A gamer (or player) is a moral agent who plays videogames, and a virtual act is an act which a gamer performs, using her in-game character, on a computer-controlled (but not human-controlled) character in the game’s virtual world.Footnote 1 According to Luck (2009), gamers face a dilemma when it comes to performing certain virtual acts. This is because most gamers regularly commit acts of virtual murder (which are virtual acts that would have counted as murder had the virtual environment in which they were performed been real), and take these acts to be morally permissible. They are permissible because unlike real murder, no one is harmed in performing them; their only victims are computer-controlled characters, and such characters are neither moral agents nor moral patients.Footnote 2 What Luck points out is that this justification equally applies to virtual pedophilia (which are virtual acts that would have counted as pedophelic had the virtual environment in which they were performed been real), but gamers intuitively think that such acts are not morally permissible. The result is a dilemma: either gamers must reject the intuition that virtual pedophelic acts are impermissible and so accept partaking in such acts, or they must reject the intuition that acts of virtual murder are permissible, and so abstain from many (if not most) extant games.

There are multiple ways one might react to this dilemma. Luck (2009), and the subsequent literature that has arisen around the dilemma (specifically, Bartel 2012; Patridge 2013; Luck and Ellerby 2013), have pursued a solution which rests on finding some morally relevant distinction between the two acts, such that acts of virtual murder, but not virtual pedophilia, can be performed without moral qualms. This, however, is only one way of solving the dilemma. We can clearly see this by considering the premises leading up to the dilemma:

P1—Intuitively, gamers believe that virtual murder is morally permissible, and that virtual pedophilia is morally impermissible.

P2—Virtual murder is morally permissible because no one is (directly) harmed in the performance.Footnote 3

P3—However, it is also true that no one is (directly) harmed by virtual pedophilia.

P4—In addition, there is no other morally relevant distinctions to justify a differential attitude towards these acts.

C—Therefore, gamers must (on pain of inconsistency) reject their intuitions, either they must believe that acts of virtual pedophilia are permissible, or reject that acts of virtual murder are.

Aside from embracing the conclusion and thus adopting one of two revisionary attitudes (either partaking in virtual pedophelic acts, or abstaining from virtual murders), one may reject any of the argument’s premises, and not just the fourth premise.

Of course not every premise is equally susceptible to criticism. P2 and P3 both seem unobjectionable. With P2 it is hard to see how the harmlessness of virtual murder is not the reason for its moral permissibility. Likewise, P3 seems true, since neither act has a victim. Even if one maintains that the gamer who performs the act of virtual pedophilia is herself harmed (for instance, on virtue ethical grounds, or on the grounds that engaging in such acts increases the likelihood of performing and/or desiring the real life counterpart), it seems that the same sort of (indirect) harm will be present in cases of virtual murder.Footnote 4 This leaves P1 and P4. P4 is a promising target, and this is why the literature focuses on undermining it. But I do not think it an easy target. While there are moral differences that allow us to distinguish some instances of virtual murder from some instances of virtual pedophilia,Footnote 5 to reject P4 we need a moral difference distinguishing every instance of virtual pedophilia from every instance of virtual murder. Such a difference has not been forthcoming,Footnote 6 and indeed if my argument here is right, there is no such pervasive difference.Footnote 7 Consequently, my aim is to pursue the hitherto unexplored strategy of rejecting P1. P1 seems to me implausible because virtual acts in games serve different contexts, and these contexts matter to the moral status of a given act. The result is that it is neither the case that all acts of virtual murder are acceptable, nor that all acts of virtual pedophilia are unacceptable (in this sense, the view is both conservative and radical). This skepticism about P1 is not entirely idiosyncratic. Other gamers I have spoken to have shared my skepticism, and in addition, in the extant literature, Patridge (2013) rejects part of P1, denying that all virtual murder is acceptable. She writes “on this view not all acts of virtual murder get a moral pass….virtual murder too can be represented in such a way that reasonably connects it to our moral reality, it might also be subject to moral criticism.”Footnote 8

