Morgan Luck (2009) states the gamer’s dilemma thus:

Most people agree that murder is wrong. Yet, within computer games virtual murder scarcely raises an eyebrow. In one respect this is hardly surprising, as no one is actually murdered within a computer game. A virtual murder, some might argue, is no more unethical than taking a pawn in a game of chess. However, if no actual children are abused in acts of virtual paedophilia (life-like simulations of the actual practice), does that mean we should disregard these acts with the same abandon we do virtual murder? (p. 31).

There does seem to be a meaningful moral distinction between murdering someone in a video game and engaging in acts of virtual pedophilia. Virtual murder—defined as the unjustified killing of a character—is not uncommon in modern video games, such as in the Grand Theft Auto series, whereas one can only imagine the public outcry if a high-profile video game were to encourage, or even just to allow, acts of virtual pedophilia.

Rather than argue that virtual murder and pedophilia are equally impermissible, responses to Luck’s dilemma purport to identify either a relevant moral difference between murder and pedophilia modulo virtual context (Bartel 2012; Patridge 2013; Young 2016, Chap. 6; see also Luck 2009, pp. 32–35), or to question the basic premise of the argument that we do in fact systematically judge one to be permissible, the other impermissible (Ali 2015). To establish the first of these alternatives would solve the dilemma; to establish the latter alternative would show the dilemma to be ill-conceived. I am only aware of one response to the dilemma that presumes to uphold it, which is that of Young (2013).

Most promising has been Ali’s (2015) proposed dissolution of Luck’s dilemma.Footnote 1 Indeed, Luck himself (2018) judges that Ali’s attempt does in fact dissolve a strong version of the dilemma, and Young, in his book-length investigation of the dilemma, opines that Ali’s argument holds promise (Young 2016, p. 102), though he ultimately deems it unsatisfactory.

Ali’s (2015) argument, in a nutshell, is that context matters. Descriptively, it may be acceptable for players to engage in virtual pedophilia in certain cases, and it may not be acceptable for players to engage in virtual murder in certain cases. For example, we might, following Ali (2015, p. 272), imagine a game in which the player has unknowingly and unwillingly engaged in virtual pedophilia due to a tragic delusion on the part of the player character. Only toward the end of the game does the player learn that this is what he or she has done. We might think such a realization would provoke moral reflection in the player and judge it permissible insofar as it is properly and tactfully embedded in the game’s story world. But this contrived counterexample does not overturn the intuition that generally virtual pedophilia is wrong, whereas virtual murder is at least not as unambiguously wrong. Ali finds the source of this conviction in the kind of scenario we envision when we imagine what a game featuring virtual pedophilia might be like. In all but very unusual cases, it is hard to see why players would engage in virtual pedophilia if not for the fact that they derive pleasure from the activity as such. We therefore imagine players to perform acts of virtual pedophilia in simulation games through which the pedophilic act is performed for its own sake, as a way of experiencing what it is like, which seems deeply perverse. On the other hand, it is easy to imagine cases in which players would kill in a video game for instrumental reasons: as part of a friendly competition, or to overcome some challenge set by the game. We therefore default to imagining virtual murder in the context of a story or competition game in which the murder is performed as a means to some further end, which seems more permissible than virtual murder for the sake of virtual murder. If virtual murder were instead carried out in a context of simulative engagement, it would be impermissible. If this is true, then it seems that Ali has dissolved the dilemma, at least at a high level of description. In some circumstances, virtual pedophilia is indeed permissible, and in some circumstances, virtual murder is indeed impermissible.

Despite this effort, Luck (2018) considers that the dilemma has yet to be finally resolved. Though Ali’s (2015) proposed dissolution goes some way toward answering the dilemma—and strictly speaking does dissolve the unqualified version proposed by Luck (2009)—it does not work for all cases. Luck (2018) offers the following counterexample:

A game is developed similar to Counter-Strike. It is a sporting multiplayer game; that is, a game designed ‘with the intent of allowing gamers to virtually compete’ (Ali, 270). One team plays the role of the police, and others the role of criminals. The game has two modes. The first mode is sniper mode: where one team play terrorist snipers (whose aim it is to shoot as many innocent civilians as possible), and the other team play police officers (whose aim it is to limit the mayhem by moving the civilians to safety and/or incapacitating the terrorists). The second mode is molestation mode: where one team play child molesters (whose aim is to molest as many children as possible), and the other team play police offers [sic] (whose aim is to limit the mayhem by moving the children to safety and/or incapacitating the molesters). (p. 161).

