Vices in Gaming: Virtue Ethics and Endorsement View

1 Introduction

‘It’s just a game!’, ‘Nobody is getting hurt!’, ‘Nothing really happens!’, ‘It’s not real!’. These would be the typical replies of a ludic amoralist if confronted with the reproach that he is acting morally wrong by playing certain video games. Actually, there is no need to be morally concerned about violent video games. After all, if the amoralist is right, it’s just a game. Nobody gets hurt.

Neither utilitarianism nor Kantian ethics seem capable of remarking anything substantially regarding the moral evaluation of such video games or the virtual actions within them.¹ Virtue ethical approaches such as McCormick’s or Patridge’s also seem to fail due to incorrect or imprecise implementation.² Schulzke, for example, shows that even an Aristotelian cannot raise a general objection to violent video games and that on the contrary, some such video games can even have a positive effect on our character.³

Nevertheless, we do have the intuition that there is something morally or ethically wrong with some games or the playing of some games. Moreover, we have the intuition that this cannot just be attributed to bad taste. This intuition is not false.

I want to argue that virtual actions in video games, even if they elude any moral evaluation, can still be evaluated according to a virtue ethical approach. However, this requires a detailed analysis of the cases in which such an assessment is possible. For this, I will use Frankfurt’s theory of volition.⁴ The possibility of evaluation itself, however, is independent of Frankfurt’s theory. It could also be argued for on a basis of other theories of action, such as Searle’s concept of intention in action.⁵

I am particularly focusing on Frankfurt’s concept of second-order volitions. According to him, it is with the help of these second-order volitions that we identify with our first-order desires and, ultimately, it is with the help of these structures of desire (as I will call them) that we identify with our actions.⁶ An analysis of these structures of desire allows us to assess when we completely identify with certain actions and which actions we identify with to some extent but do not completely endorse. It is these actions, which we identify ourselves with completely, that cannot elude a virtue ethical evaluation, even if they are virtual actions.

Then, I will address a possible objection that video games also offer the opportunity to try oneself out in a role—even in a morally objectionable one. This objection entails that the opportunity to try out what it is like to be a torturer or a murderer without actually harming anyone is to be assessed positively and therefore the action of trying out oneself by playing such games cannot be morally objectionable. I then examine the extent to which my approach is compatible with Ostritsch’s endorsement view.⁷ Finally, in the conclusion, I raise some further-reaching questions that are worth exploring in the light of what I shall show hereafter. Virtue ethical approaches, I argue, can not only help us better understand the ethical aspects of video games. They also allow us to examine ethically other areas where moral theories traditionally face problems, such as the ethics of simulation, art, and literature, for example.

2 Previous Attempts

Many attempts have been made to justify our intuition that there is something morally or at least ethically wrong with some video games or playing of some of those games. Hereinafter I present a number of these attempts and explain why they were unsuccessful.

In the context of the morality of video games, a significant part of the debate revolves around the so-called desensitation effect. That is, the idea that the consumption of violent video games (and other media) increases the risk of violence by the consumer or makes it more likely that the consumer will actually show aggressive behavior. The problem is this: There is disagreement about whether or not video games actually have such an effect. There is good research on both sides.⁸ However, the investigation of this effect is empirical in nature and hence not primarily of interest to philosophy. Nevertheless, I believe it to be worthwhile to reflect philosophically on the morality of video games, regardless of such an effect. In the following, I will therefore not engage further with the empirical questions, which are certainly interesting in the context of video games. Instead, I now want to examine some philosophical approaches to the ethics of video gaming and why they have not been successful so far.⁹

For example, Matt McCormick explains that neither the Utilitarian nor the Kantian can morally criticize the playing of video games as long as they do not have a negative effect on our moral judgment or contribute to an increased risk of us doing real harm.¹⁰ However, McCormick argues that the Aristotelian can certainly raise an objection: Playing video games can have a negative influence on our character, preventing us from developing virtuous behavior and therefore impedes us in our pursuit of eudaimonia.¹¹ The question of why the Aristotelian is able to assume an effect on the player’s character more easily than, say, the Kantian, remains unanswered by McCormick. Even if McCormick wants to employ a virtue ethical approach here, his proposal is not actually virtue ethical. Instead, his approach is utilitarian in nature as well. For only insofar as video games have a negative effect on our character, i.e. corrupt us morally, are they morally objectionable. Obviously, what matters for him are the consequences of our video game actions, namely those on our ethical character.