The constitution and moral significance of virtual acts

To see why P1 should be rejected, we first need to understand how virtual acts are constituted, and how they acquire their moral significance. This will tell us how we should individuate virtual acts, and evaluate them. We can begin by noting that the gamer’s dilemma takes for granted that virtual acts have moral significance despite their lack of victims. Plausibly, this moral significance derives from the effect on the only moral agents involved in the act, the gamers and those observing their virtual acts.Footnote 9 There are many ways in which this moral effect might arise. For instance, through increasing or decreasing the likelihood of committing real life counterparts, or through the systematic effect on one’s moral viewpoint of equivalent real-life acts. For our purposes, however, we need not commit to a specific mechanism. All we need to keep in mind is that virtual acts have a moral significance, and this significance derives from those directly engaging with (but not necessarily playing) the game.

Turning to the constitution of virtual acts, we can start by noting that like real life acts, the identity of virtual acts partly depends on the context of the performance. Whether a real life act is one of murder or self-defense depends on the situation the agent is in e.g. whether the agent is being attacked or not. With virtual acts, however, a further complication arises. This is because virtual acts have two different contexts. Their is the in-game context of the act, which is the context of the game character in its virtual world, and there is the gamer’s context, which is the context of the gamer performing the in-game acts. Clarifying the relationship between these two contexts is thus the first step in understanding how virtual acts are constituted.

A plausible first hypothesis is that virtual acts are individuated by their in-game context. For instance, consider an act of virtual killing. Whether this act counts as virtual murder or virtual self-defense depends on the situation of the character in its virtual world. This situation is provided by the game designers who build the world, populate it, and allow the gamer to interact with it through the mediation of a digital display (e.g. a tv) and some means of controlling events (e.g. through a controller) in the game world.Footnote 10 As an example, in the Uncharted (2007-present) series gamers control a (virtual) human character, Nathan Drake, who is a modern day treasure hunter placed in dangerous situations, with (virtual) human characters out to kill him. Because of this, when gamers direct Drake to kill in his world, the killings done by Drake are (usually) instances of self-defense. This in turn means that the gamer’s virtual killings are instances of virtual self-defense. By contrast, in the game Dishonored (2012), the gamer controls an assassin who is free to negotiate his surroundings with stealth instead of killing. When the protagonist in this game kills, he therefore commits murder. Similarly, when the gamer commands him to do so, he commits an act of virtual murder.Footnote 11

This first hypothesis is plausible, but incomplete. While it is true that the gamer’s contribution to the virtual world depends on what the contribution amounts to in that world, it is also true that the what the act amounts to in that world may be entirely irrelevant to the gamer’s virtual performance. A fuller picture requires that we also attend to the context of the gamer performing the virtual acts. We can see this missing part by noting that a gamer can engage with a game world in various ways. She can perform acts with or without knowing their in-game significance, and with or without regard for that significance. A gamer will know the significance of the acts if she is following the in-game narrative, and will (usually) not if she simply jumps into the game e.g. by trying it at a friend’s house. She will give regard to the significance of acts if they seek to appropriately engage with that world, performing acts that are basically in-line with the game’s narrative. She will ignore that significance if she either willfully acts in a way that gives no regard to the context, or inadvertently because she is unaware of the context.

The way gamers engage with the game affects our individuation of virtual acts. To see this consider a case in which the gamer disregards the act’s in-game significance. Imagine a morally degenerate gamer who fantasizes about murdering others, notices that he resembles Nathan Drake, and so plays the game with the sole purpose of enacting his fantasies. We can imagine that the gamer entirely disregards the narrative, perhaps muting the game and skipping any story sequences. We can now ask concerning this player, what act does he perform? On the one hand it is clear that Drake continues to perform the very same act in the game’s fiction. If the act was one of self-defense, it continues to be so. However, since the gamer is not aware of the in-game context, and anyway would choose to disregard it if he was aware of it, it seems implausible to attribute virtual self-defense to him. Instead his act is plausibly one of virtual murder. What he is doing is virtually murdering, but the way he commits this act is through Drake’s act of self-defense.