Luck judges that this example once again arrays our intuitions against the pedophilia case and for the murder case, which seems right. And it does so despite provisions that, (1) there is no story to justify either gameplay mode, and (2) both gameplay modes are instrumentally motivated as means to competitive ends and thus do not foster a predominantly simulative mode of player engagement. Luck is thus able to sustain a weaker version of the dilemma. This weaker version holds only that virtual murder is sometimes permissible in cases where, all things being equal, virtual pedophilia is not permissible.

In the remainder of this article, I will attempt to advance the debate not mainly by offering an ethical resolution to the dilemma, but an explanation of its psychological basis at the level of descriptive ethics. In other words, I will attempt to explain why we think there is a dilemma in the first place. My investigation will show that virtual murder and virtual pedophilia vary as action prototypes, construed as enacted event schemas (e.g., Zacks et al. 2001, 2007), with respect to three descriptive features that may explain the moral intuitions that give rise to the dilemma. First, virtual pedophilia is suggestive of immoral desire in a way that virtual murder is not. Ali’s (2015) proposed dissolution subsumes a version of this argument, and I shall be content to develop rather than reconstruct it. Second, virtual pedophilia describes a kind of moral violation, a so-called Purity violation, which can be shown to influence judgments of character more so than other kinds of moral violation. Third, in describing cases of physical assault by a powerful adult on a powerless child, virtual pedophilia evokes a particularly objectionable perpetrator-victim relation not normally evoked by the notion of virtual murder. The two action prototypes vary, then, in ways that can be shown systematically to condition moral judgment.

I aim to show that when we consider various possible instantiations of both action types—virtual murder and virtual pedophilia—we recognize that prototypical features of the acts, rather than the acts per se, account systematically for our differing moral response. (Young (2016, p. 123) indicates a similar conviction in noting that it is not warranted simply to assume that the gamer’s dilemma is fully grounded in a “single factor”). Once the action types are contrasted in terms of their prototypical components, we may also be able to account for the precise intuitions that support the gamer’s dilemma. In addition, we may be able to account for the intuitions that support Luck’s (2018) counterexample, as well as other possible counterexamples. Finally, we can ask whether and which of these differing features seem morally relevant without yoking them under the encompassing gamer’s dilemma, which, as I shall argue, masks a finer level of moral consideration. This move splinters, rather than dissolves, the gamer’s dilemma.

Of course, explaining the psychological basis of the dilemma is not tantamount to resolving it. However, I hope also to be able to show that the low-level features that distinguish prototypical virtual pedophilia from prototypical virtual murder may contribute to a normative resolution of the dilemma. My suggestion, explored toward the end of the paper, will be that the strong negative reactions typically provoked by pedophilic representations will make us ask why anyone would want to engage with such representations if not for the fact that they respond to them in highly atypical and immorally desirous ways. Whether in a simulation game or not, the player of such a game is by that fact alone impugned as motivated by perversely immoral desire, and this would seem to change the ethical status of the player’s virtual engagement.

I will now examine the three noted distinctions between prototypical virtual murder and prototypical virtual pedophilia one by one, suggesting in each case what about the distinction causes us to judge virtual murder as being more permissible and virtual pedophilia as being less permissible.

Immoral desire

Ali (2015) proposes that

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. (p. 273).

As already noted, we do this because we can think of plenty of examples of virtual murder in storytelling or competition games that may seem to make sense of the killing, whereas we cannot think of games that would somehow justify virtual pedophilia. We therefore default to imagining a simulation game in which the player performs pedophilic acts without any justification, which seems indicative of a perversely impassioned form of engagement. However, when we explicitly consider virtual murder in a simulation context, we may find that it is also intrinsically objectionable.

The intuition that virtual murder is more permissible than virtual pedophilia may, in part, come down to this observation. Murder is not necessarily suggestive of a corresponding desire in the player to murder because we find plenty of examples of virtual murder that is narratively or competitively motivated. But Ali (2015) also adverts to a deeper source of this conviction. Prototypically, a pedophilic act is an autotelic act: an act carried out because of the pleasure or satisfaction it affords the agent as an end in itself rather than as a means to some further end. It is very hard to see why anyone would want to engage in pedophilic acts, even in a virtual context, if not for the fact that they find pleasure in the act as such. In a ludic context, therefore, pedophilic acts provoke strong reactions because they are seen to reveal a disturbing fact about the perpetrator: immoral desire.

