How digital insular social groups polarize and change behavior


Many people feel ostracized or alone in their everyday lives, a somewhat obvious thing to say. The internet has nooks and crannies and often people who have been on the fringes can find peers there. The anonymity offers equality—everyone is the same, and the etiquette is set both by the structure of the program, and by the ideological survival landscape—who is there. In a network like the defunct podium app, for example, white nationalists were more common and so the etiquette included all the common things you might expect from bigoted groups. On places like tumblr, ideas tend toward the liberal, and the etiquette becomes more astringently cautious about offense.


In sociology, social contagion theory specifically, relationships are described as either weak bonds/ties/actors or strong bonds/ties/actors. A weak actor is something that cannot change your behavior. For most people, what someone says on the internet is a weak actor, and a strong actor is a family member or a close friend. But on the internet, in these corners where outsiders collect, the relationship is often inverted. The rest of the world is a weak actor, and the anonymous people they’ve never met become strong actors on their behavior, their thoughts, their actions.

This is why the kind of social engineering being done by the Russians is so effective and dangerous, how it leads to insurrections and mob violence. It’s not merely that people believe things they see/hear more than once, or their assumptions about viable sources. It’s also that for many people, the cyberspace enclaves they populate become strong actors on their ideas. But the root of it is belonging. They don’t feel as if they belong, they feel like they are divided from others, they feel invisible in their everyday lives, they feel like their opinions aren’t heard or they cannot express themselves openly, and so the group they do have, becomes critically important. It’s also true that most outside information like news or random facts are usually absorbed through weak ties, because those casual acquaintances move in other circles and bring information back like bees returning to a hive. This is how internet social groups become insular—A space full of strong actors with no outside information. And as has been studied many times, insular groups tend toward polarized opinions—in this case not referring to duality, but an extreme perspective.

If a group focuses on a topic, they’re often referred to as deliberating groups. “As polarization gets underway, the group members become more reluctant to bring up items of information they have about the subject that might contradict the emerging group consensus. The result is a biased discussion in which the group has no opportunity to consider all the facts, because the members are not bringing them up.”1 Often deliberating becomes highly susceptible to localized “cascade effects” in which a piece or pieces of information, true or false, are spread through the group by higher status members on down the hierarchy (this status determined by expertise, age, presumed knowledge or any number of factors). Those with very little private information are overtaken by this cascade through informational externality– or the acceptance of the verity of facts based on a lack of desire to engage in confrontation, a presumption that the others know more, or that they must be wrong somehow. Often as ideas become more extreme, pressure is applied to conform, and members will allow even their own senses to be denied, in order to not be seen as part of the contested idea. This is called reputational externality. Both are means by which information is spread from one member of the group to another. Essentially, a cascade is so forceful it can often be seen as mimicking viral modeling. But acceptance of influence, susceptibility to it, isn’t uniform. It’s more likely to be uniform, if the deliberating group is a strong actor on a persons life. Meaning that if the group is seen as important to the person, their threshold for allowing a cascade to hit them is much lower.

“A further specific characteristic behavior within group dynamics is “groupthink”. It’s the phenomena within polarized groups, where concurrence-seeking and cohesiveness becomes so dominant that it tends to override realistic appraisal of alternatives. Groupthink replaces independent critical thinking. That results in irrational and dehumanizing actions directed against opposing groups.”2

In other words, deliberating groups, through complex dynamics, produce one uniform polarized answer, and if that group has become a strong actor on an individual…it changes behavior often in ways that mobilize them to be hateful or abusive.

But it’s all rather flimsy in actuality, because of that distance and anonymity. The lack of actual connection breeches many of the evolved cues of bonding—meaning the physical aspects of it like proximity, physical contact, hormone triggers for oxytocin, body language and so on. Without personal contact or a subject to deliberate, such groups usually collapse. To prevent that collapse, group leaders will often organize contact in the form of rallies, or protests. However, in social groups that are exclusively online, the groupthink produces things like retaliation campaigns such as letter writing, spamming, and negative press or reviews. The stronger an actor the group is in a member’s life, the more important the ideas generated within it become, until eventually, some see it as necessary to act independently on those ideas.

“…in a clinical study of 144 individuals who had threatened some form of violence against others, 8 were found to have threatened mass homicide… All 8 subjects said they had intended to kill as many people as possible, and all cases involved targeting a specific group against whom the would-be perpetrator held a grievance. Over the 12-month study period, none of the 8 subjects carried out or attempted to carry out their plans. However, 2 of the 8 assaulted a person unrelated to the targeted group.” 3

Obviously mass shootings are a rarer example of such actions, and are a statistical outlier. However, recent events make them a topic worth including. Extreme opinions and the group dynamics that help them to proliferate are caustic to the psyche. Often the group dynamic breaks down the filters and rational safeguards that inhibit action. As the social group becomes more of a strong actor in a person’s life, the ideas built there also become central. Without satisfactory organized paths or directions for the ideas, the groups do not often maintain cohesion. This collapse can also potentially function as a trigger for behavior, sometimes not even against the initial target of concern, by removing the support for those central ideas, something which could only exist in that deliberating group dynamic. Members are left with only an extreme idea, and a reality that cannot ever support it. This discrepancy is an obvious trigger also.

Understanding how the inversion of weak and strong ties happens, understanding how deliberating bodies work and what rules might be in place to prevent polarizing opinions, being able to see how these things are impacted or impact a digital landscape would be very useful indeed. Many of these things have undergone statistical analysis and can be somewhat predicted. Further research is needed.

1. “Deliberative trouble: why groups go extreme” Cass Sunstrin

2. “How social media causes disruptive group dynamics” Tony Saldanha

3. “Mass Shootings and Mental Illness” Agnes, Knoll

One response to “How digital insular social groups polarize and change behavior

  1. I don’t bring up how choices are made within the group, or things like confirmation bias (the tendency to select and include as evidence only the things that support the initial conclusion). I didn’t include these because I thought they would obscure the points. But it’s undeniably true that on the internet, where it’s not only possible to find groups based on shared ideas or habits, it’s encouraged and engineered for, confirmation bias is integral to the actual formation of the deliberating group. What I mean is, people who already have some thoughts on an issue, seek out a deliberating group—like a message board or chat room— that is discussing those ideas. Confirmation bias is already in play, as they’re looking to lodge in a group that already agreed with and supports their initial thoughts. So the polarization effect is much much faster.

    I should also note that informational externality and reputational externality are also sometimes discussed as “social comparison theory” and “informational influence theory”. The former refers to that ranking structure and the decision to defer to someone else, either to avoid controversy, or to avoid ridicule. The latter refers to the willingness to accept an argument from someone who seems persuasive or knowledgeable, simply because one doesn’t believe one’s own information can be trusted.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s