Guns are a virus. But not in the way people think.
I’ll let you in on a content strategy secret that we haven’t shared before. If your goal is persuasion, you have to understand the mental frame of the people you’re trying to persuade. And while we spend most of our time on gun-curious people, sometimes we think about what gun control advocates are thinking. Here’s the secret. We have a simple heuristic to model their behavior: think of guns as a virus. If guns were a virus, how would you treat them? What would you say about them? How would you model exposure to them?
Probably something like this:
Or this:
Or this:
Dear Hasbro Board of Directors:
As the holidays approach, we are reminded of our mission to protect the safety of children. As we watch holiday toy commercials, we see the Nerf Ultra One and other extreme Nerf machine guns for children and are reminded of mass shootings that have devastated American children and families for decades now.
…
Corporate social responsibility is not a slogan; it is what calls you to raise the bar in the interests of children and become a source for the non-violent creative playthings children deserve. Marketing assault weapon toys to this most vulnerable group of consumers is an assault on their dignity and their worth as human beings.
We implore you to remove assault-style toy weapons from your product offering.
Or this:
A safer society for my taste has fewer guns and not more guns. The guns that do exist are at the hands of the right people, particularly trained members of law enforcement.
– New Jersey Governor Phil Murphy explaining his decision to effectively ban gun sales in the state during the COVID shutdowns. (He reversed that decision under pressure five days later and reopened gun stores.)
These are familiar points, so I’m not including them here to say they’re surprising or unusual or new. But it’s interesting to think about what these very standard gun control views represent: the lens is epidemiology.
The epidemiological frame is so ubiquitous (and appeals so strongly to intuition) that people build entire careers on this stuff without realizing that there are other frames. There are — sociological, cultural, moral, philosophical, etc. — but seeing them requires coming at the issue with a painstakingly open mind.
In the guns-as-virus epidemiology frame, the presumption is that guns are inherently noxious. We wrote a whole article about what that belief misses, but for brevity, we like the way the sociology professor David Yamane describes the blind spot:
A quarter-century ago in 1995, sociologist James Wright included among his “Ten Essential Observations on Guns in America” that “gun ownership is normative, not deviant, behavior across vast swaths of the social landscape”. The idea that guns are normal and normal people use guns may seem common-sense to those of us gathered here, but it’s actually a dramatic departure from the standard social scientific approaches that view guns and gun owners as deviant, and research literatures that are dominated by criminological and epidemiological studies of gun violence. This theme is so constant that the New York Times ran a headline just last week declaring, “Gun Research Is Suddenly Hot”. In fact, the story was about how research on gun violence is suddenly hot. Research on the lawful use of guns is as cold as ever.
The idea is intuitive: guns cause death. Fewer guns means less death. So if you can stop guns from spreading, you stop people from dying. That’s how diseases work too, so there’s a ready answer: treat guns like you treat a virus. Get the CDC and university labs to study it. Reduce exposure. Make it hard for people to spread it to others. And keep the pressure on, stamping out the virus wherever you see it, until eventually it dies out.
Now, we (unsurprisingly) don’t agree with that framing*, but we like engaging uncomfortable ideas. If you play with them long enough, sometimes something interesting falls out.
So let’s try it. Let’s model guns as a virus.
(*It’s out of scope for this article, but if you’d like to learn a few of the reasons we don’t agree: there’s a surprisingly tiny correlation between gun ownership and murder rates, what little correlation there is vanishes when you look at non-domestic murders (which oddly drive ~all legislative discussion), unrelated factors have much more explanatory power for rates of violence, and the entire debate matches a culture war that hurts everybody much more than it matches a good-faith discussion that people would actually benefit from engaging in.)
Going viral
And lookee here, there’s immediately something interesting: viruses are a surprisingly apt metaphor because, like guns, they spread without even being alive. A virus can’t reproduce on its own. It’s just some genetic material and a little coat of protein. It needs to get into a host cell, and then it relies on the host cell’s own inner workings to create and spread new copies of the virus. Pretty crafty.
Guns and viruses aren’t alone in this, of course. Internet memes are a fine (albeit imprecise) example, and they’re actually named for it. In 1976, Richard Dawkins invented the term “meme” in his book The Selfish Gene. His idea was that genes are to biology as memes are to culture — entities that “seek” to spread and replicate:
But do we have to go to distant worlds to find other kinds of replicator [than genes] and other, consequent, kinds of evolution? I think that a new kind of replicator has recently emerged on this very planet. It is staring us in the face. It is still in its infancy, still drifting clumsily about in its primeval soup, but already it is achieving evolutionary change at a rate that leaves the old gene panting far behind.
