religion. Given the assumption that a fictionalist treats as fictionally true what a realist in the same position might treat as literally true, we can say that as a general rule the fictionalist is to imagine what a realist in the same position could believe.
While there are crucial differences in how imagination and belief function (for example, most epistemologists take belief to be involuntary, and imagination seems to be a voluntary state of mind), there are also interesting similarities. According to Neil Shinbahabu, imaginations relate to other imaginations in a way similar to how beliefs relate to other beliefs. Just as it is psychologically infeasible to believe in contradictions, it is also extremely hard to imagine the truth of mutually exclusive propositions by the same act of imagining. In Shinbahabu's view, it is almost as impossible to imagine simultaneously both p and ¬p as it is to believe simultaneously both p and ¬p. Even though the content of imagination is not truth-normed, imagination does not allow for outright contradiction (Shinbahabu (2016), 112). While not all philosophers would agree with Shinbahabu on this, I think it is safe to assume that it holds for acts of imagination with a rich phenomenological content, such as the experiential imagination involved in religious fictionalism.
Most fiction (literature, drama, representational art, etc.) is appreciated solely in the imagination while the subject remains passive as reader or part of an audience. Religious fictionalism is different. It is appreciated in an active fashion, by partaking in the creation of the fiction. It has more in common with improvisational theatre and children's games than it has with literature or representational art. Active fiction requires more than mere imagination, it requires us to pretend actively. Keeping things simple, I propose we follow Elizabeth Picciutoand Peter Carruthers ((2016), 317), and think of pretence as imagination plus action: Pretence (def): 'To pretend that p is to act as if p (without believing it) while imagining that p'. 10 Games of make-believe are often prop-oriented, where props are real objects used to generate fictional truths. For example, if we play a game where we fight each other with sticks we pretend to be swords, the sticks are props and start to generate fictional truths. If I drop my stick, I drop my sword. If you poke me with your stick, it becomes fictionally true that you have stabbed me. 11 While basic physical objects are the archetypical form of props, there are no restrictions on what can function as a prop in more advanced games. In the religious game, props include things like religious emotions and theological ideas. An idea can be a prop because an idea which is fictionally true in a game of make-believe will start generating new fictional truths. Consider the game 'the floor is lava', which is basically built around the idea after which it is named and the fictional truths generated by this ideamost importantly, that you die if you touch the floor. The idea 'God exists' works similarly in a theistic game of make-believe (for example, since it is fictionally true that God is omniscient, it becomes fictionally true that God always knows what you are doing, and you must act accordingly when playing the game). Likewise, an emotion can be a prop in a game where the player is supposed to respond emotionally to the fiction, and these emotions generate new fictional truths. 12 For example, in a religious game of make-believe, you might feel grateful to God within the fiction, and that emotion makes it fictionally true that you want to praise God.
Most games of make-believe superimpose a fictional layer of imaginings on reality. Neil Van Leeuwen refers to this as a two-map cognitive structure, consisting of one layer of belief and one of imaginings: In pretending a sofa is a spaceship, you might have the following cluster of factual beliefs. FB: the sofa is in the living room. FB: the sofa is made of fabric, wood, and cushioning. FB: the sofa is of normal dimensions, about 3 ′ × 3 ′ × 6 ′ . . . .You might then have the following propositional imaginings, which constitute the second map. PI: the sofa is a spaceship in outer space. PI: the spaceship is made of light, durable metal and has controls for flying. PI: the spaceship is about the size of a medium-sized airplane. (Van Leeuwen (2016), 290) Van Leeuwen presses the point that we are not only guided by imaginings when playing a game of make-believe, but also by factual knowledge about the real-world props we use. While games of make-believe do not have a fixed narrative, the props provide some boundaries for what can reasonably be imagined.
The game played by the religious fictionalist has at least two features setting it apart from ordinary prop-oriented games of make-believe. First, props do not need to be introduced as such to function within the game. In the game with the spaceship-sofa, we must introduce the prop by making everybody in the game understand that the sofa is a spaceship. When playing the religious game, we imagine the world to be as described by realist religion. Since religion is an all-encompassing take on the world, virtually everything we encounter while playing the religious game will function as a prop without the need for special introduction. For example, when a Christian fictionalist picks up a Bible, she does not need to introduce it as a prop by saying 'let us pretend that this book is holy scripture', and when in nature she does not need to say 'let us pretend that this is all God's work'.