Where does this leave us with respect to the constitution of virtual acts? On the one hand, individuating virtual acts wholly by the in-game context seems to misdirect the dilemma. The dilemma is aimed at gamers and their acts, but focusing solely on the in-game context turns the dilemma into a problem for designers and the sorts of acts they depict in games.Footnote 12 Indeed the dilemma would have nothing to do with gamers engaging in those acts. On the other hand, if virtual acts depend wholly on the gamer’s context, then any in-game act will turn out impermissible or permissible depending on the gamer’s intention in the performance. The morality of virtual acts will turn on whether the gamer engages with these acts in a morally perverse manner or not, and not on the type of act performed (whether virtual murder or virtual pedophilia). In this sense, depending wholly on the gamer’s context trivializes the dilemma.Footnote 13

This reveals an ambiguity in the argument leading up to the gamer’s dilemma. Specifically, P1 is not clear about which sorts of acts it is discussing: are these virtual acts that are due to the in-game context, or virtual acts due to the gamer’s context? I think the answer is neither. We should not individuate virtual acts either by the in-game context alone nor the gamer’s context alone. The first leaves the gamer irrelevant to the dilemma, the second leaves the contents of the game irrelevant to the dilemma. By contrast, I think we should adopt an appropriate engagement view of virtual acts. On this view both contexts have significance. Acts are individuated by the gamer’s appropriate engagement with the in-game context. So, when appropriately engaging a game, the gamer lets her context be dictated by the in-game context, and in so doing performs a particular virtual act.

Sport, storytelling, and simulation

I argued that the gamer’s dilemma concerns itself with those acts that a gamer performs when appropriately engaging with the videogame world. But what counts as appropriately engaging with a given world? The answer to this question depends on the ends of the game designers in presenting a given world. Game designers have reasons that are extrinsic to the game itself, such as wishing to profit from the game, or doing what the company asks. But they also have reasons intrinsic to the game world itself. In producing a game, game designers construct a virtual world and a means of interacting with it with the intention of engaging the gamer in a particular way. It is in these ways of engaging gamers that we see what constitutes appropriate engagement with a given game.

A survey of current and past videogames reveals at least three different ways in which games seek to engage their audience. A first means of engaging the gamer is providing her with a virtual space in which a sporting or competitive event is held. Some of the earliest videogames, like Pong (1972), were designed solely with the intent of allowing gamers to virtually compete. This trend continued with arcade games focused on high scores and leaderboards, and continues today with popular multiplayer shooters like Call of Duty (2003-present). Indeed it is plausible to think that games in general have had competition as one of their central constituents (consider chess).

A second mode of engagement involves providing the gamer with a virtual space in which a story is told. Storytelling games also emerged early on in the history of gaming. For example, the early Legend of Zelda (1986) tells a simple story in which the protagonist, Link, seeks to rescue a princess, Zelda. Similarly, the Sonic the Hedgehog (1991-present) games focus on the story of Sonic who seeks to free his animal friends from the machine obsessed scientist Dr. Eggman/Robotnik. This form of game is also increasingly more popular with advances in technology that allow for more complex and cinematic stories, for instance those seen in games like Uncharted or The Last of Us (2013).