Marcus Schulzke argues that neither the utilitarian nor the Kantian can object to anything of substance concerning the playing of video games.¹² One could condemn playing such games on a Kantian basis because for Kant the mistreatment of animals is morally objectionable as it degenerates our moral judgements. Therefore, one could argue, playing violent video games is wrong because video game characters are analogous to animals.¹³ Schulzke shows, however, that this is the wrong kind of similarity. Instead of a physiological similarity, in the case of video games, there is only a representational one.¹⁴ He also defends video games against McCormick’s argument, which states that from a virtue ethical perspective playing such games should be objected as these have negative effects on our character. Schulzke argues that the Aristotelian cannot fundamentally object to violent video games, because certain forms of violence, for Aristotle, can be compatible with the virtues.¹⁵ But exactly like McCormick, Schulzke seems to overlook that the Aristotelian cannot object to violence, video gaming, or any other type of action in general anyway, due to the particularity of virtue ethics, but instead must always judge whether the particular action (the token) is appropriate to the circumstances in a particular situation.

On the contrary, according to Schulzke, video games could even have a positive effect on our character by allowing us to train our moral and ethical decision-making ability in simulated moral dilemmas.¹⁶ However, Schulzke does not explain why video games could affect our character in this case, but not in the case of the Kantian argument regarding our treatment of animals and the degeneration of our moral capacities. One possible way to explain this apparent contradiction lies in the assumption that due to the previous intention to train moral judgment—i.e., insofar as identification with the training action takes place—a causal connection arises from playing the game to changing the character. But more on this later.

In Resolving the gamer’s dilemma, Christopher Bartel does not deal directly with the question of how playing video games or virtual actions in video games in general could be morally condemnable.¹⁷ However, in trying to solve the dilemma, he provides an answer nonetheless. He tries to solve the gamer’s dilemma by claiming that acts of virtual murder have no morally condemnable consequences. Acts of virtual pedophilia, on the other hand, do have morally condemnable consequences: For Bartel, it is the “eroticization of inequality inherent within child pornography [that] is harmful to women because it undermines the promotion of sexual inequality for women,”¹⁸ with which he tries to justify our intuitions about the immorality of virtual pedophile acts.

But it seems obvious that we do not consider video games that contain such virtual acts or the playing of such games to be morally objectionable because women experience long-term negative consequences due to these games by cementing sexual inequality. We condemn such games because something is wrong with the performed actions in the game itself. However, Bartel fails to explain (and justify) that intuition.

It is precisely this inherent characteristic of a morally (or ethically) condemnable virtual game act that Stephanie Patridge tries to identify: “What is particularly attractive about virtue theory for my purposes is that it provides non-consequential resources for assessing the moral status of our in-game activities.”¹⁹ She concludes that there are certain representations in games that are inherently bad. And they are embedded in a social context that makes it difficult to embed the representations in another context that allows a different, morally non-objectionable, reading of these representations.²⁰ According to her, such representations have an incorrigible social meaning. She continues that players who do not distance themselves from such representations while playing have a morally flawed character.²¹

But in doing so, she overlooks the fact that such representations are not themselves morally objectionable. The context in which they appear is relevant in their moral assessment. Watching the film Schindler’s List, for example, is certainly not morally condemnable, even though it contains representations with incorrigible social meaning.²² The same can also apply to games, just think of the Wolfenstein series.²³ This is also criticized by Sebastian Ostritsch.²⁴ He tries to show a different way in which our intuition regarding video games can be justified: For him, not playing the game, but the game itself should be the focus of the moral evaluation. The endorsement view presented by Ostritsch claims that a game—and thus also playing this game, at least as a game—is morally condemnable if morally condemnable actions or images are not only represented, but if the game itself suggests (or endorses) a certain morally objectionable worldview. This is because “[s]ometimes a piece of fiction is not merely fictional because, on a pragmatic level, it also endorses a normative view about the real world.”²⁵

