Flame War
A reader's summary of the oldest kind of online fight — the heated, escalating exchange of insults — and of what the early research and the early networks made of it.
The term at a glance
A flame war is a prolonged, hostile exchange between people online, typically on a forum, a mailing list or a comment thread, in which replies get more personal and the original subject drops out of sight. A single abusive message is a flame; a war needs at least two sides who keep answering. The word is dated now, and most of the behavior it described has moved to platforms that give it other names, but it is the template for them.
Origin
“Flame” comes from hacker slang on the early academic networks and is recorded in the Jargon File, which also notes the related “flame on” and “flame off,” echoing the call of a comic-book character. The word was in circulation before Usenet; Usenet gave it a permanent home, down to groups created for flaming alone. It is part of the same founding mythology as Eternal September, in which the old, small community of regulars is remembered as having kept its fights under control until newcomers arrived.
History and context
Because early messages were plain text, the first practical problem was tone. Scott Fahlman's 1982 smiley was a response to jokes being read as claims, and it still functions as the cheapest form of de-escalation. The first academic work on the subject came soon after: the 1984 Kiesler, Siegel and McGuire study, and a 1992 rebuttal by Lea and colleagues that complicated it. A joke about online arguments from the same period, later named Godwin's law, shows how quickly the community began to treat escalation as a pattern to be recognized rather than an accident. Another recurring flashpoint is the correction itself, as in Cunningham's law and Muphry's law.
Main ideas
A flame is heat, not argument
In the hacker slang recorded in the Jargon File, to flame was to speak or write rabidly and at length, often on a trivial subject, and a flame war was the resulting exchange. The defining feature is intensity rather than disagreement: two people can disagree for years without a flame war, and a flame can erupt over a question both sides agree on.
The smiley was a fire-prevention device
On 19 September 1982 Scott Fahlman proposed on a Carnegie Mellon bulletin board that jokes be marked with :-) so that readers would stop taking them literally. The thread that prompted it was a dispute over a joke about a mercury spill in an elevator. Plain text carries no tone of voice, so a sarcastic remark and a hostile one look identical, and a mark of intent became the first patch.
Early research saw more uninhibited behavior
Sara Kiesler, Jane Siegel and Timothy McGuire's 1984 American Psychologist paper compared groups deciding by computer with groups meeting face to face. The computer groups showed more uninhibited behavior, including swearing and insults, and took longer to reach decisions. It is the usual starting point for the claim that the medium itself loosens restraint.
The medium is not the whole explanation
Martin Lea, Tim O'Shea, Pat Fung and Russell Spears's 1992 review of flaming argued that the evidence was thinner and more mixed than the popular story. Much of the apparent hostility came from small samples, laboratory tasks with strangers and ambiguous coding of what counted as a flame. Group identity, norms and the topic under discussion mattered alongside the technology.
Disinhibition cuts both ways
John Suler's 2004 article on the online disinhibition effect listed anonymity, invisibility, delayed replies and the sense of a separate world as reasons people say things online they would not say in person. He distinguished benign disinhibition, such as unusual openness and generosity, from toxic disinhibition, such as rudeness and threats. The same conditions produce both.
Communities answered with etiquette
The early networks wrote rules down. Usenet FAQs, the 1990s netiquette guides and Virginia Shea's 1994 book Netiquette all urged readers to think before replying, to cool off before posting, and not to feed a fight. These were social tools rather than technical ones, and they worked best in groups small enough for the regulars to enforce them.
Critique
- “Flaming” is a judgment, not a measurement. What one reader calls a flame, another calls blunt speech or banter between friends. Studies that count flames depend on coders agreeing, and the 1992 review shows how much of the early effect rested on that choice.
- The golden age is partly nostalgia. Accounts of well-behaved early networks come mostly from regulars remembering their own corner. Usenet had flame wars from its first years, and a small group that polices itself is not evidence that the medium was gentler.
- It can blame the screen for what people bring to it. The disinhibition account makes the interface the cause. Critics point out that groups with strong norms stay civil online and that face-to-face groups with weak ones do not, which suggests the technology amplifies rather than creates the behavior.
Impact
Flame wars gave the internet its first vocabulary for conflict and its first moderation problem. The questions they raised — when to ban, whether to answer, whether anonymity helps or hurts — are the ones every community still argues about. They also left a commercial lesson: a platform that ranks content by engagement will find that a fight is engaging, which is one way the old behavior was turned into a feature, as in rage bait and trolling. What changed is less the behavior than who profits from it.