Scenes of justified and unjustified violence in movies activate different areas of the adolescent brain


The gun violence seen in popular PG-13 movies aimed at children and teenagers has more than doubled since the rating was introduced in 1984. The increasing on-screen gun violence has raised concerns that it will encourage imitation, especially when it is portrayed as “justified.”

What was not clear until now is whether justified and unjustified violence produce different brain responses.

In a new study, researchers at the University of Pennsylvania find that scenes of unjustified and justified violence in movies activate different parts of the adolescent brain. This research is the first to show that when movie characters engage in violence that is seen as justified, there is a synchronized response among viewers in a part of the brain involved in moral evaluation, the ventromedial prefrontal cortex (vmPFC), suggesting that viewers see the violent behavior as acceptable for self- or family protection.

Performing fMRI scans of more than two dozen late adolescents who watched scenes of movie violence, the researchers also found that scenes of unjustified violence evoked a synchronized response in a different part of the brain.

Activating that area of the brain, the lateral orbital frontal cortex (lOFC), is consistent with a disapproving response to the violence.

The research, led by a team at the Annenberg Public Policy Center (APPC) of the University of Pennsylvania, was published in the journal Frontiers in Behavioral Neuroscience.

“What this response suggests is that not all movie violence produces the same response,” said senior author Dr. Dan Romer, APPC’s research director. “Adolescents disapprove of movie violence that is seen as unjustified, which is consistent with what parents have reported in past studies.

But when the violence seems justified, adolescents’ brains appear to find it much more acceptable than when it is not.”

Romer said the growth of movie violence and particularly the depiction of justified gun violence in movies raises concerns. “By popularizing the use of guns in a justified manner, Hollywood may be cultivating approval of this kind of entertainment,” he said.

Viewing movies in an MRI scanner

For the study, researchers recruited a group of 26 college students ages 18 to 22, divided between men and women. All regularly watched violent movies and 70 percent played active shooter video games.

The researchers performed functional magnetic resonance imaging (fMRI) scans on the participants as they viewed movie clips. Each participant was shown eight pairs of 90-second movie clips from PG-13 or R-rated films.

The clips featured a scene of characters talking followed by scenes of the characters engaged in violence. Half of the clips showed scenes of justified violence, the other unjustified violence. The order of scenes varied.

The scenes of justified violence showed major characters engaging in the defense of friends, family or themselves, while the unjustified violence showed characters harming others out of cruelty or ill will. Prior evaluation of the scenes by parents and young adults confirmed that the scenes differed in justification for violence.

The researchers edited the scenes from the R-rated films to remove the graphic effects of the violence such as blood and suffering so that the scenes were more directly comparable to the violence portrayed in PG-13 movies.

The scenes of justified violence came from the PG-13 movies “Live Free or Die Hard” (2007), “White House Down” (2013), “Terminator Salvation” (2009), and “Taken” (2008). The clips of unjustified violence came from the PG-13 movies “Skyfall” (2012) and “Jack Reacher” (2012) and the R-rated films “Sicario” (2015) and “Training Day” (2001).

A synchronized brain response

The researchers found that watching the movie clips produced a synchronous response in brain activity among the study participants at the same points during the movie clips.

But the brain activity differed when the participants were watching scenes of justified or unjustified violence.

“It was exciting to observe a synchronized reaction to these movie clips,” said the study’s lead author, Dr. Azeez Adebimpe, a former postdoctoral fellow at the Annenberg Public Policy Center. “Our findings clearly show that violent movies have similar effects on viewers.”

The researchers found that scenes of unjustified violence evoked greater synchrony in a region of the brain that responds to aversive events (lOFC).

They also observed synchrony in a region that responds to the experience of pain in either oneself or others, the insular cortex.

That finding was consistent with an empathetic response to the pain experienced by the victims of this kind of violence, again suggesting that the violence was seen as unacceptable.

Justified violence and the trolley problem

The ventromedial prefrontal cortex is the part of the brain that is activated when an individual is presented with a moral dilemma such as the trolley problem.

This problem poses an ethical dilemma in which a runaway train is headed toward five people who are on the tracks.

You can pull a switch and divert the train, which will kill one person on the alternate track – or you can take no action as the train races ahead toward five people.

Most people see it as appropriate to save the five people and the ventromedial prefrontal cortex tends to respond as well.

In a different version of the problem, you can only stop the train from killing others by pushing an innocent bystander onto the tracks, which most people are unwilling to do.

Research has shown that people who lack a functioning ventromedial prefrontal cortex are more willing to push an innocent person to death to save lives.

The present study’s results are consistent with this research in showing that the same brain region responds when the violence appears justified.

