Short answer: “experts say” without names is a weasel phrase: it borrows authority while making the claim impossible to check. Ask which experts, how many, expert in what, who funds them and what exactly they said. Even a precise number can mislead: the UK ad watchdog ruled against Colgate’s “more than 80% of dentists recommend” claim because the survey let dentists recommend several brands.
“Experts warn.” “Scientists say.” “Doctors recommend.” “Research suggests.” These phrases appear in news, ads, social media posts and, increasingly, in AI-generated answers. They sound like evidence. Very often they are the opposite: a way of borrowing authority without telling you whose authority it is, so that nobody can check it.
Why vague experts are a problem
Linguists and editors have a name for this: weasel words. Wikipedia describes them as words and phrases that create the impression that something specific and meaningful has been said, when only a vague or ambiguous claim has been communicated. Typical examples include “some people say”, “it is thought” and “researchers believe”, and the article notes that such phrasing is used in advertising, popular science, opinion pieces and political statements to mislead or disguise a biased view or an unsubstantiated claim.
Wikipedia’s own style guide lists phrases its editors should avoid without a source: “some people say”, “many scholars state”, “experts declare”, “research has shown”, “science says” and similar. The reason is practical: an unnamed authority can’t be verified, challenged or weighed against other authorities.
- You can’t check expertise. An “expert” in nutrition might be a professor of epidemiology or someone who completed a weekend course.
- You can’t check numbers. One expert and a thousand experts look the same in “experts say”.
- You can’t check interests. An “industry expert” may work for the company whose product the article praises.
- You can’t check context. The expert may have said something more cautious that was later shortened.
- The author can always back away. Vague wording leaves room to deny any specific meaning if the claim is challenged.
Case study: “More than 80% of dentists recommend…”
A well-known example comes from the UK. A Colgate poster claimed “More than 80% of dentists recommend Colgate”. The Advertising Standards Authority (ASA) received complaints that this implied dentists recommended Colgate over other toothpastes. The problem lay in the survey behind the claim: it allowed dentists to recommend several toothpastes and brands and didn’t ask them to state a preference for any one of them. The ASA found that the claim suggested 80% of dentists would recommend the brand over others while the remaining 20% preferred a different one, which the survey didn’t show, and the claim had to go.
A dictionary of vague authority
| Phrase | What it may hide | Ask |
|---|---|---|
| “Experts say” | One person, an unnamed source, or the author’s own view | Which experts, how many, expert in what? |
| “Scientists have found” | One small study, a preprint, an animal study | Which study, published where, on whom? |
| “Studies show” | A single study or a cherry-picked selection | Is there a systematic review? (see our guide) |
| “Doctors recommend” | A paid ambassador, a survey with leading questions | Which doctors, in what specialty, who funded it? |
| “It is widely believed” | A popular opinion, not evidence | Believed by whom, and is it true? |
| “Sources close to…” | Legitimate anonymous sourcing or speculation | Does the outlet explain why the source is anonymous and confirm it independently? |
| “A leading expert” | Someone leading in a different field | What did they publish on this exact topic? |
| “According to a recent survey” | An online poll of the brand’s customers | Who ran it, how many people, which questions? |
When an unnamed source is acceptable
Not every anonymous source is a red flag. Investigative journalism sometimes protects whistleblowers or people at risk, and that can be necessary to report important stories. The difference lies in transparency about the method (our assessment, based on common newsroom practice):
- A credible outlet explains why the source isn’t named (e.g. not authorised to speak, risk of retaliation).
- It describes the source’s position as precisely as possible: “a senior official in the health ministry with direct knowledge of the negotiations”, not just “insiders”.
- It confirms key claims with documents or a second, independent source.
- It takes responsibility for the claim and corrects it publicly if it turns out wrong.
“Experts say” in a lifestyle article or an advert has none of these safeguards. It’s a shortcut, not a protected source.