Finally, a third way of engaging gamers involves providing them with a virtual space in which various acts and events can be simulated. We can call these simulation games. Such games are harder to characterize because unlike sporting and storytelling games, they do not make explicit demands from the player. Unlike sporting games, they do not challenge the player to meet some criteria that constitutes winning e.g. a high score, or points against the other team. Though one may acquire proficiency at these games, attaining mastery over one’s virtual freedom in the game (in the sense of being better able to control the game, and thus do what one wants), such mastery is entirely optional. They are also unlike storytelling games. They do not tell a story, though they might provide a context which allows for storytelling (e.g. a virtual world, a protagonist, certain performable actions). Instead, what characterizes these games is their focus on enjoying or exercising a virtual freedom in a given domain. In providing the player with a virtual freedom, the freedom to perform certain acts, or partake in specific events in a virtual world, such games simulate our natural freedom, and in this sense are simulations.Footnote 14

Like sport and storytelling games, simulation games emerged early on in videogame history. Microsoft Flight Simulator (1982) and SimCity (1989) are two old examples. In Microsoft Flight Simulator, the game provides a realistic plane flying simulation. This sets up a minimal context for the game, but little else is demanded from the player. One cannot win, since the game does not demand that you learn how to fly the plane. Crashing stylishly, or not, is equally permitted. Similarly the simulator provides no story. The game provides the context of flying a plane, but this alone is not a story. It is more the case that one tells one’s own story in this context, e.g. how one is learning to be a pilot, how one will perform a suicidal stunt. Stories of that sort are the gamer’s own construction, even if the construct is constrained by the virtual context. This lack of demands suggests that the in-game context of simulation games is simply to allow the gamer to enjoy the lack of demands. One way to think of this is that such games provide an in-game context designed to let the gamer’s own context define the experience. Like sport and storytelling games, simulation games have become more advanced over the years. Some recent examples are The Sims (2000-present), Noby Noby Boy (2009), and Minecraft (2011).

Before proceeding, it is worth making two clarificatory points. First, the tripartite distinction above is not meant to indicate that games for the most part, or necessarily, engage games in only one of these ways. In practice, most games involve some combination of sporting, storytelling, and simulation. A series that particularly succeeds at all three of these is Grand Theft Auto (1997-present), where the games tell the story of a corrupt protagonist in a big city, but also give the player the freedom to roam the large city aimlessly as a simulation would, and provide various sporting events within that city (e.g. racing, attempting stunts). The distinction, however, is supposed to capture the idea that there are different modes of engagement that might be central to a game. Second, the fact that there is some standard of appropriately engaging a game should not be taken to indicate that gamers must, or largely do engage with their games appropriately. When a gamer plays a sporting game, to appropriately engage, she must compete, or at least try to. But equally, the player might simply join the game with the sole purpose of appreciating the scenery, or with the intent of observing one particular player. Similarly a story game might be played with the intent of beating it as quickly as possible (a ‘speed run’), thus turning it into a sporting game. The point, however, is that in doing so gamers use the game and the available actions for their own ends, not for the ends that are specified by the game.

Virtual murder and virtual pedophilia

The gamer’s dilemma focuses on virtual acts, and such acts are individuated by the gamer’s appropriate engagement with the in-game context. Since there are at least three different ways in which the gamer might appropriately engage with any given game, it is reasonable to wonder whether our P1 intuitions hold across storytelling, sport, and simulation games. In this section, I focus on storytelling and simulation games only, arguing that our intuitions about these types of games differ, and thus sufficiently establish a case against P1.

Before turning to storytelling and simulation games, however, it is worth dwelling momentarily on the in-game representations of murder and pedophilia. Discussing the moral status of videogames, Tavinor (2009) writes “There is, then, a genuine reason for the events and actions depicted in games to be morally criticized, even if fictional: representations in themselves are amenable to moral criticism, especially when they express an objectionable viewpoint.”Footnote 15 Tavinor cites two ways in which representations in games can be objectionable, the representation itself may be objectionable, and the viewpoint these representations serve might be so too. The first of these points helps solve a limited version of the problem,Footnote 16 which Bartel (2012) endorses. On Bartel’s view, we can reject P4 of the dilemma because the relevant difference between virtual murder and virtual pedophilia is that the latter but not the former counts as child pornography, which is objectionable. While this reason explains some cases, it plausibly does not explain all, since visual representations of virtual pedophilia might be highly abstracted (e.g. presented in Hotline Miami’s (2012) visual style), or nonexistent (e.g. the act is presented off screen).Footnote 17 In addition, it is unclear that all murder imagery will be acceptable, since some imagery might involve gratuitous detail. Despite these shortcomings, I think Bartel is right to point out that we can understand why some acts of virtual pedophilia are so unacceptable: they simply represent the act unacceptably (e.g. by fetishizing the act, depicting it pornographically, or with the intent to evoke curiosity or desire). So in what follows, my argument focuses on depictions or representations that are not problematic in this way. My aim is to show that provided the representation itself is not unacceptable, acts of virtual murder and virtual pedophilia are not different in any clear-cut way.