The second difference is that a religious game adds make-believe religious meaning rather than make-believe properties to its props (like the PI-properties of Van Leeuwen's spaceship-sofa). Van Leeuwen's subject pretends that a sofa is a spaceship with spaceship qualities, but the fictionalist is certainly not supposed to pretend that things she encounters in the world have properties different from those they in fact have. She should not pretend that her bathroom is a cathedral with cathedral properties, only a real cathedral is a cathedral in the game. Instead, the religious game is characterized by an added layer of imagined religious meaning. For example, a fictionalist might superimpose 'I am praying, and God is listening' as an added layer of meaning to her belief 'I am on my knees, talking to myself in my head'.
These differences depend on the assumption that in a religious game of make-believe, realist religion sets the rules for how the props generate fictive truths, while in an ordinary game of make-believe the rules are set by the players themselves. Since religious fictional truth is determined by realist religion, the fictionalist cannot use the props to generate any fictional truths she desires. Her use must lie within the acceptable range of ways in which the object could give rise to religious beliefs. For example, a Christian fictionalist might use a crucifix in make-believe worship to generate fictional truths but she cannot pretend that the crucifix is a magic wand and generate fictional truths about spell-casting. In an ordinary game, a prop can change properties if all players agree upon it, but the fictionalist must follow the lead of realist religion.

The fictionalist as a role-player
In basic games of make-believe where the fiction is limited, it is fully possible to participate without pretending to be someone else. However, the more elaborate a fiction grows, the more pressing the need to identify who the player is in that fiction. Therefore, elaborate games of make-believe standardly contain elements of role-playing.
First, a clarification. According to Walter Coutu (1951) 'role-playing' can mean at least two different things. It can refer to having a role associated with certain functions in a certain settingat work you might assume the role of a teacher while at home you assume the role of a father. It can also refer to pretending to be another person, imagining what that person would think and feel, and trying to act accordingly. I am only interested in role-playing in the second sense (henceforth I use the term role-playing only in this latter sense, and it would presumably be more adequate to refer to the first, functional sense as 'role-taking'). I understand such role-playing as involving what Malm Lindberg refers to as 'experiential imagining', that is imagining being a qualitatively different person with different thoughts and feelings, and to use this imagined picture of another person as a basis for pretending to be that person in speech and action. 13 The fictionalist needs to role-play because she is not a believer. It is part of who she is that she lacks religious belief. Therefore, she cannot simply 'fictionalize' herself and be herself within the fiction. To be able to take on the role of the believer, she must also pretend to be a somewhat different person.
Walton argues that it is part of what it means to imagine p, to imagine that one believes p (Walton (1990), 214). It might therefore seem straightforward to suggest that it is part of playing a religious game of make-believe to imagine that one believes the relevant religious propositions. But if the fictionalist is herself an unbeliever, is it enough that she imagines believing the central religious propositions she in fact does not believe? For example, is it enough that an atheist Christian fictionalist starts imagining that she believes 'God exists'?
It seems to me that it is not. Remember that the entire world is included in the religious fiction, and that everything can function as a prop in a religious game of makebelieve. This includes all reasons the fictionalist has for being a non-believer. If the fictionalist pretends to believe p in a fiction which in all relevant aspects is like reality, she will quickly re-encounter the reasons she has for not believing p in real life. Consider a fictionalist who is a non-believer because she is convinced by the problem of evil. The suffering in the world will function as a prop in the game, and when she considers it, it will generate fictional truths of the kind 'there is much suffering in the world, so presumably an almighty, perfectly good God does not exist'. That is because the fictionalist is to imagine what a realist in the same position would be inclined to believe, and that God does not exist is what she herself concludes when treating the problem of evil as a realist. 14 If the only difference between who the fictionalist imagines to be, and who she in fact is, is that she imagines believing 'God exists' (and similar propositions), there does not seem to be anything preventing her from losing her fictive belief, ending up a fictional apostate.