With this theory, Ostritsch succeeds in explaining one part of our intuition. He can explain why the normal play of morally objectionable games is morally condemnable. However, the endorsement view cannot explain the complementary case: How is it possible that playing normal games in a morally reprehensible manner is morally condemnable? Ostritsch admits that a non-consequentialist approach, such as Patridge’s, may be promising.²⁶ In the following section, I try to outline such a non-consequentialist approach. Ostritsch’s approach, however, has another problem: one could argue that only actions may be morally objectionable. However, video games themselves are not actions. So how can they be morally objectionable—they don’t do anything. This is another problem I will address with the approach defended here.

3 Identification Theory

Our intuition towards video games seems to be partially explained by the endorsement view: Some games are morally condemnable in themselves, which is why it is also morally condemnable to play these games as games.²⁷ This can explain why games like Hatred can be called immoral:²⁸ They are immoral because they cross the line between game and reality by endorsing a morally objectionable worldview. This shows how normal gaming behavior involving morally problematic video games can be ethically condemnable. But how can it be that abnormal gaming behavior of normal games is ethically problematic? For this, I first need to say a few things about Harry G. Frankfurt’s theory of volition.

According to Frankfurt, our desires can be classified as desires of different orders.²⁹ First-order desires relate directly to our actions. The will of an agent is his action-guiding first-order desire. In addition, we can also have second-order desires that relate to our first-order desires. For example, while I may have a first-order desire to eat chocolate cake, I may at the same time have a second-order desire not to have that first-order desire for chocolate cake. A special form of these second-order desires are our second-order volitions. With their help, we identify ourselves with our first-order desires and thus take a stand on whether and which of our first-order desires we want to become effective for action.

Frankfurt illustrates this with the example of an unwilling drug addict:³⁰ Because of his addiction, the addict has, on the one hand, a first-order desire to take the drug. On the other hand, he tries to rid himself of it, i.e. he has the first-order desire not to take the drug. The addict is an unwilling one because he also has a second-order volition that his first-order desire not to take the drug should become effective. According to Frankfurt, it is through this second-order volition that the addict identifies himself with his first-order desire not to take the drug and distances himself from the other. If the other first-order desire ultimately prevails and he does take the drug because the addiction was too intense at last, he will feel, according to Frankfurt, that in a strict sense it was not he who was acting.³¹

Frankfurt also points out the normative significance of these second-order volitions, with the help of which we adopt a certain attitude towards our first-order desires. Thus, according to him, for the attribution of moral responsibility it is not necessary that the agent has free will or the possibility to act otherwise. If an addict identifies with his first-order desire to take drugs through a corresponding second-order volition, he is also responsible for this drug use—even if the addiction would have led him to continue using drugs anyway. This is because he adopts a certain attitude toward his first-order desires (and ultimately his actions), authenticating them as, in a strict sense, his own. If we identify with a first-order desire that, when it becomes effective, leads to a morally condemnable act, i.e. if we authenticate it as our own in the strict sense, we are also morally responsible for this act because, after all, we recognize it as an act of our own.³² This is also illustrated by Tim Henning. He, too, points out this normative dimension of internal desires, i.e. desires that we identify ourselves with: “The basic phenomenon […] is the fact that having an ‘internal’ desire gives rise to certain normative reasons in a way in which merely finding oneself with an ‘alien’ desire does not.”³³ Nevertheless, we should keep in mind the importance of taking into account these internal desires in our moral evaluation, i.e. those desires that we identify ourselves with through the corresponding second-order volitions.

Let us now return to video games. A problem with Ostritsch’s endorsement view is that it can only explain one half of our intuitions: normal game actions can be morally objectionable if they are performed in the context of immoral games. But it also seems possible that abnormal game actions in the context of perfectly innocent games can be ethically problematic, too. How is this possible?

