The 13,000-word gap that no recorder ever heard

Dark couples-seminar-style bar chart titled “The 13,000-word gap”, masthead “The Conversation Clinic · Couples’ Communication Workbook · 2026 edition”, plotting two confident bars of words spoken per day: a gold Women bar at 20,000 and a much shorter slate Men bar at 7,000, joined by a dashed red bracket labelled “the 13,000-word gap — a conversation deficit he can’t help”. A badge reads “2.9× — she speaks nearly three times the words he does. It’s neurology, not nagging”, an italic aside over the men’s bar says “out of words by lunchtime”, and the kicker advises planning the conversations that matter for breakfast because “by dinnertime, his 7,000 words are spent”. The only source given is a quiet footnote crediting “a bestselling 2006 book on the female brain”; no sample size, error bar, or study appears anywhere on the chart. — Showing the misleading chart

The same comparison on a light background, kicked “The same question · asked with a voice recorder running” and titled “Counted, the gap is 546 words — give or take eight thousand”: gold and slate bars show estimated daily words from Mehl et al.’s recorder study — women 16,215, men 15,669 — each with whiskers spanning ±1 standard deviation (7,301 and 8,633), the measured 546-word gap noted between them. Dashed red lines mark the workbook’s claims: “her” 20,000 crosses inside the women’s whisker, and “him” 7,000 lands almost exactly at the bottom of the men’s. Three cards explain that the 20,000-vs-7,000 figure has n = 0 — linguist Mark Liberman found no study behind it, only a self-help book’s guess the author agreed to delete from future printings; that the recorders heard 396 students speak a 546-word, 3.5% gap, not statistically significant, with the quietest (~500 words a day) and chattiest (~47,000) talkers both men; and that a 2025 follow-up of 2,197 people aged 10–94 found a real but modest 1.2× difference confined to ages 25–64. A footnote credits Science (2007) and notes both charts are our own demonstration. — Showing the misleading chart

Be the first to star this exhibit

A couples-workshop classic: women speak 20,000 words a day, men just 7,000 — two bars so crisp nobody asks how many people were counted. The answer is nobody: the figures trace to a self-help book’s guess. Clip voice recorders to 396 real people and the measured gap is 546 words — against standard deviations of seven and eight thousand.

Mehl, Vazire, Ramírez-Esparza, Slatcher & Pennebaker, “Are Women Really More Talkative Than Men?”, Science 317, 82 (2007) — EAR-recorded daily word estimates, 396 participants; the 20,000-vs-7,000 folklore as traced by Mark Liberman (Language Log, 2006); follow-up: Tidwell, Pfeifer, Mehl et al., J. Pers. Soc. Psychol. (2025), 2007Added July 26, 2026

01The claim

Women speak nearly three times as many words a day as men — 20,000 to his 7,000. It’s neurology, not nagging: her brain runs a conversational marathon while his rations a small daily budget, so plan the conversations that matter for breakfast. By dinnertime, his words are spent.

02The trick

A crisp bar chart looks exactly as confident whether it rests on half a million recordings or six — and this one rests on none. The 20,000-vs-7,000 figure carries no sample size, no error bars, and no study, because there is no study: when linguist Mark Liberman went hunting for its source, the trail dead-ended in a self-help book’s uncited guess, passed from book to book until a bestselling 2006 neuroscience title printed it as fact — then agreed, once challenged, to delete it from future printings. The chart’s whole authority is typographic. Round numbers wear the costume of measurements, the tidy 13,000-word bracket does arithmetic on two inventions, and the one citation is a vibe (“a bestselling book on the brain”) rather than a reference anyone could check. Our Steam-leaderboard exhibit shows what a percentage with 96 reviews behind it can get away with; this chart does that trick one better, with a statistic backed by zero. The precision is entirely phantom — and it worked for years, because a specific number simply sounds counted. Real counting is expensive and slow: it took researchers clipping voice recorders to hundreds of people across six years to measure what the workbook simply asserted. (This exhibit is our own demonstration in the house style of a couples-communication seminar workbook — no real workbook ran these exact charts, though for a decade plenty ran these exact numbers.)

03The fix

Ask the question the chart is dressed to deflect: how many people, counted how? Mehl and colleagues actually counted — unobtrusive recorders sampling 30 seconds of ambient sound every 12.5 minutes from 396 university students — and the honest chart wears its uncertainty on its sleeve: women 16,215 words a day, men 15,669, a 546-word (3.5%) gap that is not statistically significant, drawn with ±1 SD whiskers of 7,301 and 8,633 words. Those whiskers are the finding. Talkativeness varies enormously person to person — the quietest talker logged about 500 words a day and the chattiest about 47,000, both of them men — so sex explains almost nothing about how much a given person talks. Overlaid on the real data, the folklore confesses: the workbook’s “20,000-word woman” sits within one SD of the real female average, and its “7,000-word man” lands almost exactly one SD below the real male average — both are ordinary people on the honest chart, promoted to archetypes by a chart with no data. Honesty then cuts both ways: a 2025 registered-report follow-up (2,197 people, ages 10–94) found women aged 25–64 really do average more — 21,845 vs 18,570, about 1.2× — with no reliable difference in any other age band. “Roughly equal, with a modest midlife gap” is the counted answer; 2.9× was the sold one. The tell is a suspiciously precise statistic traveling with no n: before trusting any number that sounds measured, ask who measured it — some of the most-quoted numbers in pop science have never been counted by anyone.