I don’t feel wrong about killing over 9000 grunts (Hartmann & Vorderer, 2010)

“It’s just a game”

“I’m not really killing anyone”

“It’s just fantasy”

What google found for me

These responses should be familiar to gamers and anti-violent video game advocates, many would toss them aside as knee-jerk rationalizations undeserving of a response beyond a huff. However, they have neglected to consider the psychological significance behind them. This kind of rationalization was noted in academia and is now given a serious investigation by Tilo Hartmann (VU-Free University) and Peter Vorderer study published in the Journal of Communication, albeit it’s an exploratory study as there’s little information about this issue.


What makes virtual violence enjoyable rather than aversive? Two 2×2 experiments tested the assumption that moral disengagement cues provided by a violent video game’s narrative and game play lessen users’ guilt and negative affect, which would otherwise undermine players’ enjoyment of the game. Experiment 1 found that users’ familiarity with the violent game reduced guilt and negative affect, and enhanced enjoyment, whereas opponents’ nonhuman outer appearance and blameworthiness had no effect. Experiment 2 found that fighting for a just purpose, perceiving less mayhem, and framing the overall situation as “just a game” or “just an experiment” reduced guilt and negative affect, whereas the distorted portrayal of consequences did not. Effects on game enjoyment were mixed and suggest that moral disengagement cues may both foster and diminish game enjoyment.

An interesting sidenote, gamepolitics reported a study about the application of international humanitarian law in military-themed video games. Needless to say, many of them violated these laws in favour of Michael Bay explosions and Hollywood mayhem.

So why do gamers bring up such excuses? There are two perspectives the authors brought up. The virtual nature of mediated violence and moral disengagement.

The characters we shoot at aren’t real, they’re onscreen characters reacting to our actions according to specified computer scripts (i.e. if their HP is down to zero, death animations are enacted and such and such.). However, It’s quite complicated and the authors brought three arguments to counter this thinking. First, We do not remind ourselves that we’re playing a game or we’re playing with artificial constructs, we often times get immersed and flowed into it. Anthropomorphizing of video game characters often happen and automatically through our automatic social perceptions. For example, the love given to Portal’s Weighted Companion Cube or the various video game characters. Various game design features makes us perceive game entities as social entities, such as biological motion, eye-gazing, natural facial activity, display of emotion, breathing, natural vocal tones and displays of intelligence.

Second, due to humanity’s automatic social perceptions, we at first treat video game characters as social entities before recognizing them as artificial and non-social entities. However, even if we think of them as fakes, we would revert to our initial reactions because they’re so well done and convincingly social entities. To trump such anthropomorphism, it takes a lot of effort or displays of irritating media cues (i.e. a badly-made game or cut-scenes).

Third, constantly reminding oneself that what they’re experiencing is “just a game” is quite a drain and a downer during play. Video games are meant to be entertaining, engrossing with narratives, and an escape into a fantastical world within a magic circle of play. If we distance ourselves from this magic circle (IMO) or “discounting the perception of an apparent reality (in the authors’ words), we wouldn’t have much of an emotional attachment in the narrative and gaming and that makes the game boring. Therefore, there’s motivation to perceive video game characters as social entities.

The authors contend that these mediated characters are quasi-social: entities that “potentially fall into the scope of justice and have a moral status.” (okay…)

Given these arguments, a paradox arises that we should feel some guilt when we kill video game characters recognized as social entities and yet non-existent beyond the screen. But then we feel pleasure. We feel pleasure because it offers gratification of superiority, and excitement with little or no costs to our inner moral standards (to which would causes feelings of guilt). To this end, the authors investigate how certain game design features might reduce guilt by shaping players’ moral processing.

The second perspective, moral disengagement is the act of avoiding moral concerns and related aversive feelings by restructuring violent and reprehensible acts into enjoyable experiences during play. In relation to quasi-social entities, moral disengagement would make them fall out of the players’ “scope of justice” and “moral status” into entities “undeserving, expendable, and therefore eligible for harm.”

Fortunately, people don’t morally disengage in any situations. Conditions have to be met for people to morally disengage during a violent act and this is done without any conscious awareness. Albert Bandura et al. (via Richmond and Wilson, 2008) enumerated eight moral disengagment mechanisms: moral justification, sanitizing language, advantageous comparison, displacement of responsibility, diffusion of responsibility, distortion of possible consequences, attribution of blame, and dehumanization. Other related theories, moral-sanction theory of delight and repugnance and the disaparagement/disposition theory of drama, added more moral disengagement mechanisms, namely characters’ immoral behaviours. In short, these mechanisms reframe violence, both externally and internally, that doesn’t cause any cognitive dissonance within the individual or else they would feel guilt or remorse and wouldn’t find the play experience enjoyable.

Which common video game tropes corresponds to the moral disengagement mechanisms? The authors identified four mechanisms for this study: dehumanization (humans vs. non-human entities), characters’ immoral behaviours (players’ fighting against aggressive enemies killing civilians or passive enemies not killing civilians), distortion of possible consequences (characters dying realistically vs. disappearing in smoke), moral justification (raiding a torture camp as justified violence vs. defending said torture camp as unjustified violence).