Consider, for example, Bus Simulator 2018 as such a perfectly innocent game.³⁴ The reason I have chosen Bus Simulator is that it stands in such a strong contrast to many so-called (ultra)violent video games such as Mortal Kombat or Call of Duty. Modern Warfare 2.³⁵ Not only is there no mission objective in the game, according to which one should harm any virtual people. The game is also not designed in such a way that the player is encouraged to try out such a gameplay, as in the case of GTA V, for example (or who hasn’t—to choose another example—cruelly trapped their Sims in a swimming pool with no exit just to see what happens?).³⁶

Imagine a player playing this Bus Simulator in a very peculiar way. Instead of playing towards the mission goal and, for example, completing a certain route in the given time, this player seems to be only interested in killing as many pedestrians as possible in the most brutal, devious, and creative way. At this point, I suppose, our moral intuition kicks in that there is something not quite right here. I want to claim that this intuition is not wrong. The endorsement view cannot help us at this point since it is an in itself innocent game.

I argue that the ethically problematic actions of the player playing Bus Simulator in this way differ from other, ethically unproblematic game actions in that they are based on an ethically problematic structure of desire on the part of the player. By that, I do not mean the desire to perform these actions in the real world instead of the virtual world. That would mean facing the same empirical problems as the utilitarian, for example. What makes this player’s actions (killing innocent virtual humans in the most brutal way possible) ethically problematic is that the desire to perform these actions is, in a strict sense, the player’s desire. Because of this kind of identification with this desire, I argue, it is irrelevant whether the desire relates to virtual or real actions—and therefore this game action is morally condemnable even if the game itself is not morally objectionable.

At this point, Frankfurt’s distinction between desires and volitions of different order becomes relevant. This is because with the help of the internal desires, i.e. the desires with which we identify ourselves by means of a second-order volition (i.e., adopt a certain attitude towards them), it is now possible for us to determine in which cases ethically problematic behavior on the part of the player is present. If the aforementioned player plays Bus Simulator in a normal way, there is nothing ethically wrong with it. For example, at some point in the game, he may have a first-order desire to play the game in the spirit of the game, i.e., he may have the desire, for example, to steer the bus around a corner, to have passengers get on the bus, and so forth. In order to be able to conceive of his actions as, in a strict sense, his own and not as something that is merely happening to him, the player also needs to identify himself with them or authenticate them as his own. Thus, for example, the player has the second-order volition that his first-order desire to cross the red light (if he consciously does so) becomes effective in causing (or guiding—whichever theory of action is preferred) his action. That could be, for example, because he is convinced that this will enable him to keep to the timetable, which in turn is important to achieve the mission objective. Through the corresponding second-order volition, this first-order desire becomes an internal desire of his with which he identifies himself.

However, an analysis in a case involving morally objectionable gaming can show where the characteristic element of the morally objectionable gaming action lies. Imagine the player is playing the game in the problematic manner described above, i.e. for him it just seems to be about killing as many people as possible in the most devious, brutal, or creative way imaginable. If we examine the structure of his desires and, above all, that of his internal desires, we will see that they differ in one essential trait from the desire structure mentioned before. The player may have a desire to kill people in the most brutal, devious, or creative way possible and he believes that he can achieve this by running them over with a bus. Furthermore, he has the second-order volition that his first-order desire becomes effective in causing (or guiding) his action and thus identifies himself with his virtual actions, making the first-order desire an internal desire.