The new soup is the soup of human culture. We need a name for the new replicator, a noun that conveys the idea of a unit of cultural transmission, or a unit of imitation. “Mimeme” comes from a suitable Greek root, but I want a monosyllable that sounds a bit like “gene”. I hope my classicist friends will forgive me if I abbreviate mimeme to meme. If it is any consolation, it could alternatively be thought of as being related to “memory”, or to the French word meme. It should be pronounced to rhyme with “cream”.
Examples of memes are tunes, ideas, catch-phrases, clothes fashions, ways of making pots or of building arches. Just as genes propagate themselves in the gene pool by leaping from body to body via sperms or eggs, so memes propagate themselves in the meme pool by leaping from brain to brain via a process which, in the broad sense, can be called imitation. If a scientist hears, or reads about, a good idea, he passes it on to his colleagues and students. He mentions it in his articles and his lectures. If the idea catches on, it can be said to propagate itself, spreading from brain to brain. As my colleague N. K. Humphrey neatly summed up an earlier draft of this chapter: “… memes should be regarded as living structures, not just metaphorically but technically. When you plant a fertile meme in my mind you literally parasitize my brain, turning it into a vehicle for the meme’s propagation in just the way that a virus may parasitize the genetic mechanism of a host cell. And this isn’t just a way of talking—the meme for, say, ‘belief in life after death’ is actually realized physically, millions of times over, as a structure in the nervous systems of individual men the world over.”
The other ubiquitous modern expression of the concept is in marketing and social media, when things go viral. The concept itself is ancient, but that term is surprisingly recent:
The earliest known use of the term “viral marketing” is in a 1989 PC User magazine article about the adoption of Macintosh SEs versus Compaqs:
“At Ernst & Whinney, when Macgregor initially put Macintosh SEs up against a set of Compaqs, the staff almost unanimously voted with their feet as long waiting lists developed for use of the Macintoshes. The Compaqs were all but idle. John Bownes of City Bank confirmed this. ‘It’s viral marketing. You get one or two in and they spread throughout the company.’”
So viral marketing initially denoted seeding designed to kick-start the copycat effect whereby people “catch” the idea and adopt it by seeing it adopted by others.
(By the way, big shout out to John Bownes over at City Bank. Whoever you are, apparently you were the first person in history recorded using the phrase “viral marketing”. Congrats.)
This seems like a fertile place to focus: the public health approach casts guns as a virus in the disease sense, but in doing that, it sells its own metaphor short. You miss all the complexity of a virus if you only look at is its effects. (And on guns, effects are a worn out debate that at this point it’s just boring to read object-level opinions about.) The most interesting thing about viruses isn’t what they cause — it’s how they spread. So let’s really embrace the richness of the viral metaphor. Let’s talk about contagions spread.
A meme’s levers
The first concept here is R0, or the basic reproduction number. This describes how many new infections each person with a virus will cause in a population where nobody’s immune. So if R0 is 2 and I have the virus, I’ll give it to two people, each of them will give it to two more, and the population who has it will be 2# of steps. That grows surprisingly fast: after 10 steps, there’ll be 210, i.e. 1024 infected. Ten more steps, and 220 is 1,048,576.
That’s interesting as far as it goes, but it’s not the whole story. For one thing, 233 is more than the human population of Earth, and since no meme or virus, no matter how viral, gets quite that big, there must be some sort of system-level brakes.
The writer and software engineer Kevin Simler wrote an interactive essay, “Going Critical”, to illustrate the dynamics of this visually. I strongly recommend you take the time to play with Kevin’s essay. It’s built around interactive animations of an outbreak spreading through a population, and you can adjust the variables to change the outbreak’s dynamics right before your eyes.
But I’ve pulled out some examples — which I promise are going to wind back around to being gun-related — that illustrate the key levers.
First, let’s look at a basic Susceptible–Infected–Susceptible (SIS) model. The population starts out as susceptible with a small group of infected, the infection spreads, and when people heal they go back to being susceptible. The alternative is a Susceptible–Infected–Removed (SIR) model, where after people recover, they cannot be reinfected. That matches biological viruses well (because people develop antibodies after being infected), but exposure to cultural memes can trigger “antibodies”, the opposite (i.e. increased future susceptibility), neither, or both. Basically, anything can happen. So for memes — and gun culture is a quintessential meme — SIS is a better fit.
Here’s an SIS model with a 25% transmission rate. I.e. each infected node has a 25% chance of transmitting the meme to a neighboring node (above, below, left, right).
Here’s how the same exact outbreak plays out if the transmission rate is 22% instead of 25%.