I want to push this point even further. It is not just that we happen to have a hard time imagining why anyone would engage in virtual pedophilia if not to experience perverse gratification. It is not just a matter of a contingent association, as assumed by Luck and Ellerby (2013, pp. 231–232) and Young (2013, 2016, Chap. 3). The concept of pedophilia describes a drive. The very type of the offense slots immoral desire under its intensional scope. If the above example of a permissible form of virtual pedophilia, whereby the player’s moral reflection upon learning that he or she has performed a pedophilic act, succeeds, it does so precisely by blocking a core feature of the pedophilic concept: the suggestion that the player has engaged in the act with a view to obtaining perverse gratification. Indeed, what makes the example succeed is that, technically, neither the player nor the character has enacted virtual pedophilia because the concept picks out a desire in addition to the behavior motivated by that desire.

Much of the philosophical literature on the permissibility of (the enjoyment of) fictional representations discusses the desires that underlie, and may be affected by, the production or reception of such representations (for pertinent recent examples, see Ostritsch 2017; S. Patridge 2011; Schulzke 2010). At least since Plato, the worry has been that media representations feed the base or immoral drives behind our fictional engagements, whether or not that worry has any basis in reality. We do not worry that playing Super Mario Bros. will make us jump compulsively in real life, but we might worry, when we consider the goal in many Super Mario games of rescuing Princess Peach, that playing Super Mario games will make us think of women as a sort of prize that you get for overcoming grave obstacles. We worry, in other words, about the goals and desires that media might be seen to inculcate in, or to “prescribe” for (Gaut 1998), their audiences. With this in mind, it is not surprising that we tend to moralize an activity in the play sphere whose very definition includes immoral desire.

The yoking of means and ends inherent to the concept of pedophilia concept helps explain people’s intuitive revulsion when considering virtual pedophilia as one of the two prongs of the gamer’s dilemma. After all, the formulation of the gamer’s dilemma only evokes the two prongs as conceptual categories, and the concept of murder does not imply the presence of immoral desire. It can be instrumental. Thus, the prongs of the dilemma differ in that virtual pedophilia implies antisocial desire whereas virtual murder does not, though virtual murder can certainly be (de)contextualized in such a way as to solicit desirous engagement, as in a simulation game. The fact that our single-minded judgments when faced with the abstract formulation of the gamer’s dilemma give way to nuanced and contextually sensitive judgments in concrete cases, as Ali (2015) illustrates and Luck (2018) concedes, illustrates the point that it may not be the action types that cause us to respond more negatively toward virtual pedophilia than virtual murder. It may instead be assumptions about how these action types are, or would be, concretely instantiated and responded to.

Luck’s (2018) multiplayer scenario, quoted in full above, can be used to nail these observations down to a concrete example. Luck contrasts the abhorrent notion of a sporting molestation mode in a video game with that of a sniper mode not unlike the gameplay of many mainstream titles, such as Counter-Strike. The fact that in the first case the object of the game is pedophilic whereas in the second case it is the shooting of civilians makes both games morally problematic, but Luck contends, I think rightly, that audiences would find the molestation mode more objectionable. One reason for this disparity may be that the molestation mode makes the player engage in an activity whose sole motivation in a realistic context would be the obtainment of perversely immoral gratification. The framing of the competitive game in the very terms of immoral desire makes us suspect that the virtual experience of gratifying such a desire is, or is supposed be, central to the player’s experience. If we adjust the sniper mode’s representations likewise to invite desirous engagement, I believe we evoke a parallel conviction that the gameplay is immoral. What happens, for instance, if instead of shooting digitized “civilians” the player were tasked with shooting Blacks, or women, or Japanese? In these three cases, the type of victim is specific and evocative enough to make us think that it has been chosen for a particular reason, and that, disconcertingly, players would prefer it for that reason. (Patridge’s (2013) analysis would agree on this point in noting that pedophilic representations “possess representational details that make it more reasonable to see [them] as a reflection of our lived moral reality” (p. 33, my emphasis).) Each case suggests that the representational nature of the game’s violent content is intrinsic to the gratifications afforded by the game. Therefore, the nature of the representation suggests another goal than the putative goal of the competitive gameplay: for example, the enjoyment of the killing of women qua the killing of women, and not merely as a means to score points in a competitive bout.

