The Weapons Effect – Does the Mere Presence of a Gun Increase Aggression?

Anime girl with a gunA provocative question indeed: Do we seeing lots of gun deaths in America because of people’s aggression? Or is it possible that aggression is actually increased because there are so many guns in the country? The “weapons effect” is the theory that just by being in the mere presence of weapons, aggression levels increase. It’s a controversial theory, but there is of course a lot of research that has been published on it. In this article, let’s look at the studies that shed light on this area, and I will give my take on the research literature.

The Mere Presence of a Gun

Hormones and Aggressiveness

Let’s start in 1998, when researchers Jennifer Klinesmith, Tim Kasser, and Francis McAndrew measured the effects of interacting with an object on testosterone levels and aggressive behavior. Testosterone is of course linked to aggression, and this study found what it expected to find:

We tested whether interacting with a gun increased testosterone levels and later aggressive behavior. Thirty male college students provided a saliva sample (for testosterone assay), interacted with either a gun or a children’s toy for 15 min, and then provided another saliva sample. Next, subjects added as much hot sauce as they wanted to a cup of water they believed another subject would have to drink. Males who interacted with the gun showed significantly greater increases in testosterone and added more hot sauce to the water than did those who interacted with the children’s toy. Moreover, increases in testosterone partially mediated the effects of interacting with the gun on this aggressive behavior.

So just the brief interaction with the gun resulted in higher testosterone levels and more aggressive behavior than with the toy. And if that’s just a simple interaction, imagine what this might mean for someone near a gun who was actually provoked. This suggests that simply passively interacting with a gun has a testosterone-raising, aggression-inducing quality. A 2005 study by Bartholow, Anderson, Carnagey, and Benjamin, looked into this research, saying “Recent research (Anderson, Benjamin, & Bartholow, 1998) indicates that the presence of guns increases the accessibility of aggressive thoughts via automatic priming. Our research examined whether this ‘weapons priming effect’ differs depending on the structure of an individual’s knowledge about guns, and if so, whether that difference results in corresponding differences in aggressive behavior.” For each of Bartholow and colleagues’ three experiments, they used male introductory psychology students at a large midwestern university. Here are their findings:

Experiment 1 revealed that individuals with prior gun experience (hunters) have more detailed and specific information about guns than do individuals with no direct gun experience (nonhunters), and that hunting experience interacts with gun type (hunting versus assault) in predicting affective and cognitive reactions to guns.

Experiment 2 revealed that pictures of hunting guns were more likely to prime aggressive thoughts among nonhunters, whereas pictures of assault guns were more likely to prime aggressive thoughts among hunters. Experiment 3 showed differences in aggressive behavior following gun primes that correspond to differences in affective and cognitive responses to gun cues.

The main takeaway from this is that people who are familiar with guns get aggressive thoughts more from assault weapons than from hunting rifles. The fact that the type of gun that experienced gun users feel more aggressive when seeing (i.e., assault weapons) is better at killing people than a hunting gun does not bode well for the US. It’s no wonder that, despite the lucrative business of assault weapons, the majority of Americans want to see assault weapons off the shelves and out of public reach. That is, Americans are beginning to understand what science has suggested all along. So let’s talk science.

What’s the Deal with Aggression Studies?

In many aggression experiments, like the 1998 one above, participants are often measured for their malicious behavior on a certain task. For example, participants may do things like blast painfully loud noise into another room with someone else inside, or do some other relatively harmless but momentarily negative behavior (in actuality, no one is there to receive this punishment, but the participants is lead to believe so). So that could mean, for example, the participant thinks they are blasting noise, and the degree to which they intensify that noise is called “aggression.” There are of course problems with this – the obvious one being that blasting noise doesn’t necessarily translate to dangerous behavior outside the laboratory setting. But the reason for these types of procedures is because of ethical considerations. If we didn’t have to worry about ethics, we could conduct the “best” study in the world.

For example, give people guns, and see how far you can provoke someone until they pull the triggers. What factors effect that? How much does prior experience matter?… Well, there’s no point in even thinking much about this, because such a study will never happen – ethical guidelines are too important. Therefore, this is the best we have. Otherwise, you get research such as the Dutch shooting club paper of 2009, responsibly covered by the British Psychological Society (BPS). It was reported that members of a shooting club were actually less aggressive than an age-matched controls who were not affiliated with a shooting club.