Four tricks with experts
- Wrong field: a physicist commenting on vaccines or a celebrity doctor on investing. Expertise doesn’t transfer automatically.
- The lone dissenter: one expert presented as “the scientists”, while hundreds of others say the opposite. Ask about consensus, not just a quote.
- False balance: one expert for and one against, presented as a 50/50 split when evidence is overwhelmingly on one side.
- Undisclosed interests: an “independent expert” who consults for the company, sells a course or represents a lobby group.
How to check “experts say”: 7 steps
Ask: which experts?
Look for names, affiliations and the exact field of expertise. If the article gives none, treat the claim as unverified.
Check the expertise
Search for what the person has published or worked on in this specific area, not just their title.
Count them
One expert, a survey of 200, or a consensus statement from a professional body? The weight is very different.
Find the original statement
Look for the full interview, paper or report. Shortened quotes often lose their caveats.
Look for conflicts of interest
Who pays the expert or funded the study? Is the expert affiliated with the product, party or lobby in question?
Check the consensus
Search for reviews or positions of professional bodies. Is this expert in the mainstream or a lone voice?
Decide what you know
If you still can’t identify the experts, treat the claim as an opinion of the author and look for a source you can verify.
Rewriting vague claims (examples)
A useful exercise is to imagine how a claim would look if it were fully transparent. If the article can’t be rewritten this way, the information to do so is missing (illustrative examples by us):
| Vague | Verifiable |
|---|---|
| “Experts warn that screens harm children’s sleep.” | “A 2024 review of 40 studies by researchers at University X found an association between evening screen use and shorter sleep in children aged 6–12.” |
| “Doctors recommend our supplement.” | “In a survey of 150 GPs commissioned by the manufacturer, 62% said they would consider recommending the product to patients with a confirmed deficiency.” |
| “Scientists say the diet works.” | “One 12-week trial with 48 participants found greater weight loss than the control diet; no long-term studies exist yet.” |
Notice how much the second column changes your judgment, even though the first column isn’t technically false.
Frequently asked questions (FAQ)
Because it presents a claim as authoritative without letting you check who said it, what their expertise is, how many of them there are or whether they have conflicts of interest.
Words and phrases that create an impression of a specific, supported claim while saying something vague, such as “some people say”, “it is thought” or “research has shown” without a source.
No. Serious journalism sometimes protects sources, but it explains why, describes the source’s position precisely and confirms key claims independently.
The UK Advertising Standards Authority found that the claim misleadingly implied dentists preferred Colgate over other brands, while the survey had allowed them to recommend several toothpastes.
Look for publications, positions and experience in the specific field being discussed, and check whether professional bodies or peers recognise their work.
Presenting two sides as equally supported, for example one expert for and one against, when the evidence or expert consensus is overwhelmingly on one side.
Ask it to name the experts and sources, then verify them independently. Chatbots can produce authoritative-sounding phrases without real sources.
Yes, it’s a close relative. One study is not the same as a body of evidence. Check whether a systematic review exists.
Yes. The way the question was asked, who was surveyed and who funded the survey can change what the number actually means.
With full names, affiliations, relevant expertise, a link or reference to the original statement or study, and disclosure of any conflicts of interest.
Get the “How to Search?” guide
Practical methods for searching and verifying information, step by step, with dozens of examples.
See the guide → ☕Support us on Buy Me a Coffee
We create these materials in our spare time. Every bit of support helps us write more, and more often.
Support → ▶️Watch us on YouTube
The same topics in a shorter video format, with concrete examples on screen.
Subscribe →More from our blog
Sources: Wikipedia: Weasel word · Wikipedia: Manual of Style/Words to watch · Osborne Clarke: Colgate’s 80% of dentists recommend claim under fire · Cosmetics Business: ASA don’t recommend Colgate. The dictionary of phrases, the examples of rewritten claims, the list of tricks and the steps are our own analysis.