The pedophilic representation, then, is particularly offensive because it would seem to infuse unequivocally immoral gratifications into gameplay that may explicitly aim at other ends, such as sporting competition. This may be one factor that psychologically distinguishes the two prongs of the gamer’s dilemma.

Purity violations

The concept of a Purity violation evolved from the Divinity ethic in anthropologist Richard Shweder’s mappings of cross-cultural morality (e.g., Shweder et al. 1987). The concept has been developed as part of Haidt and colleagues’ Moral Foundations Theory (MFT) (Graham et al. 2013; Haidt 2007; Haidt and Joseph 2004), which has dominated the moral psychology literature of the last decade and proven viable for media psychological research into the moral intuitions of media users (e.g., Grizzard et al. 2019; Matthew et al. 2014; Tamborini 2011) including specifically moral decision-making in video games (Joeckel et al. 2012; Krcmar and Cingel 2016). MFT assumes that humans have a number of different moral “palates” that respond to different kinds of moral violation (see Graham et al. 2011 for theoretical backdrop and tests of validity). The theory describes these different palates as the foundations of human moral judgment. For instance, we have a Care/harm (henceforth Harm) foundation that causes us to object to harmful actions if there is no clear justification for the harm done, such as the justification of acting in self-defense. Pertinently, we also have a Purity/sanctity (henceforth Purity) foundation that causes the repugnance people feel toward acts perceived to be physically or spiritually degrading—to somehow corrupt us. Such transgressions are often self-centered: masturbation, uncleanliness, “impure” thoughts. But they can also be directed toward others, as in the cases of rape and pedophilia. This latter type combines the Harm and Purity foundations in being both disgustingly impure and manifestly harmful.

When we take stock of a rapist or pedophile offender, what we tend to feel is not best described as anger or outrage, but as repugnance, or moral disgust (Hutcherson and Gross 2011). The philosopher Leon Kass (1997) has expressed this sentiment in its strongest form:

Repugnance … revolts against the excesses of human willfulness, warning us not to transgress what is unspeakably profound … [it] may be the only voice left that speaks up to defend the central core of our humanity. Shallow are the souls that have forgotten how to shudder. (p. 20).

We feel a powerful visceral aversion—a “soul-shuddering”—toward the kind of mind that would perpetrate such an act and sense immediately that something is deeply wrong with such an individual. This type of transgression is typically associated with the sexual domain. As Prinz (2007) notes,

Sexual mores are obvious candidates for moral disgust because sex is a carnal act that saliently involves the transfer of bodily fluids. Since these things can elicit disgust on their own, it is unsurprising that violations of sexual rules are regarded as disgusting. (p. 73).

Psychological studies have shown that transgressions that target the normative Purity domain and thereby evoke moral disgust are seen to be highly diagnostic about the moral character of the perpetrator. Chakroff and Young (2015) summarized this effect in the title of a recent study: “harmful situations, impure people.” Observers of impure acts tend to short-circuit normal processes of moral deliberation: they do not stop to think about why the perpetrator did what he or she did (Giner-Sorolla et al. 2018; Russell and Giner-Sorolla 2011). As a consequence, they do not find occasion to mitigate the offender’s moral responsibility. Crucially, this asymmetry is greater in fictional contexts. Experiments by Sabo and Giner-Sorolla (2017) show that the fictional nature of literature, film, or video games causes observers to attribute relatively more blame to Purity than to Harm violations in these media as compared to their judgments about empirical reality. Interestingly, this effect extends to the readers, viewers, and players of fictional scenarios. If such media users are described as enjoying depictions of Purity offending, such as acts of sexual deviance, they are judged more negatively than if they are described as enjoying Harm offending. The authors conclude that Purity violations in fictional contexts are not granted the same “fictive pass” as Harm violations. It seems that the visceral nature of our reactions to Purity violations are allowed to predominate in an imaginative context, perhaps precisely because it is visceral: immediately felt and unreasoned. It is not the kind of thing that people will reason away by saying that “it is just a game.”