How can we hinder the fictionalist from losing her fictive belief? There seem to be at least two possible ways forward. Either we can make changes to the fiction regarding the world, or we can construct a role-character which differs more from the fictionalist's true self.
The first way might seem tempting. Could we not simply change the fiction, exclude pressing reasons for non-belief ,and replace them with firm reasons for belief? Why not exclude all instances of horrific suffering for which no plausible theodicy can be found, 15 and add firm evidence that the cosmos was ordered by a benevolent being? The problem is that this is public evidence, and by changing it we end up with a fiction which differs from the real world in important aspects. Since the fiction is based on realist religion the fictionalist cannot imagine things which would make her fiction significantly different from the world experienced by the realist community. Untrue fictions about the public evidence for religious belief seems hard to reconcile with this demand.
Given these considerations, it seems more feasible to develop the role-character than to change the fictional world. As a constraint on the enterprise, I will assume that the fictionalist wants to minimize change, so that her role-character remains as similar to her real self as possible without becoming a fictional apostate. I will also assume that she does not wish to make a role-character which is religious in a way she herself would deem irrational. How could the fictionalist proceed to invent a role-character meeting these demands?
It might be tempting to suggest that we only need to adjust the role-character's credence in the relevant religious propositions. 16 However, doing so would make the rolecharacter seem irrational from the fictionalist's own point of view. The fictionalist believes that the relevant evidence does not support any higher level of credence than the one she actually has and increasing the credence while everything else remains unchanged would therefore, from the fictionalist's perspective, amount to 'believing' on insufficient evidence.
One way of dealing with this would be to keep the fiction open regarding the fictionalist's evidence. In such a case, we would say that the fictionalist has some evidence to support her religious belief but refrain from going into details on what this evidence comprises. This could probably be sufficient if the fictionalist is only engaged in what I in the introduction labelled 'limited fictionalism'. However, I doubt it is enough if we consider 'extended fictionalism'. If the fictionalist plays the game of make-believe in interaction with other people, she will presumably at some point find herself in a situation where she is asked to clarify her reasons for belief. Can the fictionalist handle such a situation without inventing made-up reasons for belief or making her character irrational?
While I do not want to reject completely the idea of leaving the character's reasons for believing open, there are other methods for creating a believing role-character that are more in line with how fictional characters are usually created. It is common procedure when creating a fictional character to let the character be defined by some pivotal happening in the character's past. Consider for example how Harry Potter is defined by Lord Voldemort's early attempt on his life. Furthermore, it is uncontroversial that a powerful religious experience can support religious belief on a personal level. Therefore, I suggest that the fictionalist should create her role-character by introducing a significant religious experience into her own backstory. 17 The experience should be strong enough to provide a sufficient ground for religious belief. That is, given the fictionalist's other reasons for and against belief, the experience should be powerful enough to shift the balance in favour of belief (if the experience had been real, the fictionalist would have been a believer). However, to minimize differences it should not be a completely life-changing experience and to avoid creating a fiction in conflict with other people's perception of reality, it must be a private experience.
While the experience's main function is to enable fictive belief, it must be integrated into the role-character to work properly within the fiction. The fictionalist needs to give much thought to what such an experience would have meant for her, and the impact it would have had on her. For example, if she imagines having experienced a powerful, benign force lifting her up in a time of personal crisis, she might conclude that such an experience would have made her a more humble and thankful person. The alternative version of herself which emerges from these counterfactual considerations is the rolecharacter the fictionalist is going to play when engaging with the religious fiction.
I do think it is easier for most people to imagine a different version of themselves that has been formed by a religious experience, than a version with altered credence due to some unspecified 'open fiction' reasons. However, not all might agree with me on this point, and it is not essential for present purposes that the role-character is created by my suggested method. If the fictionalist is able to create a role-character with the right credence by way of an open fiction (or any other method) without flouting the constraints for the process I have identified, I have no objection.