Ok, so there’s an important lever! Transmission rate alone can change an outbreak from supercritical to subcritical. Let’s see what else matters.
To model cultural spread, we have to account for spontaneous activation. That’s where someone “catches” a meme on their own, not from a neighbor who already has it. SIS plus spontaneous activation is called SISa. (All of this is in Kevin’s essay, and again, I’d encourage you to spend some time playing with the graphs there.) Here’s a SISa outbreak with a 25% transmission rate and a 33% spontaneous activation rate.
Spoiler alert: spontaneous activation alters the path that an outbreak happens to take through a population, but actually won’t swing a subcritical outbreak to supercritical (or vice versa). Transmission rate is what matters. But we’ll stick with SISa anyway, just because that’s more what real life looks like for memes.
Now let’s add in the concept of immunity. For memes, like viruses, a certain portion of the population will just be immune to catching it. Here’s a population with a 25% transmission rate, 25% spontaneous activation rate, and 5% immunity rate.
Ok, still supercritical. What if we turn the immunity rate up to 20% and keep everything else the same?
The outbreak keeps popping its head up, but every time, that 20% of the population that’s immune is enough of a firebreak that the meme peters out. Firmly subcritical.
There’s one last thing to add to the model: degree. So far, we’ve assumed that each node has four neighbors to potentially spread memes to — i.e. a degree of four. In real life, that number is highly variable. So let’s see what happens when we change it. What if we keep everything the same as the previous network — 25% transmission rate, 25% spontaneous activation rate, and 20% immunity rate — but change the nodes from having a degree of four to a degree of six (i.e. six neighbors).
Whoa, we’re supercritical again, completely blowing through the brakes that the 20% immunity rate is trying to apply. Degree matters a lot. For our last network (and one last time: this is all Kevin’s thought and modeling, not mine), let’s approximate real life with a network that has varying degrees within the same network. In any population, there will be high-degree, dense areas (e.g. cities, tight-knit internet subcultures, etc.) and low-degree, sparse areas (e.g. people who don’t communicate much with others). Let’s see what that looks like, with a 25% transmission rate.
The outbreak chugs along and then it hits a high-density patch, accelerates, and blazes through. What this feels like when you’re in the population is that one day you’ve never heard of the meme (or, given current events, the virus), and in the blink of an eye, it’s everywhere. That non-linearity matters, because it means that today’s state tells you nothing about what tomorrow’s state will be.
For a virus, that’s bad. But for a meme you want to spread, it’s an unfairly powerful button to push. And I’m just going to say right up front that, surely, if we have a big red button that’s unfairly powerful, we shouldn’t push it. It wouldn’t be right. We should resist that urge. Because, well, I mean, ok, if maybe we sidle up and definitely don’t push the button but just get a bit closer, and maybe just a little…
Oh wait, right, this was about guns
Ok, so we leaned into guns-as-virus, and for our trouble, we’ve now got a solid foundation in memetics and contagion, plus a basic list of the levers that really matter: transmission rate, immunity rate, and how many contacts each node has. Let’s analyze each one.
Transmission rate
For any given node-to-node relationship where Node A has caught the meme — i.e. has “caught” gun culture (this can mean being a gun owner, or just engaging with gun stuff on YouTube, reading or writing about it, supporting gun rights groups, participating on the legislative side, whatever) — and Node B hasn’t, what are the odds that A will transmit the meme to B?
In other words, how effective are you as a transmitter? If you’re intimidated out of being a transmitter, or ashamed of being “out” as a carrier of the meme, that will reduce the transmission rate. If in attempting to be a transmitter, you drag in polarizing political baggage, that’ll kneecap your effectiveness and reduce the transmission rate. (The trap here is that politics will actually make you more effective among the subset of people who happen to agree with you. But your seemingly effective strategy will suddenly backfire you as soon as you try to expand outside your bubble.)
Legal and regulatory hurdles can also reduce transmission rate. If buying a gun requires hundreds of dollars in fees, or multiple trips to the gun shop, or intimidating legal trapdoors where the wrong-shaped grip changes a rifle from being legal to being a felony, that will all brake transmission.
This explains the confusingly specific battles people have on this stuff. It’s why, for example, gun control groups got Citigroup to ban businesses that sell 30-round mags, and YouTube to ban videos on how to build an AR. For brands like Citigroup and YouTube, allowing those items normalizes them. A ban makes them unclean. It’s about normalization and stigma — i.e. increasing or decreasing the meme’s odds of jumping from Node A to Node B.