While gun violence in movies has been increasing, it has not been evident whether different types of violence elicit different brain responses.

The findings are also consistent with a form of ethics based on evaluating an actor’s character and motives, known as virtue ethics. Virtue ethics proposes that people judge behavior as acceptable – even when it might otherwise be seen as prohibited, as in harming others – when an actor has virtuous motives for the behavior.

In the movie scenes with justified violence, the young viewers in the study rated the use of guns by the main character as more acceptable and their brains displayed a similar response.

The current research is consistent with previous APPC research that found that parents were more willing to let their children see the same movies clips when the violence appeared to be justified than when it had no socially redeeming purpose.

This research also found that parents became more accepting of justified movie violence as they watched successive movie scenes that showed such violence.

Will justified screen violence encourage imitation?

In this MRI study, the researchers concluded: “The finding that brain synchrony discriminated between justified and unjustified violence suggests that even youth who are attracted to such content are sensitive to its moral implications.

It remains for future research to determine whether the brain responses to justified film violence we have observed foster tendencies to imitate or consider the use of weapons for self-defense or other justified purposes.

Laboratory research finds that justified film violence can encourage aggressive responses in response to provocation…

What is less clear is whether the use of guns in movie portrayals of justified violence encourages their acquisition and use for purposes of self-defense.”

In addition to Romer and Adebimpe, who is currently a postdoctoral fellow in Psychiatry at Penn, the study was conducted by Penn Bioengineering Professor Danielle S. Bassett, who is also an APPC distinguished research fellow, and Patrick E. Jamieson, director of APPC’s Annenberg Health and Risk Communication Institute.

Movies are a popular form of entertainment that not only hold people’s attention but also provide models of behavior (Bandura, 2001Bushman and Huesmann, 2006). Years of research have found reliable evidence of the socializing effects of repeated exposure to violence in entertainment media and aggressive tendencies in youth (Anderson et al., 2017). However, some portrayals are more likely to be emulated than others.

Violence that is seen as justified, in which characters kill others in self-defense or to protect friends and family are more likely to register approval than violence that has no socially redeeming value (Samson and Potter, 2016Romer et al., 2018).

Furthermore, laboratory studies have found that exposure to justified film violence reduces inhibitions to act aggressively to a greater degree than exposure to unjustified violence (Berkowitz, 1984).

Here we investigate whether different brain mechanisms underlie responses to these two types of violence as portrayed in popular movies. We also compare competing neuroscience models of moral evaluation regarding their predictions for responses to these different forms of violence.

Most research regarding the neural underpinnings of moral evaluation focuses on dilemmas that pose a conflict between following rigid moral norms such as “do not kill” and more flexible utilitarian approaches that allow killing if it saves more lives in the process. The most famous of these dilemmas involves various versions of the trolley problem where one is asked to decide if it is justified to kill one person in order to save several others.

In the various models that have been proposed to explain brain responses to these dilemmas (e.g., Greene et al., 20012004Shenhav and Greene, 2014), the ventromedial prefrontal cortex (vmPFC) plays a central role. This replicable observation has led some (Moll and de Oliveira-Souza, 2007Mendez, 2009) to suggest that the vmPFC is at the center of a prosocial neural network that codes for morally appropriate responses to harm-doing.

For example, persons with lesions in the vmPFC are more open to pushing an innocent person to death in order to save others’ lives in the footbridge version of the trolley problem (Koenigs et al., 2007), a response that is otherwise seen as an unjustified form of violence. Based on such evidence, Moll and de Oliveira-Souza (2007) argue that vmPFC activation reflects an aversive emotional response to harmdoing that promotes prosocial behavior (namely not killing an innocent person). 

Greene (2007) also focuses on the vmPFC, but in that model, the region is posited to register conflict between harming an innocent person and utilitarian action to save more lives.

Despite the plausibility of Moll and de Oliveira-Souza’s interpretation, it is not altogether clear that the vmPFC has the explicit role of guiding prosocial behavior.

One could argue that both moral norms against killing and utilitarian ethics are prosocial under the right circumstances. Thus, the viewer of violence will still have to decide whether the person engaging in violence is justified in doing so. It is here that another model of moral evaluation may be relevant. This model based on Aristotelian ethics (Casebeer, 2003) suggests a different role for the vmPFC. According to this model, the virtues reflected in characters’ motives determine the evaluation of their behavior.

When characters exhibit noble motives, their behavior is seen as acceptable whether it is undertaken in support of moral norms or utilitarian purposes. Moll and de Oliveira-Souza’s interpretation of the pro-social function of the vmPFC is anti-utilitarian as suggested by results of lesion studies in the footbridge dilemma (Koenigs et al., 2007). However, from a virtue-ethics perspective, even if someone engages in violence for a utilitarian purpose, it may be seen as acceptable, such as in defense of self or others.