Religious fictionalism and role-playing games
If the fictionalist's game of make-believe must include a substantial amount of roleplaying, does that mean that we should understand religious fictionalism as a role-playing game? Of course, every game where you pretend to be another person is a role-playing game in a trivial sense. However, there is also a narrower sense of the term 'role-playing game', used to single out a certain type of games and game-like hobby-activities.

Well-known examples include the tabletop game Dungeons and Dragons and the computer game World of Warcraft.
In what follows, I argue that religious fictionalism can be fruitfully viewed as a role-playing game in the narrow sense, 18

by demonstrating how it can satisfy Michael Hitchens and Anders Drachen's (2009) qualified definition of such games.
There are many kinds of role-playing games. The most well-known genre is arguably the tabletop game, to which the archetypical Dungeons and Dragons belong. These games differ markedly from the role-playing involved in religious fictionalism. In tabletop games, participants play characters which are not based on themselves (rather, they tend to be highly fictional, like dwarven princes or space pilots) in settings vastly different from the ordinary world (typically from the science fiction or fantasy genre). The major difference, however, is that tabletop games take place solely in the imagination. Tabletop role-players create a shared fictional world in their minds. While they do pretend to be their role-characters in that fictional world, there is no game of make-believe in the traditional sense. Instead, the story is advanced by dialogue between the participants and the outcomes of actions are determined by dice-rolling according to probability-based sets of rules.
Other kinds of role-playing games bear a greater resemblance to religious fictionalism. This is especially true of 'live-action role-playing' (standardly shortened 'LARP'). While tabletop role-players only imagine being their role-characters, LARP-players dress up and act like their characters in large-scale enactments. Like fictionalists but unlike other role-players, participants in a LARP superimpose a fictional layer on reality (Montola (2009), 24). There is even a sub-genre of LARP known as 'pervasive role-playing' which is conducted in the real world rather than a set environment. In such games, everything can function as props in much the same manner as in religious fictionalism and in order to integrate the fiction with reality, the role-characters are sometimes fictionalized versions of the participants (Jonsson et al. (2007)). There is a great resemblance between the fictionalist's game of make-believe and a LARP. But is similarity all there is, or should we go one step further and consider religious fictionalism a proper role-playing game?
Criticizing earlier definitions for being overly inclusive, Hitchens and Drachen offer a qualified definition of role-playing games. It is their explicit aim to exclude role-playing activities which are not proper role-playing games. This makes their definition especially interesting for present purposes. The definition consists of six conditions: 1. Game World: A role-playing game is a game set in an imaginary world. Players are free to choose how to explore the game world, in terms of the path through the world they take, and may revisit areas previously explored. The amount of the game world potentially available for exploration is typically large. 2. Participants: The participants in the games are divided between players, who control individual characters, and game masters (who may be represented in software for digital examples) who control the remainder of the game world beyond the player characters. Players affect the evolution of the game world through the actions of their characters. 3. Characters: The characters controlled by players may be defined in quantitative and/or qualitative terms and are defined individuals in the game world, not identified only as roles or functions. These characters can potentially develop, for example in terms skills, abilities or personality, the form of this development is at least partially under player control and the game is capable of reacting to the changes. 4. Game Master: At least one, but not all, of the participants has control over the game world beyond a single character. A term commonly used for this function is 'game master', although many others exist. The balance of power between players and game masters, and the assignment of these roles, can vary, even within the playing of a single game session. Part of the game master function is typically to adjudicate on the rules of the game, although these rules need not be quantitative in any way or rely on any form of random resolution. 5. Interaction: Players have a wide range of configurative options for interacting with the game world through their characters, usually including at least combat, dialogue and object interaction. While the range of options is wide, many are handled in a very abstract fashion. The mode of engagement between player and game can shift relatively freely between configurative and interperative [sic!]. 6. Narrative: Role-playing games portray some sequence of events within the game world, which gives the game a narrative element. However, given the configurative nature of the players' involvement, these elements cannot be termed narrative according to traditional narrative theory. (Hitchens and Drachen (2009), 16) According to (1), a role-playing game is set in a fictional world, and that clearly holds for religious fictionalism. Some fictional worlds are entirely fictional, like Tolkien's Middle Earth, but most are fictionalized versions of the real world, with some minor change to the details (this is generally true of fictional worlds outside the science fiction and fantasy genres). Other worlds lie in between: they are based on the real world, but add substantial fictional elements (as in Harry Potter, where the fictional world is based on the real world but also contains elements like wizards and dragons). The world of religious fictionalism is of this third kind. It corresponds to the real world in all non-religious aspects (if we do not choose to leave the fiction open in some regard, as discussed in the previous section), 19 but it also contains a substantial addition: the central claims of some specific religion are all true, as are their implications (insofar as these implications are recognized by the fictionalist). It is the world of realist religion taken as a fiction.
