Researchers sampled 467 people’s real conversations and found 74.3% of gossip was neutral; women gossiped more than men, but only when the talk carried no positive or negative judgement

Gossip has a branding problem. The word brings to mind whispered cruelty, unverified rumours and someone enjoying another person’s embarrassment. Researchers use a much broader definition: talking about someone who is not present.

Under that definition, “My sister starts her new job on Monday” is gossip. So is an update about a colleague moving house or a friend learning to drive. The content can be positive, negative or neither.

When researchers recorded fragments of real conversations from 467 people, that ordinary version dominated. Nearly three quarters of the gossip they identified contained no positive or negative judgement at all.

The result does not make malicious gossip harmless. It shows that the behaviour grouped under one morally loaded word includes a large amount of mundane social information.

A recorder sampled life rather than memory

Psychologists Megan Robbins and Alexander Karan combined data from five studies that used the Electronically Activated Recorder, known as the EAR. Participants wore a small device that periodically captured brief snippets of sound while they went about daily life.

The method offered an unusual advantage. Most gossip studies ask people how often they gossip or place them in artificial scenarios. Self-reports are vulnerable to reputation management because few people want to describe themselves as a gossip. Random audio samples allow researchers to observe what people actually discuss outside a laboratory.

The combined sample included 467 participants, 269 women and 198 men. Most were university students, although one of the five samples included adults from the wider community. Trained coders identified 4,003 instances in which a participant discussed a person who was not present.

In the analysis published in Social Psychological and Personality Science, gossip occupied about 14 per cent of the sampled conversations. Extrapolated across 16 waking hours, the average participant spent roughly 52 minutes a day doing it.

Only 34 people produced no gossip in the available recordings. That does not prove they never gossiped. The EAR sampled short windows rather than recording every word, so any behaviour outside those windows remained unheard.

Most gossip was unevaluative

Coders rated each instance as positive, negative or neutral. Neutral gossip accounted for 74.3 per cent. About 15 per cent was negative, while roughly 9 per cent was positive.

The proportions challenge a common assumption that gossip is defined by hostility. In this study, the typical example was closer to passing along a social fact than attacking an absent person.

Neutral does not mean unimportant or necessarily safe. A factual update can still be private, inaccurate or unwelcome. Telling a colleague that someone is interviewing elsewhere may contain no praise or criticism while carrying real consequences. The coding captured evaluative tone, not consent, truth or potential harm.

Nor did the researchers treat talking about celebrities as the standard case. Participants mostly discussed people they knew. Celebrity gossip occurred, but familiar acquaintances supplied far more material.

Women did not do more “tear-down” gossip

The study tested several stereotypes about who gossips. Women produced more gossip than men overall, but the difference was confined to neutral gossip. Women and men did not differ significantly in how much positive or negative gossip they produced.

That distinction matters because the stereotype itself often turns women’s ordinary social conversation into evidence of a character flaw. In the recordings, the extra volume was informational rather than more judgemental.

Income and education also failed to support a familiar class stereotype. Less affluent and less educated participants did not gossip more than wealthier or more educated participants.

Age and personality showed some differences. Younger people engaged in more negative gossip than older people. Extraverts gossiped more than introverts, which may reflect having more conversation and more social contact rather than a special appetite for other people’s business.

The sample limits those comparisons. It contained many young adults, and the community group was not large enough to represent every age, occupation or culture. The findings describe these five datasets, not a universal daily quota.

Another study found a more negative mix

Definitions and methods change what researchers see. A later project asked 309 adults to report gossip they sent or received about members of their real social networks. Across more than 5,000 reports, negative portrayals were slightly more frequent than neutral or positive ones.

That 2021 study of gossip and reputation focused on identifiable network members and asked participants to report episodes themselves. The earlier EAR research captured spontaneous conversation, including uneventful statements that a participant might never remember as “gossip”.

The later study also found that people used gossip to update their impressions of a target’s trustworthiness, warmth and competence. Most reports concerned information the sender claimed to know through direct experience, and people tended to share with receivers they valued and trusted.

Gossip, in other words, can function as informal reputation data. That function makes accuracy and motive especially important. Useful warnings, idle updates and deliberate character attacks can travel through the same conversational channel.

The word covers several social jobs

An integrative framework for organisational gossip defines it as a sender communicating to a receiver about a target who is absent or unaware of the content. The definition does not require the message to be negative or false.

That triangle can serve several purposes. People exchange information about who is dependable, interpret confusing behaviour, test social norms and signal closeness to the listener. They can also exclude, manipulate or damage someone who has no opportunity to reply.

The same sentence can perform more than one job. A warning about a manager may protect a colleague, strengthen trust between the speakers and harm the manager’s reputation at the same time. Classifying the sentence as positive or negative cannot settle whether sharing it was justified.

Neutral is not innocent, and negative is not always malicious

The 74.3 per cent figure is best read as a correction to the idea that all talk about absent people is an attack. It is not a moral permission slip.

Negative gossip can communicate a legitimate warning about exploitation or broken trust. Neutral gossip can violate privacy. Positive gossip can flatter someone sincerely or serve a strategic purpose. Tone is only one dimension.

There is also a measurement complication: people knew they were wearing recorders. The devices sampled intermittently and participants could review recordings before researchers used them. Those safeguards were ethically necessary, but awareness of recording may have reduced the most sensitive talk.

Even with those limits, the study revealed how much ordinary conversation revolves around people who are elsewhere. Humans use one another’s experiences to map a social world too large to observe directly.

Most of that mapping is not scandal. It is the quiet circulation of who went where, what happened next and how everyone is connected. Gossip earns its bad reputation from the damaging edge of the category, while most of the category sounds almost boring.