There already seems to be a difference between the structures of desire of a player₁ performing morally unobjectionable gaming actions and a player₂ performing morally condemnable gaming actions. Based on the various structures of desire we see that with player₁, in contrast to player₂, there is a certain distance between him and his virtual actions. Of course, he somehow identifies himself with them. This is the prerequisite that they appear to him as his actions in the first place. Nevertheless, he is aware of them as game actions. He identifies with them as game actions and in this respect, there is no major difference in whether he identifies with game actions that include following a timetable in time, betraying a (virtual) friend of his game character, or solving a puzzle in a puzzle game. All game actions are game actions insofar as they can be said to have something to do with the game in a strong sense: In performing them, the player wants to get closer to the goal of the game, try out game mechanics or just have fun with the game. Player₁ identifies with his actions while being aware that they are just virtual game actions.³⁷

However, player₂ does not seem to distance himself from his gaming actions this way. For him, it is not about enjoying the game (as a game), trying out game mechanics, or pursuing the goal of the game. He has (concerning his game actions) just one internal desire: killing people in the most brutal, devious, and creative way possible. He seems to have no interest in interacting with the game (as a game) or, for example, in trying himself out in a different role.³⁸ The distance in identification to his internal desires, which was the case with player₁ through the awareness of his actions as game actions, is no longer given in the case of player₂. But an undifferentiated identification, i.e. an identification as such) with the first-order desires of player₂ seems at least ethically questionable. Why is that so?

David Lewis’ axiology works quite similar and may help to illustrate the difference between morally objectionable gaming actions and morally unobjectionable ones. For Lewis, second-order desires are identical to values. He, too, explains this using the example of a drug addict. For him, values cannot be first-order desires, because “[t]he thoughtful addict may desire his euphoric daze, but not value it.”³⁹ Lewis concludes that values just are second-order desires: “We conclude that he [the addict] does not value what he desires, but rather he values what he desires to desire.”⁴⁰ So, Lewis’ axiology once again emphasizes the normative dimension of desires as second-order volitions.

It should be noted, however, that for Lewis there are no intrinsic values that are independent of the second-order desires.⁴¹ For him, the only requirement for something being valuable is that the respective second-order desires come about under ideal conditions. Having said that, in the following I want to argue that Lewis’ theory is well suited for investigating the problem of ethically problematic behavior in video games, even if one neglects his further ontological assumptions about values. So here again we see the normative significance of higher-order desires (or volitions). But, as said, Lewis’ axiology is quite assumptive. And Frankfurt himself did not adequately flesh out the relationship of desires, volitions, and values.⁴² Therefore, what follows must be somewhat tentative. Nevertheless, I think that against the background of what I have outlined up to this point, one can plausibly assert the following: with the help of our second-order volitions, we identify with and distance ourselves from our first-order desires. Doing this means that we adopt a certain attitude towards them. Under certain circumstances, this attitude can be a virtue or a vice. (This is not to say, of course, that all of our second-order volitions are either virtues or vices.) For example, because of his virtuousness, the brave person identifies with his first-order desire to withstand a dangerous situation and distances himself from his first-order desire to flee. Thus, our structure of desire (i.e., the presence of a particular pair of a first-order desire and a second-order volition) may indicate the presence of a particular virtue (or vice).

Now it becomes clearer why the gaming actions performed by the two players may be assessed differently. While player₁ maintains a certain distance from his actions and only identifies himself with them as game actions and positions himself accordingly, this distance is not present in player₂. As stated earlier, he is not interested in the game, or anything related to it. The first-order desire which he identifies himself with (that is, his internal desire) seems to be completely independent of the gaming context. This does not mean that he de facto also has a (first-order) desire to kill people in the real world in the most brutal, devious, or creative way possible. Nevertheless, there seems to be a certain blending of reality and virtuality in the sense that his desires are detached from the game context and that it is of no importance to him that his game action merely happen in the context of a virtual game. (One could also say that the actions of player₂ are ethically objectionable because he does not value the fact (in the sense of his structure of desire) that his actions are only game actions instead of real actions.) A normal player is aware (even if at times only subconsciously) that what he is doing is merely play and his actions of stealing, murdering, etc. are merely acts of pretending. Because he identifies with his actions as game actions that are in a certain sense not real, the normal player keeps some distance between him and them. The problematic player, on the other hand, identifies with his actions as such. In this respect, he does not care that what he is doing is merely taking place in a game. Every normal player has the first-order desire to carry out these actions within the game and the corresponding second-order volition that this first-order desire is (and remains) effective in causing (or guiding) his actions only in a gaming context—thus making his first-order desire to perform these actions an internal desire exclusively in a gaming context. Player₂ lacks this second-order volition.