If these mechanisms aren’t enough to moral disengage players from feeling guilty, they can resort to morally rationalize (i.e. “it’s just a game”) by reframing post-hoc the play experience.


Two sets of experiments have been conducted to test the four moral disengagement mechanisms.

Participants: 84 undergraduates somewhere in the West Coast of the U.S.. Average is 19.82 years old. 51 are female and 33 are male. Average play time per week is 50 minutes (SD = 73) during weekdays and 1 hour and 50 minutes (SD = 2 hours) during weekends. I suspect males have an higher playtime average. From these 84 participants, 29.8% participants listed RPGs as their favourite genre, followed by action-adventures games (21.4%).


State Guilt: 3 items (!) from the Differential Emotions Scale are used to assess guilt. Good thing they explained themselves why they used three guilt items. They considered other guilt-measuring questionnaires, but they could not modify them to suit their study (i.e. translating the questions that relates to the video game stimuli).

Negative Affect: Good ol’ and proven Positive and Negative Affect Scale (PANAS). They used the short version which is composed of 20 items. The items are rated on a 1 to 5 scale. This measure is also used for assessing guilt.

Enjoyment: 5 items to ask how much they enjoyed the game on a 1 to 5 scale. The questions include how interesting it is, how boring it is, how enjoyable it is, how much of a waste of time it is, and how fun it is.

Manipulated factor measure: this is used to assess the effectiveness of their experimental factors. Questions include (for experiment 1) rating the enemies’ behaviours, whether they are fighting against human or non-human enemies; (for experiment 2) how much support they give in fighting a cause (if they’re in the just cause condition or the unjust cause condition), and how much mayhem they perceive during gameplay.

Control factors measure: three items used as a control variable for analysis. (1) familiarity with the video game. (2) Participants’ thoughts that “this is just a game”. (3) Participants’ thoughts that “this is just an experiment in which they need to follow instructions.” All rated on a 1 to 5 scale.

Video games used: The first experiment used a modded Half-Life 2 and the second experiment is a modded Operation Flashpoint. The specific modifications will be described later. Common modifications are that the player is invulnerable, can only use one type of gun, unlimited ammo, and the number of kills appeared on the screen. I find these modifications confounding factors since they can be “it’s just a game” rationalization cues.

Experiment 1 procedure

Participants played a video game scenario where city is invaded by enemy soldiers. Participants are randomly assigned to play against armed opponents who appear either human or zombie-like. This is their experimental manipulation for dehumanization. Second experimental manipulation is the immoral behaviours of characters, the participants are either told that their opponents are either actively aggressive and attacking civilians and the player unprovoked or they are passively protesting and won’t attack unless provoked. Play time is 45 minutes.

Experiment 2 procedure

Using the participants from the first experiment. Participants are shown a cutscene where a torture camp where innocent people are murdered is going to be attacked. Participants are randomly assigned to play either as a UN soldier who will attack the torture camp and restore humanity or the paramilitary who will defend the torture camp from UN forces. This is the experimental manipulation for moral justification. The second experimental manipulation is the consequence of killing the opponent, either they are killed realistically (e.g. screaming and tumbling to the ground) or unrealistically (e.g. a ping sound and vanished characters). In addition, participants received radio messages from their commander commenting the situation in a harsh, realistic way (consequences) or euphemistically (no consequences). I don’t know about these radio messages. But I would like to hear what these messages consist of.

There’s one limitation in these procedure is that all participants go through the same path (i.e. from experiment 1 to 2), but were not counter-balanced (i.e. randomly assign participants to start on experiment 1 or 2) to account for fatigue effects or relevant confounding effects.


Experiment 1

Dehumanization: there were no significant differences between groups in terms of guilt, negative affect, or enjoyment whether participants are fighting against human or zombie-like opponents. Possible explanations followed that opponents’ aggressive behaviours may trump their physical appearance; participants lacked automatic social perceptions and the opponents weren’t considered social entities; differences between the conditions is too low because characters’ appearances were just different, but their behaviours were not. Perhaps, the opponents were not portrayed or behaved realistically. If the human opponents were controlled by better A.I. or had more human-like behaviours, say taking cover or coordinating with team mates, perhaps it might render the characters more human whereas the zombie-like opponents should be given lesser A.I., say disregard for self-preservation.

Immoral behaviours of characters: They’ve hit a snag in that the participants in the actively aggressive opponent condition reported significantly higher scores in thinking of “this is just an experiment” and that their opponents’ aggression were not significantly from the passive condition. This showed that they failed to manipulate the factor and therefore confounded their analysis. Their results revealed that participants in either group did not differ significantly in terms of guilt. However, a small and significant difference was found that those with the passive opponent condition felt more guilty, ashamed, nervous and irritable (from the PANAS) than the active opponent condition. However, this is a small difference that wasn’t picked up by the guilt measurement, although it does show that participants’ moral standards were violated. Possible explanations followed that opponents’ “attribution of blame” might have a role because opponents in both conditions have guns and those in the passive condition will attack the player if attacked, such provoked attack would lead players to attribute the same level of blame and misconduct as it would to the active opponents. Another explanation is that players in the active condition might’ve been unconsciously trying to keep their moral standards from being violated towards moral rationalization, hence the higher level of “just an experiment”.