In this approach, the vmPFC is more certainly to function as part of a neural system that responds to rewarding events whether they involve the self or others as this region is known to be involved in social decision making (Ruff and Fehr, 2014).

Thus, rather than reflecting aversive emotion or conflict toward the killing of others, vmPFC can be seen as part of a network that responds to justifiable motives for harmdoing and that would be expected to track a film character’s engagement in violence if it furthers a social good, such as defense of self or others.

Based on a virtue-ethics approach, one should expect the vmPFC to respond to justified film violence; however, it is less clear what its role would be for unjustified violence. Research on brain responses to violent videos suggests that watching brief video clips of violence elicits activation of the lateral orbital frontal cortex (lOFC) (Kelly et al., 2007Strenziok et al., 2010Alia-Klein et al., 2014).

Such brief portrayals of violence may well appear unjustified given that viewers are unlikely to have a context to judge its justification and may therefore assume that the violence is initiated by the violent character rather than being an act of self-defense. Consistent with this interpretation, activation in lOFC has also been observed when people imagine themselves attacking innocent persons in a video game (Molenberghs et al., 2016).

Given the oft-observed activation in lOFC in response to unjustified violence, we hypothesize that when a character engages in violence without a socially justified motive, attention focuses on the harm inflicted on the victims and vmPFC decreases in activity. At the same time, lOFC responds reflecting disapproval of the unjustified behavior. This response parallels the role of the vmPFC in that lOFC tends to react more strongly to aversive events (Berridge and Kringelbach, 2013Rolls, 2015) while vmPFC responds more to reward. Thus, the motives of the characters engaged in movie violence may determine which region of ventral PFC is elicited.

In our study, we exposed participants to both justified and unjustified film violence so that we would be able to detect differences in vmPFC versus lOFC response to these different forms of violence. Our prediction is that justified violence will elicit significant response in vmPFC while unjustified violence will elicit significant response in lOFC.

In order to test these predictions, we exposed late adolescents who frequently watch violent entertainment to examples of scenes from popular movies that involved either justified or unjustified violence. We chose this population in order to observe the effects of violent movies on viewers who are most likely not to be emotionally disturbed by exposure to violent content.

At the same time, our research with adults (Romer et al., 2018) showed that repeated exposure to justified movie violence enhanced the acceptance of such violence relative to unjustified violence and therefore could maximize our ability to observe the same phenomenon in young people.

The scenes were selected based on the criterion that justified violence was an acceptable response to prior aggression or wrongdoing (i.e., defense of self or others), and that unjustified violence had no apparently acceptable motive.

These characteristics were verified by independent ratings (Romer et al., 2018). For each movie, we first showed the characters engaged in interaction prior to violence followed by a clip of the violent scene. Our reasoning motivating this experimental manipulation was twofold.

First, we aimed to enable viewers to begin an evaluation of the motives of the characters and second, we aimed to enable a comparison of brain response between the scenes with and without violence that nevertheless involved the same characters.

In comparing the models, we used intersubject correlation (ISC) to identify the degree to which regional brain activity became synchronized across viewers in response to movie viewing (Hasson et al., 2004).

ISC is a model-free approach to fMRI time series data that uncovers similarities of response across viewers without the need for comparisons with a control condition (Wilson et al., 2008).

ISC is particularly suited to studying movies, which require sensitivity to neural activity over relatively long-time windows of narrative content rather than to short-term changes in sensory stimulation (Hasson et al., 2010Pajula et al., 2012). The approach is also informative for exploring how differences in the violent content of popular movies are perceived by audiences.

The approach has also been shown to be more sensitive to the narrative properties of a movie than to its surface features (Pajula et al., 2012Nguyen et al., 2019) and to induce synchrony in brain regions that are sensitive to narrative content (Nummenmaa et al., 2012Schmälzle et al., 2013).

If the motives of the characters in violent movies determine brain responses as predicted by a virtue ethics approach, there should be differential synchronization of lOFC and vmPFC, with significant synchronization of vmPFC for justified violence and lOFC for unjustified violence.

Media Contacts:
Press Office – Frontiers

Original Research: Open access
“Intersubject Synchronization of Late Adolescent Brain Responses to Violent Movies: A Virtue-Ethics Approach”. Azeez Adebimpe, Danielle S. Bassett Patrick E. Jamieson and Daniel Romer.
Frontiers in Behavioral Neuroscience doi:10.3389/fnbeh.2019.00260.



Please enter your comment!
Please enter your name here

Questo sito usa Akismet per ridurre lo spam. Scopri come i tuoi dati vengono elaborati.