I argue that this is ethically objectionable because there is an alienation of one’s own values from certain intrinsic values. One such intrinsic value would be, for instance, to respect the lives of others. (Presupposing this I consider to be uncontroversial.) Such alienation occurs when a person’s structure of desire assesses something to be of no value to them, although it actually is of intrinsic value. If a person has a structure of desire (i.e., a pair of a first-order desire and a second-order volition) which indicates that the person does not identify himself with such an intrinsic value—for example, respect for other (real) people’s lives—but instead considers it valuable to kill other people in the most brutal manner possible, the resulting behavior is ethically condemnable. After all, he stands behind this value as a person because “[t]he desire does not move him either against his will or without his endorsement.”⁴³

By stating that the behavior resulting from this attitude, which is indicated by the corresponding internal desire as an action-causing (or action-guiding) desire, is ethically objectionable, I mean that someone has a character flaw in virtue ethical terms who identifies with such desires (and therefore actions), i.e. authenticates them as his own. Identification with such internal desires or the resulting actions seems to be incompatible with the identification of oneself with virtuous actions or with internal desires that result in virtuous actions. Ultimately, such identifications simply indicate a certain attitude. In our case, the structure of player₂’s desires indicates nothing but a vice.

This also seems to be what Jens Kjeldgaard-Christiansen has in mind when he investigates and (as he calls it) splinters the gamer’s dilemma. In the case of virtual pedophile acts, Kjeldgaard-Christiansen says, we must assume that the player who performs such virtual pedophile acts is playing this game not despite the ability to perform such acts but rather because of it.⁴⁴ From this we conclude that the player has certain immoral desires, Kjeldgaard-Christiansen says:

My suggestion […] 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.⁴⁵

In this respect, Kjeldgaard-Christiansen also seems to suggest a virtue ethical assessment: We consider certain gaming actions to be ethically problematic insofar as they indicate certain vicious attitudes (or, in Kjeldgaard-Christiansen’s somewhat misleading terms, immoral desires) in the player. This is the only way we can explain to ourselves why anyone would play such games (more on the relation between vice and immoral games in the following section).⁴⁶

In this section, I have argued that a non-consequentialist approach can justify our intuition about the ethical demerit of some game actions even if they are performed in the context of morally unobjectionable games. This holds whenever the player does not consider things to be valuable despite them having intrinsic value. He does so by identifying with a particular first-order desire through the help of a second-order volition. Some pairs of first-order desires and second-order volitions indicate a virtue or vice of the agent. In games, this can happen when the player’s normal distance from his game actions, which usually prevails due to the awareness of the gaming context, is lost. He, so to speak, gets lost in the game and no longer cares about the difference between game and reality. However, it is not necessary to claim that the player actually has the same first-order desires regarding his real-world actions that he has regarding his gaming actions. It is sufficient that he no longer is aware of the ludic character of his actions and that his identification with his actions is detached from the game context, i.e. concerning his identification with his actions, it is not important to him whether they are game actions or real actions.

Now, one could object that the approach I defend here is, at least at first glance, rather similar to Christoph Bartel’s approach.⁴⁷ He, too, refers to the Frankfurtian identification with certain desires in order to argue for the possibility to hold players morally accountable of certain game actions. Bartel, however, mainly uses Frankfurt’s framework in order to argue for the moral responsibility of the player in video games as deterministic worlds. Thereby, he focuses primarily on the player’s freedom of action in video games as well as the role of free will with regard to moral responsibility. However, the question of whether video games constitute such deterministic worlds and whether this leads to conclusions regarding the moral responsibility of gamers does not interest me at all in the context of my inquiry. Likewise, the question of free will is negligible for my purposes. In contrast to Bartel, I defend a virtue ethical approach to video games. Within such an approach, we can conceive certain structures of desire as indicating the presence of virtues and vices in the player. In addition, pursuing such a virtue ethical approach has the following advantage: even if it turns out that there is nothing morally wrong with playing certain video games (or playing any video games in a condemnable way), this play can still be ethically problematic, i.e. imprudent.⁴⁸ Ludic amoralists, such as Johan Huizinga, claim that the moral rules of everyday life are suspended within games.⁴⁹ Bartel tries to prove the ludic amoralist wrong. Instead, I argue that even if we grant the Ludic amoralist that moral issues must be set aside in the context of games we may be facing ethical problems regarding video games nonetheless. That is precisely when certain vice-indicating desire structures are present in the player. These are present when the player no longer cares whether or not he performs any (normally) morally objectionable actions only in the game instead of the real world. Anticipating another possible objection, in the next section I examine to what extent the approach proposed here is compatible with Ostritsch’s endorsement view.

