Violent entertainment has made it into the public discourse due to rising concerns about the graphic nature of highly popular video game franchises including Grand Theft Auto (GTA) and Call of Duty. But what about violence in films which are enjoyed by a much larger and more diverse audience? After the R Rating was established by the Motion Picture Association of America (MPAA) film rating system in 1968, there was an increase in violent content in films thereafter.
The relationship between media and culture is bidirectional, making it important yet challenging to disentangle and analyze their effects. On one hand, movies can shape cultural tastes, beliefs, attitudes, and behaviors. For instance, numerous studies have shown that media violence can lead to desensitization to violence, reduced empathy for victims, and negative mental health impacts such as increased fear and anxiety in children. On the other hand, movies often reflect societal realities, including the prevalence of violence. To better understand this bidirectional dynamic, a crucial question arises: how is the portrayal of violence in media changing over time?
To further explore this relationship, Computational Social Science Lab (CSSLab) Postdoctoral researcher Amir Tohidi (University of Pennsylvania), Babak Fotouhi (University of Maryland), Rouzbeh Touserkani (Institute for Research in Fundamental Sciences), and Brad J. Bushman (The Ohio State University) analyzed comprehensive film data and found that violent language in the film industry has increased over time. Their findings have been featured in several outlets, including The Guardian, EurekAlert, and The Daily Telegraph.
The study took place from September 2022 to September 2024. During this time, researchers used movie dialogue from a total of 166,534 films that were released between 1970 to 2020, extracting murderous words, including kill and murder.
The study considered only murderous verbs used in active construction (such as “She killed X.”). They didn’t count passive constructions (“He was killed by X.”) or negations (“She didn’t kill X.”) or questions (“Did she murder X?”), or other forms of violence such as shooting or stabbing.
“We designed this to be a conservative estimate,” Fotouhi says. “It is likely there was more violence in the movies than what we calculated in terms of the dialogue.”
One reason this can be considered a conservative estimate is the narrow focus of the study, says Tohidi.
“We focused exclusively on murderous verbs in our analysis to establish a lower bound in our reporting, and avoiding discussions about what constitutes violence and what does not” Tohidi notes. “Including less extreme forms of violence would result in a higher overall count of violence.”
They then used natural language processing techniques to match the murderous dialogue to the characters; the analysis revealed that 6.97% of these films contained violent language, showing that dialogue involving characters murdering others is increasing over time- not just limited to the crime genre but across all film categories.
While other studies have shown increases in film violence, the value of this analysis was the huge dataset, the researchers say. No other study has examined such a massive number of films.
Violence is often incorporated into films by producers as it appeals to younger audiences and boosts profits, but this comes at the expense of the wellbeing of children and adults. As this kind of content becomes more commonplace, it becomes a greater concern for many people including parents, healthcare workers, and policymakers. The effect of films on culture is challenging and controversial so further studies on the impact of film dialogue on society and individual behavior would provide insights into addressing harm reductions in media production and consumption.
AUTHORS
DELPHINE GARDINER
–
Communications Specialist