To see this, it will be sufficient to contrast virtual murder and virtual pedophilia as they occur in storytelling and simulation contexts. First, consider storytelling games. Since such games provide a narrative, it is reasonable to think that their narrative provides a moral viewpoint of the sort Tavinor discusses. Provided that neither the representations nor the viewpoint of the game are objectionable, it seems we should accept instances of both virtual pedophilia and virtual murder. To see this, consider an example using the God of War (2005-present) games. These games are undoubtedly violent, portraying a vengeful and morally questionable protagonist, Kratos. But the viewpoint of the game is not itself objectionable. The games seeks to contribute to a mythical history of ancient Greece. Just as we hear stories of betrayal, war, and other morally questionable acts taking place amongst Greek heroes and Olympian gods in Homer’s Odyssey, so we see in the God of War games. The game asks us to be moved by Kratos’ plight, and we experience, indeed enact, his heroism, resolve, and anger. At least intuitively, our acts of virtual killing in such a game are acceptable, much as the killings in The Odyssey are.Footnote 18 But now consider the possibility of a future iteration depicting Kratos committing pedophilia. The God of War games already contain scenes in which the gamer controls Kratos as he has sex off-screen, and we can imagine that in this instance Kratos, by way of cruelly punishing (as is typical of Kratos) a human colluding with the Olympians, takes his young son or daughter and molests the child. There is no question that what Kratos does is wrong. Yet is it equally clear that what the gamer does is wrong? If the gamer performs the act by way of appropriately engaging the narrative, it is unclear why that act should be singled out as questionable. By stipulation the image itself is not pornographic, and all of Kratos’ actions are at least equally questionable. Indeed, the case seems to be comparable to reading a fictional book about an immoral protagonist who, amongst others things, is a pedophile.

One may remain unconvinced on two counts. First, one may think that there is a relevant difference here, and it is that the gamer controls Kratos. This is unlike a fictional book, where the reader only observes the story. This objection is not convincing. The act of using the game controller is obviously not what is at stake, it is rather that the inputs we give allow the representation of pedophilia to unfold. But it is hard to see why this is significantly different from a book. After all, the act of reading allows the representation of pedophilia to unfold, and the reader can, just as much as the gamer, put the book/controller down.

Perhaps the thought is that the difference arises only in specific conditions, namely, those in which the gamer controls Kratos, and chooses the act freely. For any token act in a videogame, the game gives the gamer some level of freedom in performing the act. Some acts are entirely not up to the gamer, as when the game enters into a cutscene where the player watches the character do something without being given control. Such cases are more like watching a movie than playing a game. Other acts give the gamer control over the act, but do not give her a choice in whether to perform it or not. The act is simply required if the story is to make progress. Finally, the gamer may be given control over the act, and also given a choice in whether or not to perform it, since there are multiple paths of progress. Perhaps the claim is that it is only virtual acts of the last sort that cannot involve virtual pedophilia.