4 Identification and Endorsement View

Another possible objection to the approach I defend here comes immediately to mind: Player₂ only tries himself out in another role. For example, as part of him playing GTA, he could try out what it is like to be a murderer or a traitor without actually harming anyone. Not only is it morally commendable that this trying out a role happens without harming others but the act itself may also hold ethical benefits.

The objection is partly correct; however, it does not affect my approach. I agree that trying oneself out in a role—even a morally reprehensible one like the role of a murderer—can have some ethical merit and it is to be evaluated positively that this is done in the context of a video game where one does not cause any real harm. However, I do not agree that the behavior of player₂ can be understood as trying out such a role because the player who simply tries himself out in a role exhibits a different structure of desire than the one found in player₂.

This is because the player trying out himself concieves of his gaming actions as self-orientation actions. But is it not true that he has to identify himself with his actions, has to make them his own, and has to have the appropriate internal desires? It is, but the difference lies in the content of these internal desires. Both players identify themselves with their gaming actions through an internal desire. The player trying out himself perceives his actions as actions of self-orientation. Thus, although he makes his actions his own, he adopts a certain attitude towards them, namely the attitude that his actions serve to try out a role. In this respect, he distances himself from them to a certain extent and does not make them entirely his own. That is to say, he only identifies with his actions insofar as they are actions of self-orientation. Player₂, on the other hand, does not seem to care about whether or not such a distance remains. He identifies himself with his actions as such. For him, it is just as irrelevant whether they are actions of self-orientation as it is irrelevant to him that they are gaming actions.⁵⁰

That is probably what Ostritsch means in the context of the endorsement view by saying that some games—those which endorse an immoral worldview—may not be played as games.⁵¹ But what does it mean not to play a game as a game? He does not mean that one is obliged to never play these games: “It may be legitimate to play such games […] but only if the player is aware and critical to the moral status of the game.”⁵² And Ostritsch continues: “An approving or even a missing disapproving emotional and/or cognitive reaction to an immoral game is itself immoral.”⁵³

Ostritsch claims that in immoral games one is morally obliged not to enjoy them in the strong sense but only in the weak sense.⁵⁴ My approach is largely compatible with this claim. Because if someone experiences joy in the strong sense while playing such an immoral game, he thereby seems to identify himself with his gaming actions in such a way that one can no longer speak meaningfully of a distance between the player and his actions. The player in this case identifies himself with his actions entirely. However, if a person plays an immoral game not as a game, but “as part of a research for an article about the ethics of computer games,”⁵⁵ then there is the same distance to his actions as is the case with the player trying out himself in another role. He identifies with his actions, makes them his own. But he does not make them his own as game actions (and certainly not as actions as such) but as research actions. That is to say that he is aware that he is doing research and that the purpose of his actions is not to enjoy the immoral game. Insofar as his actions are described as game actions, he distances himself from them (i.e., he does not identify with them), but insofar as they are described as research actions, he does make them his own.

A game that is immoral according to the endorsement view may then be described as one that is designed in such a way that it endorses an immoral worldview. That is, that to be enjoyed as a game it requires a certain structure of desire from the player, which leads to immoral game actions with which he identifies himself. The game is immoral in that it requires the player to identify with immoral actions or to adopt an ethically problematic (i.e., vicious) attitude in order to be played as a game. So, the question is quite interesting to what extent Ostritsch’s endorsement view may also be analyzed through structures of desire.