Supporting this interpretation, Haidt (2001) has influentially argued that moral judgments are predominantly felt, in the Humean sense of a targeted sentiment, rather than reasoned out. Moral reasoning often amounts to a post hoc justification of an antecedently prevailing attitude. The most infamous example of this effect stems from a Purity violation: the fact that study interviewees made to judge a hypothetical consensual, safe, and private incestuous encounter between a brother and sister failed to provide coherent reasons for their unequivocal condemnation of the act. The standard explanation for this finding is that the interviewees felt a strongly aversive response to the description of the incestuous liaison, which made them judge it to be wrong, and then proceeded to offer token objections that were met in the description of the act: the couple risked passing on recessive genes to the offspring of their affair (but they both used protection), risked ruining their relationship (but the affair was conducted in complete mutual trust), or risked inspiring the same kind of behavior in others, who might not be as careful as them (but it was done in privacy). The same “moral dumbfounding” (Haidt 2012, Chap. 1) effect could also characterize responses to virtual pedophilia.

If it is true that Purity violations are especially salient in fictional contexts, and if this fact helps explain our visceral aversion to the thought of virtual pedophilia, then other types of Purity offending should produce a similar response.

Consider rape. You might feel a sting in response to my unceremonious introduction of this fraught topic. This response would not be unlike the immediate response one might have to contemplating pedophilia. Both types of act blend harm with a sexual violation of the body of the victim, and both bring to mind unpleasant images of prolonged, corporeal, desire-driven offending. If in Luck’s (2018) counterexample we replace the “molestation mode” with a “rape mode” whereby the object of the game is to hunt down and rape innocent adult victims, I think, again, that we succeed in capturing at least some of the revulsion of the pedophilic case. Such a game would be highly objectionable. Incidentally, Vaage (2015a, Chap. 5, b) has argued that rape may exert a strong antipathetic effect across media. She notes that, whereas antihero characters in TV often kill and sometimes even murder, they do not rape. This, she suggests, is because a raping antihero would not be an antihero at all, but a singularly contemptible villain. We feel a strong aversion to fictional characters who rape and tend to be even more repulsed by fictional rapists than by fictional murderers. Something analogous may well happen when we judge a player who would engage in virtual pedophilia.

Similarities of responding to virtual pedophilia and virtual rape do not establish that the impact of a character-assailing Purity violation eo ipso is driving our asymmetrical condemnation of virtual pedophilia compared to virtual murder. Like an act of pedophilia, rape suggests (though it does not imply) that the offender is driven by an immoral desire whose satisfaction lies in the act itself rather than in something external to the act. Indeed, Purity violations typically imply such self-reflexive desire (Giner-Sorolla and Chapman 2017). For example, impure, shameful thoughts are entertained because of prurient urges that become the contents of those thoughts. To what extent the efficacy of virtual Purity violations in producing moral censure comes down to this fact is difficult to say, but the Purity perspective at least illustrates the point, already made, that our worries about the potential negative impacts of fictional representations track the goals and desires such representations appear to promote. Debates surrounding the potentially immoral contents of video games, for example with regard to the notorious Atari 2600 pornographic games, such as Beat ‘Em & Eat ‘Em and Custer’s Revenge, often center on the sexual domain, although depictions of non-sexual violence are much more common. In a more recent example of this tendency, a normally inaccessible mini-game in the extremely violent Grand Theft Auto: San Andreas, which allowed the player to engage in consensual sex, stirred controversy. Players found ways to access the mini-game by hacking the game’s code. As a direct consequence of this discovery, Grand Theft Auto: San Andreas was banned in Australia, withdrawn from many US outlets, and the Entertainment Software Rating Board (ESRB) adjusted its rating of the game from Mature to Adults Only (Parkin 2012).

Sex and sexually suggestive themes in video games regularly draw headlines, not to mention age restrictions, even though sex between consenting adults is generally deemed perfectly moral. Sex, in a virtual context, becomes a Purity violation if not appropriately hemmed in by ESRB ratings and stickering. The rationale is that children and other vulnerable minds should not be exposed to matters that naturally belong in the domain of reasoned adulthood. Tellingly, this argument targets character. Sexual representations, like other Purity violations, may “damage” children’s sexuality by instilling in them what for a young child would be unnatural and potentially harmful desires (e.g., Ross 2012). Whether or not such worries have any legitimacy, it seems a matter of empirical fact that we do tend to moralize sexual, Purity-connoting acts more strongly in virtual environments than we do acts anchored solely in the Harm domain (Sabo and Giner-Sorolla 2017). The unreasoned, character-assailing nature of Purity violations, therefore, is likely a contributing factor to our revulsion at virtual pedophilia.