Condition (2) concerns the participants. The religious game clearly has players, and I address the issue of whether it also has something like a game master in the discussion of (4). However, while LARPing is a genuinely social activity, many fictionalists will be playing in a realist religious setting where they are the only player. Can one role-play in solitude? The answer to that question seems to be affirmative. Most popular computer role-playing games are single player experiences, like the Elder Scrolls or Fallout-series. Likewise, I can see no reason in principle why you could not have your own, private LARP or solo play a tabletop game (of course, the latter requires a second non-player participant as game master). Just because role-playing games other than single player computer games are commonly social affairs does not mean that they are necessarily or even essentially so.
Condition (3) is about role-characters. Given our discussion in the previous sections, it seems reasonable to suggest that it holds. The emphasis put on character progress here is especially interesting, since it ties in well with the instrumentalist justification of religious fictionalism in terms of personal development.
Condition (4) requires the game to have one or several game masters. In traditional role-playing games like Dungeons and Dragons, one person assumes the function of a game master who controls and defines the game world. It is the game master who presents the fictional world and gives the boundaries of the game, deciding what actions are possible and not. The game master controls narrative events which are not directly affected by the players, and in role-playing games using rules, it is the game master's task to interpret these rules. In short, the game master is the final arbiter concerning what is fictionally true in the fictional world.
Obviously, the religious game does not have game masters in the traditional sense. Does this mean that religious fictionalism cannot be a role-playing game? Not necessarily. Other writers have rejected the inclusion of a game master in the definition (Arjoranta (2011)), and not even Hitchens and Drachen require a literal game master. In digital roleplaying games, they notice, the game master's functions are performed by the software: the software presents the fictional world to the player, sets the range of actions available to the player, and determines how the players' actions affect the world.
I would like to suggest that for religious fictionalism, the game master's functions are upheld by realist religion in much the same fashion as they are carried out by the software in a digital game. Remember that what is fictionally true in the religious fiction is determined by what is literally true in the realist religious community. Much like the player of Dungeons and Dragons cannot make her fiction different from what the game master dictates (she cannot, for example, fight a troll if the game master says that there are no trolls around), the religious fictionalist cannot make her fiction much different from the realist religion in which she wishes to participate (if she is a Christian fictionalist, she cannot worship Shiva since realist Christianity denies Shiva's existence). Realist religion defines and controls the fictionalist's game world, shaping its narratives and defining its 'rules'. Much like a real game master, it decides what is fictionally true.
Also (5), which concerns interactive freedom, seems to hold. Since a fictionalist can pretend to do whatever a religious realist in the corresponding situation could do, there are no artificial limits to the ways in which she can interact with the game world.
According to (6) role-playing games contain 'narrative elements' rather than genuine narratives. In role-playing games, players create their own stories, and we cannot expect these stories to conform to the plot-structure by which a narrative is defined according to standard narrative theory. Religion (especially Abrahamic) is often understood in terms of a great, all-encompassing narrative, and an important aim for anyone engaged with religion is to partake in that story. It might therefore seem like the 'make your own story' aspect of role-playing is hard to combine with religious fictionalism.
However, it is important to distinguish between the bearing narrative of a fictional world, and the story of the individual role-character. Many role-playing games take place in fictional worlds defined by a great narrative, such as Tolkien's Middle Earth. But playing in Tolkien's world does not imply that your character's story should conform to the stories told by Tolkien. When role-playing in Middle Earth, you make your own story in the fictional world created by Tolkien. The point is that even though the grand narratives which define Middle Earth (as told by Tolkien in Bilbo and Lord of the Rings, etc.) cohere to narrative theory, the story of your player-character can still go wherever you want.
