Why Everyone Gossips (Yes, Including You)

Gossip has a reputation problem, which is ironic, because gossip is largely how reputations get built. Say the word and most people picture whispering, cruelty, someone's secret spreading through an office. Say it to a researcher who studies it, and you get a much less dramatic answer: talking about people who are not present is one of the most common things humans do with their mouths, and most of it is neither malicious nor secret.

How much of it we actually do

In 2020, Megan Robbins and Alexander Karan had participants wear small audio recorders that sampled snippets of their day for several days, rather than asking them to self-report - a method designed to sidestep the awkwardness of admitting how much you gossip. Coded across everyone in the sample, talk about absent third parties added up to close to an hour of the average day. That is a striking amount of a person's waking conversation, and it held up regardless of age or gender in their sample.

The more interesting finding was not the quantity but the content. The large majority of what got coded as gossip was neutral - simply exchanging information about someone's life, plans or circumstances, with no evaluation attached. Negative gossip, the kind people picture when they hear the word, was a minority of the total, and positive gossip - praise, admiration, good news about someone - was common too. Talking about people who are not in the room, in other words, is mostly just how people keep track of their social world, not primarily a way to tear it down.

Why we evolved to do this at all

The anthropologist Robin Dunbar proposed an answer in a 2004 review that is now a standard reference point in this field. Humans live in social groups too large to monitor through direct observation alone - you cannot personally watch everyone in your workplace, extended family and friend circle to know who is trustworthy, who is struggling, who just had a baby or lost a job. Gossip, in his account, is the tool that closes that gap. It lets information about who did what to whom travel through a network far faster than any individual could gather it firsthand, and Dunbar linked this directly to a broader idea he is known for: that language itself may have evolved substantially as a way to track social relationships at a scale grooming could manage for smaller primate groups, but talking could manage for us.

This framing helps explain why gossip is not confined to any particular culture, class or gender - despite a persistent stereotype that it is mainly something women do. The audio-recording studies find men and women gossiping at similar rates; the content sometimes differs, but the overall appetite for it does not.

The useful kind: prosocial gossip

A 2012 set of experiments by Matthew Feinberg, Robb Willer, Jennifer Stellar and Dacher Keltner looked specifically at gossip that warns others about someone's bad behaviour - a colleague who cheats, a partner who lies. Their participants who witnessed such behaviour and then had a chance to pass on a warning experienced a drop in frustration and distress after doing so, and the people who received the warning changed how they interacted with the person described, avoiding being taken advantage of themselves. The researchers labelled this prosocial gossip: sharing reputational information not to enjoy someone else's downfall, but because groups that share accurate information about who cannot be trusted function better than groups that do not.

That does not make every act of gossip virtuous. The same channel that spreads an accurate warning can just as easily spread something exaggerated, out of context or simply false, and unlike a direct conversation, the subject of the gossip has no chance to correct the record. The mechanism that keeps groups honest also has no built-in filter for accuracy.

What this does not mean

None of this is a case for careless talk about people behind their backs, and it is not a licence to treat every private conversation as fair game once someone leaves the room. The research describes what gossip typically does at a population level - inform, warn, bond - not what any specific instance of it is doing, and a cruel or false rumour is exactly as damaging as it has always been regardless of how ordinary gossip is in general. The recording studies, useful as they are, still rely mainly on samples of American adults wearing devices they knew about, which likely shaped what they said near the microphone.

There is also a second function that operates between the people doing the talking, not just about the person being discussed. Sharing information, opinions and reactions with someone else is itself a way of signalling trust - you are telling them something you would not tell just anyone, and their reaction tells you something about whether your values line up. Seen this way, a lot of ordinary gossip is less about the third party than about the two people doing the talking, checking in with each other about what counts as normal, admirable or worrying in their shared social world.

What this means for you

The next time you feel a flicker of guilt for talking about someone who is not there, it is worth asking what kind of gossip it actually was. Swapping news about a friend's new job, checking in with someone about how a mutual acquaintance is doing after a hard year, or warning a friend about someone who treated you badly are not character flaws - they are, by the evidence above, close to the default way humans maintain their social world. The distinction that matters is not whether you talked about someone who was absent. It is whether what you said was true, and whether you would be comfortable saying it to their face.