Say it again: repetition and the illusory truth effect

Series: How minds are moved · Foundations: how persuasion works · 3 of 7

On 12 November, 362 websites published the same headline, word for word, within six hours: “iHeartRewind 2026: Check Out iHeartRadio’s Top Artists & Songs Of The Year”. My news engine, LSN News, found that every one of them was a station site of the same radio company, promoting its own year-end list. To a reader it looked like hundreds of sources. In fact it was one voice, said 362 times. That gap between how often you hear something and how many people are actually saying it is where repetition does its work.

Contents · 9 sections

TL;DR #

• Hearing a statement again makes it feel more true. Psychologists call this the illusory truth effect, and it has been measured since 1977.
• Knowing the right answer does not fully protect you: a 2015 study found the effect even for statements people knew were false.
• One exposure is enough to shift belief in a fake headline, even when it is labelled as disputed.
• Slogans, catchphrases and repeated headlines work through the same mechanism: familiarity feels like truth.
• The defence is to ask, the first time you meet a claim, “is this true?”, and to count voices, not repetitions.

The experiment that started it #

In 1977 Lynn Hasher, David Goldstein and Thomas Toppino asked students to rate how true 60 plausible statements were, on topics like politics, sport and the arts: statements most students could not know for sure. They did it three times, two weeks apart. Some statements came back each session; others were new each time.

The result was simple. The repeated statements were rated as more and more true, whether they were actually true or false. The new ones were not. The authors concluded that how often we have met a statement is one of the criteria we use to decide whether it is valid. Repetition does not add evidence, but the mind treats it as if it did.

Daniel Kahneman explains why in Thinking, Fast and Slow, which is in my library. When something is familiar, it is processed more easily, and that ease feels like a signal that all is well. “A reliable way to make people believe in falsehoods is frequent repetition,” he writes, “because familiarity is not easily distinguished from truth.” The next sentence is the one to remember: authoritarian institutions and marketers, he says, have always known this.

Knowing better does not protect you #

For years researchers assumed the effect only worked on things we do not know. If you know the answer, surely repetition cannot fool you.

In 2015 Lisa Fazio, Nadia Brashier, Keith Payne and Elizabeth Marsh tested that assumption and published the result under a blunt title: “Knowledge does not protect against illusory truth”. People who could correctly answer a question in a later test still rated a repeated false statement on the same topic as more true than a new one. The authors called it “knowledge neglect”: when a statement feels fluent, we often do not bother to consult what we know.

In 2018 Gordon Pennycook, Tyrone Cannon and David Rand took real fake-news headlines as they had appeared on Facebook. A single prior exposure raised how accurate people judged them, both in the same session and a week later. The effect held even when the headlines carried a “disputed by third-party fact-checkers” label, and even when they went against the reader’s own political side. There was one limit: completely absurd statements, like “the Earth is a perfect square”, did not become believable.

How slogans use it #

A slogan is repetition made portable. It is short, it rhymes or has a rhythm, and it is meant to be heard thousands of times. Every side uses them, because they work:

• War: “Remember the Maine” in 1898, “Keep calm and carry on” (printed in 1939, mostly unused until it was rediscovered decades later), “Mission accomplished” in 2003.
• Politics: “Yes we can”, “Take back control”, “Make America great again”, “Get Brexit done”. Each was repeated by supporters until it stood for a whole argument.
• Public health: “Stay home, protect the NHS, save lives” was repeated at every UK government briefing in spring 2020.
• Commerce: “Because you’re worth it”, “Just do it”. Nobody checks whether they are true. That is the point: they are not claims, they are habits.

Jacques Ellul, whose ideas opened this series, saw repetition as the core of propaganda: slow, constant impregnation, over a period long enough to exceed a person’s capacity for attention and resistance.

Three cases from my news engine this week #

LSN News groups articles into stories and counts “voices”: outlets that run the same headline count as one voice. Our article history starts on 7 November 2026, so these are snapshots from one week, and the counts will grow. The public story links expire about seven days after a story goes quiet, so I give the numbers here.