Perpetrator-victim relation

If based in raw probability, the cognitive prototype of a murder should be that of an adult male killing another adult male. This is because adult males are disproportionately likely to be murderers and to be the victims of murderers compared to the rest of the population (Buss 2006; Daly and Wilson 1988). A pedophilic offense differs in this respect: by definition, a pedophilic offense is carried out by an adult toward a minor.

I rehearse these unsurprising facts as a reminder that murderous and pedophilic acts differ not just in terms of the acts described as physical transactions between a perpetrator and a victim, but also in terms of the types of moral agent and patient we expect to encounter in each scenario. What happens to our intuitions in response to the gamer’s dilemma when we level this associational asymmetry with a concrete example?

Consider once again Luck’s (2018) counterexample of a game that features both a sniper and a molestation mode. Luck does not match the two modes in terms of gameplay elements not explicitly specified by his abstract formulation of the gamer’s dilemma. In molestation mode, the player molests children. In sniper mode, the player does not kill children, but “civilians,” presumably grown-ups of an unspecified background. It is also noteworthy that the less offensive sniper mode evokes very different imagery from the molestation mode. You snipe from a distance, but you do not molest from a distance. Sniping is methodical, even clinical; molestation, at least as a prototype, is personal and corporeal. The two modes specify very different types of violence. It is not too surprising that they should provoke very different reactions.

Consider “slicing mode” as a third alternative mode. As in Luck’s (2018) sniper mode, the aim of one team in slicing mode is to murder. However, in slicing mode, you murder children, and you do so by seizing them and stabbing them. It takes time to do this, of course. You want to make sure they are good and dead, which is typically ensured by slitting their throats.

Notice that slicing mode comes closer to the gameplay that one would expect to find in molestation mode than does sniper mode. I take it that our moral intuitions toward sniper mode and slicing mode come apart. We think slicing mode is more disturbing and objectionable than sniper mode. This difference is explained by features below the level of description of the gamer’s dilemma. The abstract nature of the dilemma obfuscates this disparity because the relation between perpetrator and victim that is intrinsic to pedophilia is not intrinsic to the concept of a murder. However, when we envision such a relation between perpetrator and victim in the case of virtual murder, our intuitions may change accordingly.

Findings from moral psychology confirm that “personal” (corporeal and graphic) harm is judged more blameworthy and disturbing than “impersonal” harm (e.g., Nichols and Knobe 2007; Wisneski and Skitka 2017), and that some categories of victim are more distressingly salient than others (Batson 2011, Chap. 2; Gray et al. 2012; Schein and Gray 2018). “Children,” as Schein and Gray (2018, p. 56) note, “are the most prototypical moral patient.” A young child appeals to the “tender sentiments,” to once again reference Hume, and harm toward children produces more moral outrage than harm toward adults. We appeal literally to this sentiment when we use the clichéd phrase, think of the children. Of course, the phrase is clichéd for a reason. The reason may be that children activate innate dispositions to care for the young and the helpless, or it may be that we have culturally constituted children as symbols of hallowed innocence, or it may be both. Whatever is the case, important facts of contextually sensitive responding may be obscured by the gamer’s dilemma. Apart from a very few exceptions, such as Deux Ex, child murder simply is not a live option in commercially available video games. Thus, the streets of the notoriously amoral Grand Theft Auto games are completely devoid of children. It is not a stretch to suppose that child murder is not made possible in Grand Theft Auto because that would lead to significantly greater condemnation and sanctioning of the controversial series.

When we adjust the mental images evoked by abstract notions of virtual pedophilia and virtual murder to align in respects other than the abstractly definitional, then, per hypothesis, our moral intuitions also begin to align. Once again, this flexibility illustrates that the intuitive punch of the gamer’s dilemma rests on contingent assumptions about how virtual murder and virtual pedophilia would be concretely instantiated.

The gamer’s dilemmas

The gamer’s dilemma is an important primer to important questions. It highlights deep inconsistencies between our moral convictions as they apply to the realm of the real and the realm of the virtual. However, it is easy to forget that the dilemma is an abstraction. And it seems we do not merely wish to examine abstractions. We wish to be able to judge concrete, determinable cases.