Naturally, this study has several problems (one of which the BPS didn’t mention). For one thing, we don’t know why the aggression was rated as lowered, but it’s certainly likely that another factor (unrelated to guns) was the true cause of the low aggression measurements. For example, as BPS quickly mentions that “most of them became a member for relaxation, or to socialise.” It’s certainly possible that the real cause of the lowered aggression was simply because the club members were part of a club in which they could socialize in a community of people with similar interests (or at least one similar interest); or that another factor caused this altogether.

Furthermore, the study relied entirely on self-reports, which is notoriously unreliable, especially if the subjects have something to prove. The researchers tried to keep results valid by manipulating the questionnaires (in a good way; see the BPS article for more), but you can probably see the problem with such research. So you might be tempted to argue that such studies are totally useless. I agree that they are in large part meaningless on their own, but science doesn’t work by saying “we proved this by getting these results.” It works by reviewing the research in the context of other studies that dealt with the same issue, and how convincing their evidence is. This brings us to a classic study on the “weapons effect.”

Does the Gun Pull the Trigger?

The Classic Study

Sometimes the most interesting evidence comes from the classic studies. The first time I ever understood what an “interaction effect” was came from this research by Leonard Berkowitz and Anthony LePage from 1967. I’m not sure if this would have passed an ethical review board today (ethical standards were not nearly as strict back then as they are now) because in this experiment, participants actually did receive (albeit mild) shocks:

An experiment was conducted to test the hypothesis that stimuli commonly associated with aggression can elicit aggressive responses from people ready to act aggressively. 100 male university students received either 1 or 7 shocks, supposedly from a peer, and were then given an opportunity to shock this person. In some cases a rifle and revolver were near the shock key. These weapons were said to belong, or not to belong, to the available target person. In other instances there was nothing near the key, while for controls 2 badminton racquets were on the table near the key. The greatest number of shocks was given by the strongly aroused Ss (who had received 7 shocks) when they were in the presence of the weapons. The guns had evidently elicited strong aggressive responses from the aroused men.

So the researchers manipulated a few things: How many unpleasant shocks each participant received (to anger them) and the stimuli that was present in the room (guns vs. badminton racquets). The variable the researchers were interested in was how many shocks the subjects attempted to give in return. In general, neither the seven shocks nor the “aggressive stimuli” (gun) had caused the participants to try to shock their peer. However, when the two of those variables were together (i.e., when they interacted) the subjects were indeed more likely to try to return the shock to their peers. This makes a lot of sense when you think about aggression. Sure, some people may “snap” because something or someone triggered their aggression (like seven shocks), but it’s more likely that there was a predisposing factor (like the gun in the room), without which the trigger would not have had any effect. As the authors wrote in their article:

It has been observed… in field settings as well as the psychological laboratory…. It is clear that the presence of a weapon—or even a picture of a weapon—can make people behave more aggressively. In essence, the gun helps pull the trigger.

But there are several problems with this research. At least, that’s according to Paul Gallant and Joanne Eisen, two Senior Fellows at the Independence Institute – a conservative think tank.

Harsh Critics

Gallant and Eisen published a very well-researched paper in 2004, arguing, among other things, that replications of Berkowitz and LePage’s research have not been met with consistent findings. In fact, the meta-analysis they cite to make this point says that the weapons effect was “true for the overall analysis and for name-mediated cues, but confirmation of the weapons effect was restricted to cases wherein subject sophistication and evaluation apprehension were low.”   (((やっぱまだ駄目だ。。。)))   Indeed they claim that some replications contradicted their results, but I haven’t seen a study that has done so. They also, in the following excerpt, rebut what Berkowitz and LePage said, and then dismiss a later study.

The authors, however, cited no evidence to support their claim that the weapons effect has been observed “in field settings,” or that people behave more aggressively in the presence of a weapon outside the artificiality of a laboratory setting. They simply identified research that was done in the outdoors instead of in a laboratory room. In a 1975 experiment, for example, Turner, et al, had a confederate stop his pick-up truck at an intersection and remain stationary after the traffic light had turned green. The subjects in the experiment were the drivers stuck behind the immobile truck. Sometimes the truck had a rifle mounted on a gun rack plainly visible to the driver of the car, and sometimes there was no rifle present. Aggression was measured by the amount of horn-honking that ensued when the driver of the car was unable to proceed. While this may technically be classified as an experiment conducted in a “field setting” since it was conducted outside of a conventional laboratory, to claim that the outcome had any relation to the “weapons effect” is scientific sleight-of-hand and outright misrepresentation. All the horn-honking in the world does not translate to a real-world manifestation of the “weapons effect” as defined by Berkowitz and LePage

They do make a very good point at the end, despite neglecting to mention that the results of that 1975 study were consistent with Berkowitz and LePage’s findings. Nevertheless, they write a conclusion that suggests the weapons effect is bogus:

Researchers who say they are “measuring aggression,” and then perform a bait–and–switch to redefine what they measured as a “weapons effect,” are fooling themselves. Having aggressive thoughts does not translate into the lethal kind of “weapons effect” that Berkowitz and LePage hypothesized.