This holds for religious fictionalism as well. A Christian fictionalist is not supposed to pretend to be Jesus or St Paul, neither is she supposed to repeat their stories literally. True, the imitation of Christ is an important part of being Christian, but the fictionalist is playing a religious believer who imitates Christ, not Christ himself. And unlike the story of Christ, which is fixed and corresponds to narrative theory, there are no limits concerning how the story of the believing role-character might develop.
Since it is able to meet Hitchen and Drachen's qualified definition, it seems both warranted and fruitful to view religious fictionalism as a role-playing game. 20 The impact of the role-playing perspective Adopting a role-playing perspective on religious fictionalism comes with far-reaching theoretical consequences. In this section I explore these consequences by revisiting two key issues and consider how they should be re-addressed given the current approach.

Does the fictionalist need independent reasons for action?
It is commonly assumed that fiction cannot provide adequate grounds for rational action, which is considered a fundamental problem for fictionalism. In order to participate in religious life, the fictionalist must act in accordance with the fiction, but how can she do so without being irrational? Natalja Deng asks us to consider a divorced fictionalist participating in a religion which does not permit its adherents to remarry. Clearly, this divorcee would be irrational if she refrains from remarrying on fictional grounds alone (Deng (2015), 201).
Proponents of fictionalism generally accept the restriction that a fictionalist needs independent reasons for her actions. They hold that a fictionalist can only engage in activities like prayer and worship if she has specific non-fictional reasons to do so (Eshleman (2005), 189;Deng (2015), 209;Le Poidevin (2016), 185-187).
According to the present perspective, the demand for independent reasons seems misguided. It does not take into account the difference between passive fiction, which is appreciated in the imagination only, and active fiction, which requires pretence. It is true that passive fiction, like literature, does not provide a suitable ground for action. This, however, does not generalize to active fiction. In active fiction, it is trivial that action can be based on fictional reasons alone. This is especially true when it comes to roleplaying games. If in a LARP I am attacked by other players pretending to be assassins, I do not need non-fictional reason to respond to that. The fiction tells me to fight back or attempt an escape. If I respond by doing nothing since I have no real-life reasons to do anything, I spoil the game for everyone. In a role-playing game, I must act as my rolecharacter would, and I do not need independent reasons to do so. This holds for religious fictionalism as well. If I am a Christian fictionalist, I will participate in the celebration of the mass, not because I have independent reasons to do so, but because it is what my rolecharacter would do. Role-playing becomes practically impossible if we demand nonfictional reasons for individual actions.
This does not mean that reality puts no constraints on fictionalist action. The fictionalist role-player is free to act as she sees fit within the fiction, but the fiction can never take precedence over real-life concerns. If the fictionalist is a vegan, she will not eat meat when in character, and Deng's divorcee will remarry if she wants to. And of course, the fictionalist also needs some independent, non-fictitious reasons to participate in the fiction in the first place.

The benefits of fictionalist religion
The fictionalist seeks the benefits associated with religion while rejecting its truth-claims. Is this a feasible approach? Are not the benefits of religion dependent on a realist understanding? As previously mentioned, proponents of fictionalism often think of these benefits in terms of moral awareness and spiritual growth. But what reasons do we have to assume that religion's potential as a means for personal development remains intact on fictionalism (Palmqvist (2021), 57-58)?
Viewing religious fictionalism as a role-playing game radicalizes this problem, since it seems plausible to suggest that it is the role-character rather than the fictionalist herself who will receive the sought-after benefits. It is the role-character who is religious, and the player is supposed to leave everything that happens in character behind when stepping out of the game. If you spend a LARP being afraid of the barbarians, you should certainly stop being afraid when the LARP is over. Not doing so is the route to Quixotical delusion. While the role-character might well experience moral and spiritual development, this development will be part of the character and left behind when the game ends.