But even this is implausible. If as the gamer you are given the option of either having Kratos violently and graphically murder the entire family (which, in the game’s typical style, will be on-screen), or given the option of having Kratos molest the child off-screen, it is not clear that you as a gamer have chosen a virtually permissible act in one case but not the other. Moreover we can once again compare the situation to that of reading a book. A few books allow the reader to pick one or more path in progressing a story, or have more than one ending. Now imagine a case that parallels the God of War case. Is it clear that one must avoid the ending that contains pedophilia but not the one that contains murder? My intuition, at the very least, is that the answer is no. It is not that the acts do not differ, it is that both are consistent with the story being told in the game’s fiction. So it seems to me that the extent of our freedom in a storytelling game is not relevant to the permissibility of virtual murder or impermissibility of virtual pedophilia.

One might remain unconvinced for a different reason. When Luck defines virtual pedophilia, he says that such acts are ones that would have counted as pedophilia had they been real. Perhaps one could think that in the context provided in the God of War games, the act of molesting a child is not pedophilia. One might appeal to the fact that the games occur in a mythical age, or in the distant past when sexual interactions with children were acceptable. This strikes me as implausible. But even if we assume its truth, it is just as easy to come up with an example that is clearly a case of in-game pedophilia. A good example may be derived from the survival horror series, Silent Hill (1999-present). In these games, the gamer takes on the role of a protagonist who is, for one reason or another, psychologically disturbed. Silent Hill 2 (2001) is a particularly good example. In it, the gamer controls a character who has murdered his own wife. The gamer controls this character as he uncovers the repressed truth about what he has done. Consider now the possibility of a Silent Hill game that takes on an equivalent scenario involving pedophilia. This would be clearly a case of pedophilia since the game is set in modern times and in a realistic setting, yet it is not clear that a virtual pedophelic act in that game would be impermissible. Part of the point of depicting and allowing the player to perform that act is to evoke a sense of psychological disturbance in the player, and this is what Silent Hill games aim to do. Moreover Silent Hill is not an exception in the videogame world. Other games have also focused on morally disturbing scenarios. For instance, Heavy Rain’s (2010) focus is on a father forced to perform dangerous, self-harming, and immoral acts in order to retrieve his kidnapped child.

Given the above examples, I think we should conclude that when it comes to storytelling games, acts of virtual murder and virtual pedophilia can be equally acceptable/unacceptableI. This is because their unacceptability hinges on the very same features of the game, namely, the moral viewpoint of the story, and the use of objectionable or non-objectionable representations. This alone refutes P1 of the dilemma, it explains how some virtual murders are unacceptable (they occur in a story with an objectionable moral viewpoint), and how some virtual pedophilia are acceptable (they occur in a unobjectionable story). But alone, this explanation does not explain away our original intuitions. To supplement the response, we must turn to our intuitions about simulation games. Is it the case that acts of virtual murder and virtual pedophilia are acceptable in such games?

I think our intuitions shift in simulation games. We do not think that either act is morally acceptable. This is because simulation games do not provide their own narrative, they simply allow the gamer’s context to define the in-game context. So, when a gamer enacts murder or pedophilia in these games, the act is one of virtual murder or virtual pedophilia because the gamer defines it in this way. As such, the act here reflects the sorts of act the gamer finds desirable. Perhaps it is easier to see how a virtual pedophelic act in a simulation context may be impermissible, but harder to see why virtual murder is. An example, however, can shed light on this issue. Consider the popular Grand Theft Auto series, which, recall, is a particularly good case of a game with storytelling, simulation, and sporting aspects. This series of games has caused much controversy outside the videogame world, and this seems largely because the game gives gamers free reign to commit a fairly large variety of (mainly) murderous actions. But for gamers, Grand Theft Auto is not a violence simulation, it is rather a story with a particular mood. The game provides a beautifully designed, expansive, and detailed virtual worlds, where the gameplay serves as a means of unfolding a darkly humorous story that sets the stage for the immoral protagonist. Every Grand Theft Auto game provides an elaborate storytelling context, with which it critiques and lampoons society’s violence and injustice. But to those outside videogame culture, the only salient aspect is the freeroaming violence that can be enacted. Grand Theft Auto looks like a simulation game, although it is a story game with simulation elements.