If indeed we want to be able to make such discriminating judgments, then it would be helpful to know exactly which prototypical features of the encompassing dilemma make for its aporetic grip. I have suggested three such features here. First, the case of virtual pedophilia is suggestive of immoral desire. The very concept of pedophilia yokes means and ends together, whereas murder does not. Second, virtual pedophilia is a Purity violation. Purity violations attract character judgments and moral disgust, and this effect is asymmetrically accentuated in fictional and virtual settings. Third, the pedophilic case specifies a particularly objectionable schematization of a powerful adult offending corporeally against an innocent child that the concept of virtual murder does not specify. When instantiated to mirror each other in all but strict definition, virtual murder and virtual pedophilia attract more comparable judgments.

Of course, explaining our judgments in this way does not by itself validate them. But neither does it debunk them. It may be possible to integrate these explanations of our differing responses in arguments that are relevant to a normative resolution of the gamer’s dilemma. For one thing, insofar as we object more strongly to Purity violations in fictional media because we allow ourselves to rely on gut feelings in response to purely fictional wrongs, then that may be an argument that we should not oblige this moral bias. It seems arbitrary to privilege gut feelings asymmetrically in virtual over real settings. It is hard to see how this asymmetry could in any meaningful way be morally “truth-tracking” (Brosnan 2011; see also Shafer-Landau 2012 for related discussion). One can keep sampling the distinguishing features of the prototypical pedophilic case as they contrast with the prototypical murder case to ask whether these distinctions are morally meaningful. In many cases they may not be, which would strengthen the case that virtual murder and virtual pedophilia are equally impermissible—or permissible, as the case may be. It may thus be possible to expose as ethically inert the intuitions that split our judgments.

On the other hand, I do think that the three distinguishing features of the pedophilic case I have discussed highlight an ethically relevant theme. The theme is that the pedophilic case seems to invite, and seems made to attract, a perversely and immorally desirous form of player engagement. Pace Ali (2015), engagement with pedophilic representations seems not just to be informative about the player’s motivational set in the context of a simulation game.

How so? First, and as already discussed, the fact that the very concept of pedophilia specifies an immoral desire makes us suspect that the player of a game revolving around virtual pedophilia will also be driven by such a desire. The act contains its own immoral reward in a way that virtual murder does not. Why, we may ask, would anyone want to play through pedophilic scenarios if not because they desire to play through them—to simulate them, even? Why frame otherwise innocent competition, for example, in so disturbingly suggestive themes?

A similar argument works for the Purity case. If it is just empirically true that we have strong intuitive objections to virtual Purity offenses, we have to ask ourselves why anyone would want to play a game designed around them. Games are meant to be inviting and fun, after all. It stands to reason that anyone who would enjoy pedophilic gameplay would not be experiencing the moral and visceral aversions that people normally feel. Disconcertingly, they would be motivated toward, or at least not repelled by, such gameplay. Engagement with inherently unpleasant representations suggests that the person who would so engage is choosing to do so because of these representations rather than in spite of them. Why else design and play such a game?

The relation between the perpetrator and victim in the pedophilic case, finally, suggests something similar to the two preceding cases. We have deep-seated aversions to the acts and imagery implied by this relation. We do not like to see children getting hurt, and we especially do not like to see innocent children getting hurt up close and in graphic detail. We ask ourselves: why design or play a game that revolves around these offensive representations if not for the fact that some people would play such a game because, rather than in spite, of such representations? To reiterate, there may be good alternative answers to this question, as well as to questions about the inclusion of other kinds of offensive content. One could imagine a game in which it would be permissible for the player to hurt children for reasons to do with the thematic import of the game, such as is arguably the case in Bioshock, a game in which the choice between helping and exploiting others is essential to both gameplay and narrative. But that is precisely the point: if the offensive activity is not embedded in some justificatory structure, we will conclude that the activity is supposed to contain its own reward. We reach this conclusion because aversive representations, such as violence against children, presumably do not feature in a commercial product in order to make it less appealing.