One voice, 362 copies #

The iHeartRewind headline above is the cleanest example. The main story (on LSN News) had 384 outlets but only 5 independent voices. The same pattern holds for news: in the same few days, 562 sites ran one identical headline about a New York sightseeing helicopter, and 491 ran one about a children’s clothing chain filing for bankruptcy. Most are radio-network or wire copies. That is normal syndication, not a plot. But it means that whoever writes the copy decides what hundreds of “local” sites repeat, and a reader who sees it everywhere feels that it is confirmed everywhere.

A rumour repeated by its debunkers #

From 7 November a rumour spread that a Simpsons episode had “predicted” a worldwide blackout on 17 November. It started on a Brazilian site and spread through Latin America to Spain, Greece and the Arab world: 85 headlines from 74 outlets in 6 languages (main story). There is no evidence for it, and many of the articles said so in the text.

But look at the headlines. 43 of the 85 were questions that did not answer themselves: “Will there be a global blackout on 17 November?”. Only 23 called the rumour false in the headline. A fact-checker, Maldita.es, published its check six days after the first article. Most people read only the headline. So most of the coverage repeated the claim, once more, to people who would never read the correction. If Pennycook’s results hold here, each of those questions nudged the rumour towards feeling true.

One phrase, both sides #

The third case is a phrase, not a headline. In the week from 7 November, the words “hybrid war”, “hybrid threats” or “hybrid attacks” appeared in the titles or leads of 476 articles from 270 outlets in 15 languages, spread over 294 separate stories. Western outlets used it about bomb hoaxes at European schools, cyberattacks and drones, blaming Russia. Some Russian outlets put it in quotation marks, as a Western accusation, while Belarusian state media used it for “external players” turning the post-Soviet space into a theatre of hybrid war. A Spanish party leader used it about Morocco.

I make no judgment here about which of those claims are true; some of the underlying incidents are well documented and others are not. The point is the mechanism. A phrase repeated across hundreds of unrelated stories stops being a claim and becomes a frame: a ready-made box into which the next event is placed before anyone has checked what it was.

What it can and cannot do #

• The effect is reliable but modest per exposure. It matters most for statements people are unsure about, and it adds up over many exposures.
• It does not make absurd claims believable, and it does not usually overturn strong first-hand knowledge.
• It works for true statements too. Repeating accurate public-health advice makes it feel truer, which is why governments and campaigners use the same tool.
• It can be reduced. In 2020 Nadia Brashier, Emmaline Eliseev and Elizabeth Marsh found that asking people to judge accuracy when they first read a statement stopped the later effect, at least for people who had relevant knowledge. A study the same year with news headlines found a similar, smaller protection.

How to notice it #

1. Have I heard this before? If yes, ask whether I heard evidence before, or only the sentence.
2. Count voices, not copies. Fifty sites with one headline are one source.
3. Read past the question mark. A headline that asks “will X happen?” is repeating X.
4. Watch for phrases that appear in every story. Ask what they assume before the story begins.
5. On first contact with a claim, stop for two seconds and ask: is this true, and how would I know?

Frequently asked questions #

What is the illusory truth effect? #

It is the tendency to rate a statement as more true after hearing it before. It was first measured by Hasher, Goldstein and Toppino in 1977 and has been replicated many times since.

Does knowing the facts protect me? #

Only partly. Fazio and colleagues showed in 2015 that people fall for repeated false statements even on topics they know, because a familiar sentence feels true and they do not stop to check.

Do fact-check labels stop it? #

Not reliably. Pennycook, Cannon and Rand found in 2018 that a single exposure raised belief in fake headlines even when they were labelled as disputed.

Is repetition always manipulation? #

No. Teachers, doctors and safety campaigns repeat true things on purpose. Repetition is a tool. The question is what is being repeated and whether it was ever checked.

The lesson #

Familiar is not the same as true, but it feels the same. The cheapest defence is to count voices instead of repetitions, and to ask “is this true?” the first time you meet a claim, before the tenth time makes the question feel unnecessary. Next in the series: fear, and what it does to the way we think.