The gun control tribe knows that, which is why they lobbied for the bans. The gun rights tribe also knows it, which is why gun YouTubers and gun Twitter were apopleptic about it. (“Tribes” as in the classic Slate Star Codex piece on outgroups.)
Immunity rate
You can think of immunity as an asterisk on transmission rate: “With respect to this specific node, the transmission rate is 0.” And a critical mass of immune nodes — i.e. herd immunity — has the second-order effect of creating an impassable firebreak in the network, making it impossible for the meme to move from one set of carriers to other, entirely susceptible nodes.
Like transmission rate, this is also about normalization and stigma. The easiest illustration is the silly-seeming Nerf gun example from the beginning of this essay. Why would a nonprofit bother complaining that Nerf guns are, essentially, too cool? Because if you want to increase immunity to the gun culture meme, you need to introduce a stigma around it. And if you understand path dependence, then you know a childhood stigma is both easier to instill and harder to dislodge. And the same is true of normalization. That’s why people who grew up where guns are normal tend to think guns are normal, and people who grew up where guns are haram tend to think guns are haram. So for the stigma to have a chance of sticking, you need to make sure that people are never exposed to guns.
The effect isn’t symmetrical though. The stigma is much easier to dislodge than the normalization is. That’s because all that’s required is exposing people to positive gun ownership. And positive gun ownership is the overwhelming norm — there are 423 million guns in the US, and each year roughly 35,000 are involved in a suicide, murder, or accidental death. Let’s round that up wildly to 100,000 to capture nonfatal negative uses. So that leaves 99.9763593% of guns being used positively. In pithier terms: unless you’re taught otherwise, guns are cool.
This illustrates the difficulty of maintaining immunity to the gun meme. Allowing people to be exposed to gun culture is a non-starter, given our back-of-the-napkin math is that 99.9763593% of gun use is positive (or, more precisely, “innocuous and possibly positive, but measuring just how positive is out of scope for this essay so we technically can only say ‘innocuous’”). If someone’s exposed to guns, they’re statistically almost certain to be exposed to positive uses. So the only pro-immunity option is to not let them be exposed at all. Which is how people end up promoting ideas like “We can’t let kids have Nerf guns that are too much fun.”
It follows, then, that the meme’s pro-transmission, anti-immunity path is simply more exposure. The outbreak math will take care of the rest.
That’s why, for example, we have a section on our site about how to take a newbie shooting, and why we write so much about the importance of producing high-quality gun content. It’s why video games and movies matter. It’s why production values matter. It is pure cosmic luck that exposing people to gun culture tends to make people enjoy gun culture. As our friend Jon Hauptman of PHLSTER Holsters puts it, gun people get to have 100% of the fun — it’s just plain fun to go shooting. We didn’t do anything to make that true, it’s just the way the universe shook out. And it happens to be extremely powerful, as demonstrated by the contagion animations earlier in this essay. So it would be foolish not to bake this into our approach.
How many contacts each node has
We saw above that network density can blow through the otherwise-impassable hurdles that immunity rates put up. There are a couple lessons to take from that.
First, there are individual-level gains from growing your network. A bigger network increases your effective transmission rate, especially in tough times. It also feels nice. (Ok, somehow this essay has wound itself around to the profound insight that having friends is a good thing. Thanks for coming to my TED Talk, everyone.)
Second, remember the final animation, where the outbreak blazes through high-density patches in the network? That matters not just for the high-density patch; it also makes nodes in the low-density patches more likely to catch the meme. The lesson, then, is simple: build high-density patches. And again, the gun meme benefits from the happenstance of existing in modern times: the internet is the most powerful tool ever created for building high-density networks. The networks are easier to build than ever, possible to build denser than ever, and possible to build bigger than ever. So we’ve got an exponentially growing set of networks (because the internet itself is a contagion with outrageously high R0) that are unprecedentedly numerous, dense, and large. If your goal is “build high-density patches”, that’s very good news. If your goal is to put brakes on the contagion, it’s … well, it’s bad.
Takeaways
We started out with a kneejerk aversion to the guns-as-virus epidemiology lens, because it’s a textbook begging-the-question fallacy. The standard view focuses exclusively on disease, i.e. the harms of guns, missing all the truly interesting stuff that the virus lens reveals.
By being open to uncomfortable ideas and embracing the guns-as-virus metaphor more fully than its originators do, we’ve found some unexpected fruit: a treasure map on how to spread gun culture, and an understanding of the science behind the map. If you understand how and why a system works, you’re in a good spot to modify its outcome. So increase your transmission rate, build a positive culture that makes people receptive to the meme, and grow your network. The math will take care of the rest.