404·Quotes

Editorial policy

How we check

This museum makes one claim: that the name under each quotation has been looked at. That claim is worth nothing unless you can see how it is decided, so here is the whole method.

The five labels

Every quotation we have examined gets exactly one of these. The label is the product; the quotation is only the occasion for it.

The three tiers of source

Not all evidence is equal, and pretending otherwise is how misattributions spread in the first place. Every source we cite is tagged with its tier.

Primary is the work itself — the book, the letter, the speech, the recording. Reference is a scholarly compilation or a specialist investigation of the attribution. Reporting is journalism about the question.

Every investigation in the Forgery Wing cites at least one primary or reference source. Reporting alone is never sufficient — a newspaper repeating an attribution is evidence that it is popular, not that it is true. Across the 17 investigations there are 52 sources.

What we do when we have not checked

We show nothing. A quotation in the galleries with no verdict beside it has not been examined, and we would rather say less than guess. The alternative — printing a confident name on everything — is precisely the habit that produced the Forgery Wing.

Where a line cannot be traced but is universally credited to someone, the byline itself says Attributed to. That is a statement about the evidence, not about the person.

Why the galleries do not display verdicts

Reading a quotation is not a forensic activity, and a badge beside every line would be this site auditing itself in front of you. The checking still happens — it decides what is published — but the apparatus lives in the Forgery Wing, in the Restoration Room, and at the desk, where you can look anything up.

The exception is a correction that is genuinely interesting in itself. “Fortune favours the bold” is almost always printed as an anonymous Latin proverb; it is Virgil, and it has a line number. That is not an audit result, it is a fact worth knowing, so it appears on the card in ordinary prose.

Where these quotations came from

The collection was assembled from four English-language quotation sites, all of them ours, and then re-checked rather than trusted. That process removed quotations outright — “be the change you wish to see in the world” was sitting in a themed collection until the audit caught it — and corrected the name on several others, including lines credited for years to Aristotle, Confucius and Plato that belong to Will Durant, Oliver Goldsmith and Ian Maclaren.

Translations

Hindi and Chinese renderings shown beneath a quotation were written in-house and are ours. They are offered as renderings, not as scholarly translations, and the English line is always the one being examined. We do not present a modern translation of an old text as though it were the original: popular English versions of Rumi, in particular, are modern copyrighted works rather than ancient wording.

Corrections

If we have a verdict wrong, the page is corrected and the correction is described on the page rather than removed quietly. A site that exists to catch misattribution has no business editing its own record in silence. Each investigation carries the date it was last reviewed.

The current tally

650 quotations are on display across the galleries. 24 have been fully investigated: 12 misattributed, 3 misquoted, 2 verified in the Forgery Wing, plus 7 incomplete quotations restored. That ratio is not a failure of the collection — most famous quotations are shakier than people assume, and finding out which ones is the entire purpose of the building.

How do you decide whether a quote is real?

Each quotation is checked against primary sources where they exist, then against reference works, then against reporting. A quote is only labelled sourced when it can be traced to a named work with a chapter, line or date. Where no primary source can be confirmed, the label says attributed. Where nothing has been checked yet, no label is shown at all rather than a guess.

What counts as a source?

Three tiers. Primary is the work itself. Reference is a scholarly compilation or a specialist quotation investigator. Reporting is journalism about the attribution. Every investigation in the Forgery Wing cites at least one primary or reference source; reporting alone is never enough.

What happens if you get one wrong?

The page is corrected and the correction is described on the page rather than removed quietly. A site that claims to check attributions has no business editing its own record silently.