If this is possible (and I see no problem why it should not be), another problem may be avoided: Because when we judge morally, we usually evaluate actions or (in the case of virtue ethics) people or their attitudes.⁵⁶ But what do we mean by saying that a game is immoral? The game itself does not do anything. Ostritsch’s point can also not be taken to mean that the act of playing is morally objectionable because it is through playing that the game is actualized as an interactive system of signs (and thus the action of actualization may be immoral). He points out that in this regard there is no difference between games, movies, or literature.⁵⁷ But then what could be morally objectionable about a game? Only playing a game may be morally objectionable. For the virtue ethicist, however, this does not pose a challenge. As far as he is concerned, the problem lies not in the game itself but in the player’s reaction to the game. And Ostritsch also thinks that what is relevant is how the player reacts to the game. Does the player realize that this game should not be played as a game (but, for example, only as a research object)? And does he react accordingly? When asking such questions, however, we must look for the flaw not in the game but in the player. Here, as I pointed out, a close examination of the player’s structure of desire may tell us a great deal about his ethical integrity.

5 Conclusion and Further Implications

In summary, structures of desire may help us decide more easily what exactly is ethically (or morally) objectionable about a video game. A game is immoral if it is designed in such a way that playing the game as a game (and not, for instance, as a research object) requires the player to have a structure of desire that excludes certain virtues or requires certain vices. Like this, the game necessarily prevents virtuous attitudes or leads to vicious actions when played as such. Due to the virtue ethical character of my approach, however, no moral obligation not to play the game (as a game) can be derived from this necessarily. (For this, we would have to examine more closely what normative status virtue ethical assessments have and whether moral obligations may be derived from them. This, of course, would be a question concerning the meta-ethical status of virtues.⁵⁸)

But regardless of how the question about the player’s obligation is answered, my proposal is not so much of a new approach than an attempt to reconcile the endorsement view with a virtue ethical approach.⁵⁹ I do so by on the one hand explaining to what extent games may be considered immoral and, on the other hand, to extend the ethical evaluation framework from (immoral) video games to all video gaming activities.⁶⁰ Through this, ethically objectionable gaming actions can also be evaluated without having to resort to an immoral game. The case of an immoral game just is a special case because in this case the game to be played as a game already requires the player to have an attitude towards his actions that is problematic in virtue ethical terms.

In general, one can say that gaming actions are ethically objectionable if the player has an attitude (due to his structure of desire) that excludes certain other virtuous attitudes. A game itself is ethically (and possibly morally) objectionable if, in order to be played as a game, it already requires the player to have a structure of desire which indicates an attitude that excludes certain virtuous attitudes in the respective player.

But what are the further implications? As stated in the beginning, traditional moral approaches (such as utilitarianism and deontology) have difficulty in evaluating virtual game actions morally. In contrast, virtue ethical approaches apparently have no difficulty bridging this normative gap from reality to virtuality. Therefore, I believe that a serious engagement with virtue ethical approaches is also worthwhile regarding other normative questions in the field of virtuality. It seems likely that virtue ethical-expressivist approaches also hold answers regarding an ethics of simulation, for example. Here, too, we can expect traditional approaches of moral theory to fail because of the reality-virtuality gap, while virtue ethics may be able to bridge this gap for us.

But we encounter gaps like this in other places as well: in art or literature, for example. With regard to these areas, similar questions arise as to whether (and if so, to what extent) a book or a painting can be morally objectionable. Here, too, moral theory is faced with the question: Why should there be anything wrong with this at all? After all, nothing of this is real.

And finally, there is also the question of the ethics of games (or of play) in general. Here, too, we are faced with a normative gap. Namely, the transition from the real world to a special world of play, as Johan Huizinga notes.⁶¹ We may encounter this special world of play in many other areas, such as society in general.⁶² A serious engagement with virtue ethical approaches of games and play may therefore have much more far-reaching ramifications than may have been apparent at first.