In the LARP community, there is something known as the bleed-out effect, which means that the experiences of the role-character might 'spillover' and affect the player's thoughts, feelings, and actions outside the game. Is this something the fictionalist could take advantage of, to partake in her role-character's spiritual development? The suggestion that a fictionalist might exploit the bleed-out effect recalls Pascal's famous idea that a religious unbeliever might participate in religious life and act as if she believed, as a means to start believing (Pascal (1958), 68). While it has often been pointed out that acquiring belief this way is epistemically objectionable, perhaps the idea has more merit when it comes to moral and spiritual development?
Unfortunately, bleed-out is a mental-health hazard which can lead to social and emotional delusion and mental illness. As explained by Diana Leonard and Tessa Thurman in their neuroscientific treatment of the phenomenon, bleed-out is caused by an overexhaustion of the player's self-regulatory resources, which are used to compartmentalize the fictional experiences of the role-character (Leonard and Thurman (2018)). Role-players usually take great precautions to avoid bleed-out, which is why a LARP seldom lasts more than a few days and ends with a debriefing to facilitate leaving the rolecharacter behind. Taking bleed-out as a path to personal development seems an ill-advised and outright dangerous approach.
Answering the original objection, Eshleman has pointed out that as long as religious fictionalism remains a means for 'human flourishing', there is no need to demand that the fictionalist develops in the same way as the realist (Eshleman (2005), 195). This response marks a retreat from the basic instrumentalist idea of preserving the benefits of realist religion. However, it also opens for a broader instrumentalist approach, which can allow for other kinds of positive developments to function as motivation. If the fictionalist cannot reap the benefits of religion, perhaps we should consider instead the possible rewards of role-playing?
I want to highlight a study by Meriläinen (2012) concerning the effects of role-playing games on personal development. Members of the Finnish role-playing community were asked to self-assess how their hobby affected their personal development. The respondents reported an increased ability in areas such as creativity and empathy, which Meriläinen interprets in terms of 'empathic intelligence'. While these results must be seen as tentative considering the scope of the study and the perils of self-assessment, they do indicate that role-playing games can be used as a tool for personal development. They also indicate that the benefits of role-playing might not be so different from the spiritual and moral development for which many proponents of fictionalism hope. In short, the prospects of offering instrumentalist justification for religious fictionalism as a roleplaying game look promising.

Concluding remarks
By focusing on the game of make-believe, the differences between religious fictionalism and other kinds of philosophical fictionalism become salient. Unlike scientific or mathematical fictionalism, religious fictionalism requires immersion and pretence on a high level. The religious fictionalist needs to create a role-character to be able to participate in the game, and religious fictionalism can be fruitfully viewed as a role-playing game, even according to a qualified definition of such games.
Taking a role-playing perspective on religious fictionalism invites us to reconsider many of the key issues associated with the approach. I have explored two such theoretical reorientations, giving some first suggestions as to how they should be approached. First, the roleplaying perspective strongly suggests that no independent reasons are required for fictional action, since making such demands in a role-playing game makes the game impossible to play. Second, it seems reasonable to suppose that it is the benefits of role-playing rather than the benefits of religion that the fictionalist will be able to receive by her engagement.
When it comes to working out the full theoretical implications of the role-playing perspective, much remains to be done. For example, I have assumed that we should prefer what I referred to in the introduction as 'extended fictionalism', since it provides a fuller religious experience than a practice limited to special religious actions. The role-playing perspective might give us reason to question this assumption. Due to the dangers associated with bleed-out, a LARP seldom lasts more than a few days and it ends with a debriefing to facilitate leaving the role-character behind. This suggests that the fictionalist's game of make-believe should be restricted in much the same way, but the question needs to be assessed in detail. How should we restrict the game of make-believe? Should we introduce a debriefing? Is anything like a religious life possible on fictionalism?
Other issues have to do with the relation between religious fictionalism and other roleplaying games. If religious fictionalism can be fruitfully viewed as a role-playing game, does that mean that it is nothing more than a role-playing game? If the game is played for the benefits of role-playing rather than religion, does this mean that it is interchangeable with other role-playing games, so that it is merely a matter of taste and aesthetic preference whether I choose to play the religious game rather than Dungeons and Dragons?