Control factors: They found players who are more familiar with the game (Half-Life 2) reported significantly less guilt, less negative affect, higher enjoyment, higher levels of thoughts of “just a game” and “just an experiment”. The authors offered two explanations: Familiar players are desensitized to the violence (to the game’s graphics, IMO). Second is personality might be a factor, specifically low empathy and low susceptibility to guilt lead participants not to perceive quasi-social characters from their “scope of justice”. Unfortunately, they’ve ruled out the idea that players who are more aware of the artificial nature of the game are more likely to think of the experience as “just a game”, the correlational link between familiarity and “just a game” thought was not significant.

Experiment 2

Moral justification: Participants in the just cause condition significantly felt less guilty and less negative affect than players in the unjust cause condition. However, their levels of enjoyment did not differ. The results support that giving moral justifications in violent acts reduces violations of players’ moral standards. As for enjoyment, the authors argued moral justification indirectly enjoyment by reducing guilt and negative affect, which are negatively correlated with enjoyment.

Distortion of consequences: Another snag is that participants in both conditions reported no differences in the amount of mayhem during play, indicating a failed experimental manipulation. Therefore, there was no difference in levels of guilt, negative affect or enjoyment. A possible limitation noted by the authors is the killing distance between the player and the enemy, the farther away the enemy is, the less details (e.g. blood) you will see. So, maybe an experiment in enclosed virtual arena or perhaps limiting the weapons’ range to shotgun? The authors then conducted another analysis using mayhem as the independent variable and found that greater perception of mayhem was correlated with greater levels of guilty and negative affect. Albeit, there was no difference in the levels of enjoyment.

One interesting note are near-significant interaction effects. Those in the just cause and realistic consequence conditions showed higher levels of enjoyment and those in the unjust cause, but unrealistic consequence condition has show higher levels of enjoyment. Given the non-significant differences of enjoyment between the conditions of moral justification and the distortion of consequences, the authors contend that there’s a sort of cost-benefit at play. Namely, we feel guilty of causing violence, but we feel joy for being effective and powerful at our prescribed role of a soldier. Continuing on this line of reasoning, they hypothesized that enjoyment is based on a certain ratio between moral engagement and disengagement (given their near significant interaction results). An excess of either ratio would result in being boring or too frightening, respectively. Alternatively, playing the bad guy might be fun too as long they have a moral disengaging reason (i.e. “it’s fantasy” or “just a game”).

It is perhaps that (aside from technical decisions) that some game features that could induce moral engagement or disengagement are present or absent from game to game. To elaborate, Soldier of Fortune featured realistic damage display mechanics where shooting someone in the crotch resulted in grabbing said damaged crotch whereas this feature is absent from say Halo or Borderlands. This is likely based on genre or artistic conventions or perhaps cultural tastes. I mean it could get weird to see aliens grabbing their bleeding crotches, that’s just pushing the convention. On second thought, I don’t recall Modern Warfare 2 featuring dismemberment. Hmmm…. I should do a prevalence study on video game gore, saying violence isn’t enough, we should ask how violent… but then it could be a useless study in the eyes of some since the act is the only thing that mattered…

Control factors: They found that participants who thought of “just a game” experienced less guilt and negative affect. Again a sign of moral disengagement or perhaps a Freudian defense mechanism in maintaining their moral standards.

So here’s the take home message: We have some suggestive evidence that moral disengagement from video game cues have a role in influencing our moral reactions to interactive violent acts in a mediated (but ludological) environment. We also have evidence that top-down processes, such as familiarity and thinking that “it’s just a game” have also some influence in our feelings of guilt and negative affect that it should deserve more attention than it does in this study. More refinements to the methods should be implemented and the results published to confirm or not the present study’s results. We should consider emotion regulation and experience of conflicting emotions into the equation. While this study has collected some interesting studies for me to peruse (especially that chapter about the aesthetics of destruction), others might ask if moral disengagement might be mediating factor between media violence and aggression? According to Richmond and Wilson, the answer is yes.

Hartmann, T. & Vorderer, P. (2010). It’s okay to shoot a character: Moral disengagement in violent video games. Journal of Communication, 60, 94-119. doi: 10.1111/j.1460-2466.2009.01459.x


2 thoughts on “I don’t feel wrong about killing over 9000 grunts (Hartmann & Vorderer, 2010)

  1. Pingback: Cannon fodder NPCs, dehumanization and aggressive behavior (Greitemeyer & McLatchie, 2011) « VG Researcher

  2. Pingback: Mirrored morality: Moral choices in videogames (Weaver & Lewis, 2012) | VG Researcher

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s