A new generation of weapons effect proponents would have us believe that ordinary American gunowners are like Pavlov‘s dogs learning to salivate upon hearing a bell: put them near a gun, and they will shoot themselves or some other innocent. As Leonard Berkowitz put it, “Gun control may not be too effective in protecting ordinary citizens against criminals or Presidents against assassins, but it may, nevertheless, save some ordinary citizens from other ordinary citizens like themselves.”

Such a profoundly pessimistic view of human nature presumes that most of us are incapable of controlling our actions. If we are to believe that simply seeing a firearm will cause us to think about murdering one another and make us more likely to commit the act, we must also concede that we are gravely lacking of free will—mere slaves to our environment—and that we can easily and completely be dominated by mind-control tactics like subliminal advertising and frenzied propaganda. Doesn‘t mankind deserve more credit?

So they do make an excellent point – is this really aggression? And even if it is, does that really translate to the same types of killing sprees that worry us today? But they also seem to be largely missing the point of the interaction effect. Their opening line of the article is “The ‘Weapons Effect’ hypothesis suggests that guns can psychologically control people and cause them to be violent.” Well… that may be true, but only to the extent that it’s only one factor among many complex factors.

They also pull up some straw men arguments, suggesting that Berkowitz and LePage, were saying that seeing a firearm will cause us to think about murdering one another. They clearly said “the presence of a weapon [. . .] can make people behave more aggressively.” Nothing about seeing a firearm will cause us to think about murder. Furthermore, the tone of their concluding paragraph seems to betray their bias, because a) no one is suggesting that people are utter slaves to their environment, completely hypnotically dominated by mind-control from inanimate objects; and b) subliminal influence is real, and the environment affects behavior, emotions, and even perception, all the time; and we are generally oblivious to it. The notion that we have complete free will has been contested and largely dismissed by scientists and philosophers alike. That doesn’t mean that most people believe we have no free will (a much smaller number of scientists take that position, and I am not among them) but clearly, the environment influences us without our awareness.


The most well-researched paper I could find on the subject suggested that the evidence to support the Weapons Effect is weak, to the extent that it doesn’t have much to say about generalization outside the laboratory. This is basically true, and we have to be careful when looking at what aggression really is. However, this does not mean that we should dismiss all of the research as useless – it’s all a matter of context. So let’s talk about the context.

The 2004 study was done by a conservative think-tank. Conservative Americans are more likely to be gun owners and want to protect their rights to own guns. They are also, according to research published in 2012 by Brittany Liu and Peter Ditto, more likely to spin reality and make up facts to suit their arguments. This all makes sense when you see gun-proponents arguing that the research on guns is bogus, and in the same breath, arguing that the real causes of gun violence are violent TV, movies, and especially video games – which they say teaches children to kill by having them act out violence. By the way, America’s National Rifle Association – the world’s biggest, loudest gun-proponent organization – recently released their very own gun-wielding first-person-shooter video game. It is recommended for children aged four and up. We can only assume the game is called “Call of Hypocrisy.”

The Bottom Line

As I mentioned above, we’ll never get the “perfect” study on aggression and guns, because it would never be ethical to conduct it. But if you’re only going to throw away the research that refutes your argument, you’re not really debating in the first place. All the research on aggression should be considered, from guns, to blasting sound, to playing violent video games. So is the weapons effect real? It’s hard to say. The research does seem to suggest that there may be cultural associations we have psychologically made with guns and aggression. It’s possible that, for example, movies – in which guns are used because they are a simple story-telling tool to elicit action in a story – have reinforced the notion that when you see a gun onscreen, it will be used. And that usually means aggression. But an important question is just what does aggression mean? The answer will probably influence the definition of the “weapons effect.” But regardless of whether or not it’s because of an increase in aggression due to the mere presence of a gun, or other factors such as the perception of a lack of security, the simple truth is that gun ownership is negatively correlated with safety.

This entry was posted in Culture, Legal Issues, Media, Psychology, Science, Technology. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *