The human need for 'chaos': Here is why extremists spread conspiracy theories and fake news

“Some men aren't looking for anything logical, like money. They can't be bought, bullied, reasoned, or negotiated with. Some men just want to watch the world burn.”

– Alfred, Batman’s butler, in "The Dark Knight," explaining the Joker’s motivations

A new paper, recently presented at the American Political Science Association’s annual convention, suggests a widespread motive driving  people to share fake news, conspiracy theories and other hostile political rumors. "Many status-obsessed, yet marginalized individuals experience a 'Need for Chaos' and want to 'watch the world burn'," lead author Michael Petersen tweeted, announcing the availability of a preprint copy.

This article first appeared on Salon.

Truth, in such a worldview, is beside the point, which offers a new perspective on the limitations of fact-checking. The motivation behind sharing or spreading narratives one may not even believe can help make sense of a variety of threatening or confusing recent developments in advanced democracies. It also sheds light on disturbing similarities with outbreaks of ethnic or genocidal violence, such as those seen in Rwanda and the Balkan nations during the 1990s.

Petersen is a professor of political science at Aarhus University in Denmark, where he directs the Politics and Evolution Lab. Together with Mathias Osmundsen of the Aarhus Institute of Advanced Studies and Kevin Arceneaux, a political scientist at Temple University, Petersen studied attitudes in both Denmark and the United States.

While the focus of the research was on individual actors, orientations toward group identities played a central role as well, and a connection was per with broader underlying conditions. The importance of groups can be seen in two scales used to draw one of the most important distinctions the paper explores: support for legal vs. violent activism. The former relates to commonly recognized phenomena, such as partisan polarization, while the later relates to the "Need for Chaos."

Both scales are based on agreement or disagreement with a number of statements. For legal activism, these were statements like “I would follow a group on social media [or] become a member of an organization [or] donate money to an organization that fights for my group's political rights and interests.” For violent activism, these were statements such as “I would continue to support an organization that fights for my group's political rights and interests even if the organization sometimes breaks the law [or] sometimes resorts to violence,” and “I would go to war to protect the rights of my group.”

There’s also a "Need for Chaos" scale, which shows far less evidence of connection with a group, with statements like “I think society should be burned to the ground,” “We cannot fix the problems in our social institutions, we need to tear them down and start over,” “When I think about our political and social institutions, I cannot help thinking ‘just let them all burn,’" and “Sometimes I just feel like destroying beautiful things.”

The phrase "need for chaos" especially drew my attention because it resonates with the findings of what is called "structural demographic theory" or SDT, as laid out in Peter Turchin’s 2016 book, “Ages of Discord: A Structural-Demographic Analysis of American History” (Salon review here). This is where the connection with broader underlying conditions comes in. SDT explains how demographic changes periodically overwhelm the capacity of social structures to meet basic needs — first in the general population, then among existing and aspiring elites — with the resulting rise in political instability eventually leading to a period of likely state breakdown and possible civil war.

Turchin describes and presents data for several contributing factors and processes, which he identifies in societies from ancient Rome and China to the present. But within those broad outlines, a great deal more remains to be explained, and this paper’s findings break new ground in doing just that.

To explore the paper further, Salon reached out to Petersen and conducted an interview via email. It has been lightly edited for clarity.

Your paper “A ‘Need for Chaos’ and the Sharing of Hostile Political Rumors in Advanced Democracies" caught my attention because it seemed to pull some things together which hadn’t been well connected before. You write about several phenomena that are very different in some respects, and are usually treated separately -- conspiracy theories, fake news, negative campaigns, etc. — describing them as "hostile political rumors."   

I'll start with three questions about them: First, what do you see as the essential characteristics that make these different things similar, at least as they function in the world today?

Journalists and researchers often treat these information types as different – and they are indeed very different if we look at their origin and their epistemic value. But when citizens meet these information types in social media posts, at online forums etc., they are remarkably similar. First of all, citizens have limited information on whether they are true or not. Second of all, they all contain negative information about political groups and elites.

Second, how are these hostile political rumors in advanced democracies similar to phenomena seen elsewhere, or in our own past history?

As the saying goes, “The first casualty when war comes is truth.” Hostile rumors have always circulated in intergroup conflicts. They are used by political entrepreneurs to incite hostility and mobilize the audience. Some of the best documented cases comes from the grim literature on ethnic riots such as those in the Balkans and Rwanda. You never see people engage in the extreme violence and atrocities related to riots without an initial period of rumor sharing, to stir up sentiments.

And thirdly, how can a focus on their similarities help us to understand them better?

Modern politics is very far from a riot situation. But across Western societies politics is increasingly polarized, and many of the themes of hostile political rumors resemble the themes of rumors in riot situations. They all focus on how another group is powerful and is out to get us and that we need to act now. We can use these similarities to understand the psychology underlying the spread of hostile political rumors in democracies. Potentially, the difference between rumor-sharing in politics and in riots is more quantitative than qualitative. That is, the psychology is the same but the hatred of other groups is, fortunately, less intense in democracies for most people. So in democracies we spread rumors rather than buy machetes. Still, if this perspective is right, the worrisome implication is that rumor-sharing is, at the psychological level, a precursor to violence.

In your paper, you write that you were providing “an initial test of the role of chaotic motivations as explanation for motivations to share hostile political rumors,” as opposed to the role of partisan motivations. To do this, you created two scales — the violent activism scale and the legal activism scale. How did you create the scales, what was the logic behind them, and what did your initial test show?  

We took these scales from the literature on radicalization and terrorism. They are designed to measure individual differences in political engagement that occur within the boundaries of the political system, and engagement that occurs outside and targets the political order. Our tests showed that there is a strong relationship between being a violent activist and being motivated to share hostile political rumors on social media. Legal activism, in contrast, was not important. So, consistent with the riot literature, we see this association between a willingness to entertain violent thoughts and a willingness to spread hostile rumors.

How did belief in rumors compare with motivation to share? And what’s the significance in how these differ?

Importantly, violent activism was a substantially stronger predictor of motivations to share than of people’s beliefs in these rumors. This is crucial. It suggests that quite a few people share these rumors, even if they don’t believe them. Fact-checking solutions to fake news are built on the assumption that people who share don’t know what is true and false. But the problem is not that people don’t know what the truth is. The problem is that people don’t care about the truth. They do not share rumors because of their truth or epistemic value. They share them for other, and darker, reasons.

In your second test, you looked more closely into why violent activists share rumors. What did you discover that might surprise some people? One popular explanation associated with the "alt-right" is that rumors are spread for amusement — for the lulz. What did you find out about that?

A key feature of people’s motivations for sharing hostile political rumors, when asked directly, is to mobilize against groups they don’t like. That is, it is an instrumental act with a particular goal. Against the assumptions behind fact-checking, they don’t do it because they think the stories are true. Consistent with some past work on online trolling, people also share because they think it is fun. But the key message here is that they don’t do it just for the lulz. Sharing is an instrumental act too. People who share do so with a particular goal in mind. They want to incite hatred and chaos. And they share whatever will help accomplish that.

In Test 3, you developed another new scale, the “Need for Chaos” scale. You note that “Multiple distinct literatures have circulated around the concept of chaotic motivations (using different labels)” including research on online trolling, cyberbullying, social media protest mobilization, and democratic dissatisfaction and populism. But there was no “comprehensive and direct empirical assessment” you could find, so you had to come up with something of your own. What exactly did you mean by “need for chaos” and how did you go about constructing the scale?

When we use the label ”need for chaos,” we refer to the fact that some people seem to be emotionally excited about the disruption of the current order. They crave it in their personal lives and, of special relevance to politics, they crave it a broader societal scale. This is an extreme mindset and is not really captured by measures of populism and political cynicism. Is it much more radical than these. Often researchers tend to focus on the mainstream. But to understand the spread of hostile political rumors, we need to understand the extremes.

Pop culture, in contrast, is circulating much more around such extremes and some of the inspiration for specific items of the scale comes from movies such as "The Dark Knight" and "Fight Club." An iconic scene in "The Dark Knight" is an attempt to explain the motivations of the Joker with the punchline that “some men just want to watch the world burn.”

In a scene in "Fight Club," the main character explains a particular brutal fight by saying that he “felt like destroying something beautiful.” This is the sentiment we were interested in because it seemed to explain our patterns of findings: People instrumentally sharing whatever incites hatred and even finding it amusing. In political commentary there has been some interest in this sentiment too. For example, in a Danish newspaper article, it was argued that to Steve Bannon, “Donald Trump is a hammer used to smash the whole Washington establishment, such that we can start over.” I am not making a judgment of whether this is true or not. But this is the sentiment we wanted to capture.

How well does that account for the motivation to share hostile political rumors? What about belief?

The "Need for Chaos" scale was a very strong predictor of motivations to share hostile political rumors – the strongest we have yet found. And, again, it is stronger for motivations to share than for beliefs. People who share due to chaotic motivations do not necessarily believe what they share.

Do you have any thoughts about how researchers working on online trolling, cyberbullying, etc. might be able to use your “need for chaos” scale in their work?

My prediction would be that this underlies a lot of other online hostility such as online trolling and cyberbullying.

In Test 4, you sought to identify who those with a need for chaos were. As you note, chaos in Greek mythology is the original state of things, out of which order (“cosmos”) is produced. So, logically speaking, a need for chaos is a longing to wipe the slate clean, and begin again. This points to “those without stakes in the present status hierarchy: those who seek but lack status,” you write. “We therefore predict that chaos-incitement is a strategy of last resort by marginalized status-seekers.” 

To test this, you employed measures of status-seeking motivations, and measures of present social status. What did you find about how well these predicted the need for chaos?

People high in need for chaos were very high on a psychological trait termed "status-driven risk-taking." So they do indeed crave status. At the same time, they were significantly more lonely and felt lower in the social hierarchy than average. So, while these people crave status, they do not have it. And this is why they want to disrupt the current order. For them, “burning the world down” feels like their best bet for a brighter future. Interestingly, they also reported more mating success than average. This goes against some of the recent attention to so-called Incels, but is consistent with what we know from past research. Thus, short-term mating success is often related to antisocial traits.

You also found that need for chaos was associated with an ability to take advantage of a chaotic environment. Could you say a little bit more about that?

Humans are extremely strategic actors. If we were not, we would not have populated the earth. So, we should not expect every lonely person to crave chaos. People seek chaos as a strategic way of bettering their own status. So we should mainly expect people to incite hatred if they believe they would be good at navigating an environment characterized by the resulting chaos. In a series of more exploratory tests, we find that people higher in need for chaos lack empathy but report higher physical strength. These are the exact traits that will make you competitive in a chaotic world.

This discussion inevitably reminds us of Donald Trump, who has spent his whole life seeking status, always coming up short, even as president. He’s America’s No. 1 promoter of hostile political rumors, and everything your paper says about them seems to describe him perfectly as well. Any comments?

The paper is not about Donald Trump. The paper is about everyone else. In order to understand the present democratic crisis, we need to begin pointing the finger back at ourselves and the fleeting impulses we feel to incite hostility and stir up sentiments. While most people are fortunately not high in need for chaos, our analyses show that just a tiny bit of these sentiments are enough to trigger some motivations to share hostile rumors. The real problem are those tiny bits of chaotic motivations that you might hear as a whisper in the back of your mind when sitting in front of the computer screen. Today, you don’t have to go to the streets to engage in disruptive behavior. A small click with your computer mouse is enough.

In your concluding discussion, you write that “societies of today are facing specific sociopolitical developments that might spur a desire for chaos to a greater extent than in several decades,” and you cite Peter Turchin’s book, “Ages of Discord.” 

For me, this was another example of how your paper pulls together some things that hadn’t been well connected before: mass immiseration, giving rise first to various forms of protest, and increased political violence, heterodox belief systems (such as apocalyptic faith), and counter-elites willing to challenge or overthrow the existing state. Do you have any thoughts about how one might go further in connecting your work with some of the specific mechanisms Turchin points to?

Most political researchers are focusing on surface-level factors when trying to understand the present situation. They explain polarization and democratic discontent with the rise of the radical right in Europe, or Fox News in the United States, or the tweeting of Donald Trump or social media platforms. In my view, all these things are, however, not problems in themselves. They are symptoms of deeper problems that like gravitational forces push everything in a particular direction. If we are to get out of the present crisis in good shape, it is the responsibility of political researchers to begin to unravel what is the underlying driver of all this. And it cannot be an explanation that just works in the United States. Because we see similar symptoms across the Western world. So what is going on?

In my view, Peter Turchin is one of the few people that tries to tackle this question head on-by pointing to how historical models of state collapse can be used to understand the deep forces at work in the present. One of his basic points is that rising inequality increases status competition across the entire status hierarchy and that mass mobilization and elite polarization are natural consequences of this, historically as today.

In our paper, we too find that status concerns are key to the sharing of hostile political rumors. There might be many other things going on, and my point is not necessarily that rising inequality is the underlying drive of everything. My point is that we, as researchers and political analysts, need to stop focusing on the surface symptoms and look after the underlying forces.