Now imagine that society is right. Imagine a Grand Theft Auto stripped of its storytelling components. The game, in such a case, would actually be a violence simulator. Is it so clear that the resultant game, whose sole purpose is to put you in a city where you can virtually choose to abuse, murder, rob, and otherwise harass others, is morally permissible? Or more to the point, is it clear that the acts you engage in, when you appropriately engage with this game, are acts that we would think morally permissible? It seems to me that the answer is no. One’s virtual acts in this game are not morally permissible, not if their only pretext is enjoying the freedom and performance of these acts.

With this last piece in place, we can now provide a diagnosis of the initial plausibility of P1. P1 seems intuitive initially because when we originally consider acts of virtual murder and virtual pedophilia, we default on acts of virtual murder presented in current games, where these games are either storytelling or sporting games, but then proceed to compare those to acts of virtual pedophilia in hypothetical simulation games. The reason for this is simple: there are plenty of games that present virtual murder in storytelling and sporting contexts, and few if any games portray virtual murder in a simulation context.Footnote 19 By contrast there are no (at least well-known) games depicting pedophilia of any sort.Footnote 20 So when we are asked to think about such cases, we default on thinking of a simulation of virtual pedophilia.

One might wonder why simulations are what we should imagine by default. There may be several reasons for this. But I think one primary reason is that when we think of a game containing pedophilia, our first instinct is to think that the reason for portraying the act cannot be anything but enjoying the act, and wanting to perform it. Unlike killing, which is often a means of acquiring something more than the death of the murdered (for instance, winning the war, defending oneself, getting one’s own way, or revenge), pedophilia seems to function as an (unjustifiable) end, and not as a means to some further outcome. ‘Why would anyone commit such an act but for the enjoyment of that act?’, the thought goes. From this, I contend, we move to the idea of a game depicting the act so that it may enjoyed for its own sake, since otherwise there is little reason to perform the act. Of course, a game that allows us to enjoy the act for its own sake would be a simulation game.

This diagnosis of the dilemma may raise a final worry about the proposed solution. If virtual murder can function as a means to an end in a way virtual pedophilia is not, then doesn’t this mean that virtual murder in simulation games is both more comprehensible and acceptable than virtual pedophilia? The first act, but not the second, suggests motivations for the act aside from the sheer enjoyment of it. This may suggest a narrower version of the gamer’s dilemma. On this variant, the dilemma focuses on simulation games, and maintains that our intuitions diverge with respect to these two acts. In this case, P1 might be reconstrued as follows: Intuitively, gamers believe that acts of virtual murder are morally permissible, and similarly positioned acts of virtual pedophilia are morally impermissible.Footnote 21

But it is not clear that this variant fares much better. While killing may be a means to an (external) end in a way pedophilia is not, if a gamer enacts virtual murder, rather than, for instance, revenge, or self-defense, then it is an act done with the intent of (virtually) harming someone who is innocent. Moreover this is not a mere fantasy or imagining. It is a way of materializing the fantasy, enacting it virtually, in a way that is perceptible to the gamer.Footnote 22 It is in having this desire, and seeking to actualize it, that the virtual murder is unacceptable. The same applies to pedophilia. It is possible that we think the gamer’s actions more unacceptable in the pedophilia case than the murder case, but this is not because we think the former is okay.

Conclusion

The standard solution to the gamer’s dilemma, which notes that we differentially treat virtual murder and virtual pedophilia despite the same justification being applicable to both acts, has been to find a further, morally relevant distinction between these acts. I have argued that in place of this standard solution, we should reject the intuitions that found the dilemma. This is because once we clarify how virtual acts are individuated, and how different games seek to engage their gamers, we see that our intuitions in P1 are not sensitive to differences between types of games. Once we point out these differences, our intuitions change, and in so doing reveal that P1 of the original dilemma is mistaken.