I started Phrase Finder in 1997. The site grew out of a computational linguistics research project I worked on at Sheffield Hallam University in the late 1980s, and over the last twenty-nine years it has become a reference for the meanings and origins of more than 2,100 English phrases, with a thesaurus of related-phrase lookups attached. The discussion forum that used to sit alongside it isn't open to new posts any more, but its 70,000-question archive is still on the site for anyone who finds it useful.
Every entry on the site has an author's name on it. Most of the meanings and origins are mine. The newer ones are by Cari Mayhew and David Saylor, who work to the same standards.
How we research origins
For each phrase the question is when it was first printed in English, and by whom. That sounds narrow, and it is, but it's the part of a phrase's history you can actually check. Theories about who really invented an expression, or what it's secretly an acronym for, are sometimes entertaining and almost always wrong. We try to find the earliest dated occurrence in print and let that speak for itself.
The reference shelf is mostly the Oxford English Dictionary, Green's Dictionary of Slang, Partridge's Dictionary of Slang and Unconventional English, Brewer's Dictionary of Phrase and Fable, and the Oxford Dictionary of Quotations. Beyond that, the actual work involves reading the source material itself: digitised newspaper archives, Early English Books Online, the British Newspaper Archive, Google Books, Project Gutenberg, and so on. Where a phrase has a known author we name them. Where it doesn't, we date it as best we can and say what we don't know.
Citation and folk etymology
Every origin claim on the site is tied to a dated source where the source exists. Quotations carry their author and year. Where the evidence is thin, or where there are competing accounts of where a phrase came from, the page says so.
A surprising number of well-known phrases have famous origin stories that fall apart on examination. 'Back to square one', 'the whole nine yards', and 'posh' are three of the more familiar examples. When a folk etymology is widely believed but isn't supported by the dated record, the entry says that plainly, and explains where the popular story came from. Repeating a confident-sounding origin story without testing it would be a lot less work; it would also produce a less honest site.
Revisions
Every entry shows the date it was first published and the date it was last updated. Phrase research isn't a fixed body of knowledge; a citation that pushes the earliest dated use back by fifty years can land in your inbox at any time, and when it does, the entry is revised and the new date attributed. A fair share of the past thirty years has gone into rewriting entries whose evidence has changed.
If you spot an error, or you have a citation that predates the one we have, the feedback link on each entry comes straight to me. Corrections are credited where the contributor wants to be named.
What we don't do
Phrase Finder is independently owned and isn't part of a content network. There are no commissioned posts and no sponsored entries. No page on the site is written to rank for a particular search term. Where the available evidence is thin, the entry says so plainly, and 'origin unknown' is a perfectly respectable answer for a phrase to have.