The Psychology of Aggression on the Net

The chapter starts with an anecdote, about a generally even-tempered character who wanders onto USENET and gets involved in a flame war. Conclusion that this is common, though not as pervasive as some would suggest.

Aggression, itself, is a natural impulse and "an essential part of the life-preserving organization of instincts." As social creatures, we learn to curb aggression when dealing with other people, but the visual cues we provide one another in meatspace are absent online, so we're more likely to cross boundaries we respect in face-to-face dialog.


Aggression is most often a response to stimulus - specifically, frustration in the face of an obstacle that prevents us from achieving our goals.

An experiment (Barker and Lewin, 1941): children are shown a room filled with intriguing toys. Allowed to enter immediately, they engage in normal play. However, if kept at a distance where they can see but not access the toys, frustration builds and they act aggressively (often smashing the toys rather than playing with them) when the barrier is removed.

The delay inherent in the medium, especially at modem speeds, is a barrier between the user and the information he seeks. This lag is a source of much frustration, such that a person who enters a forum is already annoyed by the wait, and is thus predisposed to aggressive behavior.

A follow-on case study of a chat room with a controlled delay: a two- or three-second wait kept people reasonably content, but when the lag was longer than eight seconds, the frustration level (and precipitating aggression) rose considerably.

EN: it was not reported, but might be interesting to know, if there is an upper threshold at which the participants simply abandon the site.

Conversational lag in meatspace is normal: averaging 1.25 seconds among same-sex and 3.21 among mixed. In meatspace, such time is filled with nonverbal cues. In the absence of such cues, frustration builds. For example, Japanese negotiators have learned to pause, stone-faced, to put American businesspeople off guard and gain a tactical edge.

There is also the phenomenon of "net splits," or disconnects, where a person may drop out of a forum without any visual indication: it seems that his "partners" have simply gone silent and are ignoring his attempts at a communication.

EN: This information is a bit dated, and I suspect that the increase in connection speed has considerably decreased the effects of lag and disconnect in synchronous conversation forums.

There was also a mention of the frustration inherent in locating the forum in the first place: a user arrives at a site after several minutes of dead-end searching for the information he desires on a public search engine, then has to navigate to a page. If it's a forum, he has to learn how to contribute. Especially with users who are new to a forum, they may arrive in an agitated state.


Most individuals do not handle irritation in a rational manner - specifically, they do not attribute their irritation to the specific things that caused it, but tend to vent their frustration at any opportunity that presents itself. This is the reason that an aggressive reaction can be far out of proportion to the provocation.

We are likewise inclined to see others actions as aggressive when we are, ourselves, agitated. An act that we'd dismiss as clumsy or careless in a calm mood is seen as rude and aggressive when we are agitated. Even when such an action is, in fact, aggressive, we are more likely to respond negatively than to pause and consider the other's motives, or what we may have done to elicit it.

Retaliation against aggression is both instinctive (threat-response) and social (dominance), and a common pattern is to retaliate in a plus-one manner (slightly harsher than the threat) to discourage the attacker from continuing. The result is rapid escalation. In "meatspace", our desire to avoid a physical conflict will usually help to diffuse a hostile confrontation - but online, the potential for physical violence is not a deterrent.

Even when one's intention is to diffuse the situation, much peace-making relies upon reading the nonverbal cues of aggression to determine another person's mental state, and using gestures and tone to assuage their irritation. None of this is possible in the online medium.

There are also cultural differences in conversational styles, particularly argumentative styles - what an American considers to be a mild insult might be thought an outrage to an Asian, or completely innocuous to a European. Even within a nation, there are various cultural groups whose behavior is unintentionally offensive to others.


There is also the matter of interpretation. It is not so much what a writer intended, or even the literal words that they used, as the way in which the reader reacts to them - and since many people lack communication skills (particularly in writing), it can turn an innocuous conversation into a flame war.

In an example, one person asks for a recommendation for a good class in skiing, and receives the response that a particular school has a special deal for beginners. It may have been intended as friendly advice, but could be interpreted as an insult: suggesting that the other person considers him to be completely unskilled and in need of a discount price.

The example is taken further when another person joins the conversation and suggests an alternative school, and it degenerates into a flame war, focused largely on class conflict (one school is seen as a cheap place for the unwashed, the other as an overpriced resort for snobs). Ironically, the use of emoticons (smileys) to mitigate the hostility is interpreted by the other as sarcasm, having the opposite effect than the sender intended.


There have been a number of studies of social interaction via electronic media by Sara Kiesler. Some examples:

In all, it was also commented that while the study was focused on aggression, incidence of aggressive behavior overall was "really quite rare" compared to the researchers' expectations.


The 'reproach" is an attempt on the part of an individual who has no formal authority to control a forum to discourage aggressive behavior. Three parts are identified:

  1. The Offense - A statement that is considered inappropriate for the forum
  2. The Reproach - The statement that informs the offender of the nature of their offense
  3. The Accounting - An acknowledgement or apology by the offender.

In some instances, an "accounting" does not follow. In others. The "accounting" is an attempt to justify the offense, or question the rules. In still others, this can spark a "metadiscourse" that can distract from the purpose of the forum.

The theory of cognitive dissonance posits that people feel uncomfortable when they do something that is out of line with their attitudes, beliefs, and perceptions - so the simple act of reproach can guide an individual back toward more rational behavior, unless their intention was to be disruptive.


Anonymity is seen as a contributor to aggression, as our actions are not associated with our "real" persona, or with a personal that we intend to maintain. Conversely, identity is an inhibitor to aggressing: having rude and aggressive behavior associated with a persona we wish to be perceived as something other than rude and aggressive is counterproductive.

Also, anonymity is seen as a shield against retribution: if you cannot be identified, you cannot be tracked down, and do not suffer the consequences of your behavior (which can include physical retribution or legal persecution for your actions).

Temporal, as well as physical distance, is a consideration: there is a "cooling off" period with traditional mail - an opportunity to reconsider before you drop an envelope in the mail - whereas e-mail is instantaneous. Also, studies have show that people are more casual about composing electronic messages.


The principle of "catharsis" is a healthy purgation of emotion that enables a person to return to a rational state, and that venting frustration is a form of commiseration that can be a shared experience that builds cohesion rather than creates distance.

Research results are inconclusive as to whether expressing anger, frustration, and aggression have a cathartic effect - i.e.., that they assuage anger rather than intensifying it.

The author also suggests that it is her opinion (she expressly states "I think that") acceptance of aggressive behavior encoruage4s further aggression and increases one's threshold for aggression. No evidence or support is provided for this assertion.


Aggression online is limited by the capabilities of the medium. It is, by its nature, verbal aggression that affects the emotions of the recipient, intentionally or accidentally.

While the majority of this chapter dealt with aggression in interactive communications, there is also aggression in broadcast communications in the form of "hostile Web sites" that vilifies and offender. There is a lot of gray area between the first-amendment right to express frustration and libel. Such sites often encourage others to join - the author specifically mentions "the dick list."

Impersonation is another form of aggression, with an individual using the name or handle of an offender to behave in ways that will cause others to retaliate against them, thereby doing harm to the person's reputation.

Channel wars also exist, a form of Internet gang warfare in which multiple individuals rally against a site or an organization and attempt to flood a channel with their side of the story, or disable the channel for use by others.