On all three counts, something like the following pattern emerges: virtual pedophilia is typically aversive in ways that virtual murder is not. Stating this comes close simply to stating the gamer’s dilemma. But if that is so, we come to ask ourselves why someone would play games that contain virtual pedophilia. We conclude that they would be engaging in this counter-normative activity because they desire it, and this seems to make a real moral difference. This argument fits with the observation that cases of seemingly permissible virtual pedophilia appear to deny the player such perverse engagement. Ali’s (2015) example of a game protagonist discovering facts about his or her own pedophilic history would be unlikely to stir pedophilic desire unless the bygone acts were reported or shown to the player in gratuitous detail, in which case we would object. Likewise, we would not object to a researcher playing through pedophilic scenarios in order to investigate them. At least in pertinent cases, what makes us judge a game to be immoral is whether the game appears to invite immorally desirous, or at least callously uncaring, forms of engagement. Likewise, what makes us judge players for their virtual activities is whether they appear to be motivated by immoral impulses, or, at least, whether they fail to respond appropriately to a medium that offers offensive gratifications (see Hazlett 2009 for a discussion of the reach of such response moralism.)

It may be said that this argument seems too intellective in form to ground our differing responses to the two cases. After all, I keep invoking players’ intuitions and gut reactions rather than their reasoned ethical stances and principles. Do I really think that players, and their would-be moral arbiters, reason in this way? However, there is nothing inconsistent or psychologically suspect in supposing that players’ moral intuitions could be produced by the processes I have explicated without consciously referencing those same processes. I am attempting to explain the intuitions, not just describe them. Besides, I think, along with other contributors to the debate surrounding the gamer’s dilemma, that most people do in fact have very different intuitions about the kind of player who would (prefer to) play a competitive game designed around pedophilic representations over the battle-arena or special ops-style armed combat that usually characterizes these games, such as in Luck’s (2018) imagined counterexample. We have different intuitions about the motivations that would sustain such gameplay, precisely because virtual pedophilia is normally considered aversive and immoral, and different intuitions about the motives of game developers who would funnel their creative energies into pedophilic representations. And those intuitions are not arbitrary, as I have argued, nor are they altogether implausible. I would be loath to play the molestation mode described by Luck. I would suspect that anyone who would want to play it would do so for reasons other than merely to compete with fellow players, and I venture that the reader would share in this suspicion. My normative proposal, then, is that pedophilic virtual content is prototypically more objectionable than murderous virtual content due to the intrinsic gratifications such content would appear to prescribe for its audience.Footnote 2 There are compelling reasons to suppose that virtual pedophilia may invite forms of desirous engagement that are widely and plausibly regarded as immoral, whereas virtual murder may not invite such desirous engagement.

Conclusion

I have argued that virtual pedophilia provokes harsher moral judgment than virtual murder because it is suggestive of immoral desire even in non-simulative contexts; induces moral disgust, which we may feel strongly even in a virtual context; and because it implies an unequal and particularly offensive transaction between perpetrator and victim in a way that prototypical virtual murder does not. These explanations by themselves suggest neither that we should “bite [the] bullet” (Luck 2009, p. 32) and judge virtual murder impermissible, nor that we should lift the ethical charge of virtual murder implicit in the gamer’s dilemma (why accept virtual murder if we do not accept virtual pedophilia?). What they suggest is that the moral intuitions that give rise to the dilemma can be fruitfully examined apart from the dilemma. This move splinters rather than resolves the dilemma because there will be more moral distinctions to consider, both descriptively, and, potentially, normatively. It might not seem very helpful so to cause our moral difficulties to multiply, but it certainly is not helpful to ignore this multiplicity of difficulties if indeed it exists. This is not to disparage Luck’s (2009) conception of the gamer’s dilemma. It is a powerfully provocative case, and I see much value in it as a primer to questions in virtual ethics. However, as Ali (2015) has proposed and as I hope to have substantiated, the dilemma can easily come to mask the fact that context matters, even in virtual worlds (see also Goerger 2017; Tamborini et al. 2013).

All of this leaves the final ethical status of virtual murder vis-à-vis virtual pedophilia undecided—if indeed we think that this question, when abstractly posed, could have a kind of provisional resolution. However, I hope to have offered some useful thoughts. With the exception of the immersed player, people rightly tend not to worry too much about what happens to characters in video games because such characters are not sentient creatures. What they worry about are players and immoral forms of player engagement. The case of virtual pedophilia seems, as a rule, to prescribe and invite a potentially immoral, desirous form of engagement, whereas the case of virtual murder does not. This distinction may, as a rule, render virtual pedophilia more unethical than virtual murder.