The Cottingley Fairies
Five photographs taken by two girls in a Yorkshire village in 1917. Arthur Conan Doyle published them as evidence that fairies exist. The figures were traced from a gift book they had at home.
Why this verdict
Debunked by the photographers: in 1983 Elsie Wright and Frances Griffiths said the fairies were paper cut-outs copied from a 1914 children's book and held up with hatpins — a conclusion Geoffrey Crawley's photographic analysis had reached the previous year independently.
- Period
- Region
- Reading time
- Sources
01The claim
That five photographs taken by Elsie Wright, sixteen, and her cousin Frances Griffiths, nine, at Cottingley Beck near Bradford in 1917 and 1920 show real fairies and a gnome — and, in the use made of them, that photography could supply objective evidence of a spirit world.
02Where it came from
The girls borrowed Elsie's father's camera to prove they had been playing with fairies, which the household did not believe. The plates would probably have stayed a family curiosity had Elsie's mother not taken them to a Theosophical Society lecture in 1919. Edward Gardner took them up, and through him they reached Arthur Conan Doyle, who was researching an article on fairies for The Strand. Doyle published the first two in the Christmas 1920 issue and the full set in The Coming of the Fairies in 1922. His authority — the creator of the most rigorously deductive character in English fiction — was the whole of the case's credibility.
03What the record shows
✕ DEBUNKED — Investigated and refuted. Here is what actually happened.
The photographs are not subtle. The figures are flat, sharply outlined, lit inconsistently with the girls beside them, and dressed in a fashion contemporary to 1917. Doyle's own experts were divided at the time; Kodak declined to certify them as genuine.
The specific source was identified later: the figures correspond closely to Claude Shepperson's illustrations in Princess Mary's Gift Book, published in 1914 and in wide domestic circulation. Geoffrey Crawley, editor of the British Journal of Photography, published a detailed technical study across 1982 and 1983 concluding the images were cut-out fakes.
Elsie and Frances, then in their eighties, confirmed it. They had drawn the figures, stiffened them with card and held them in place with hatpins — a hatpin point is visible in one image, which Doyle had interpreted as a fairy's navel. Both said the hoax had begun as an ordinary childish attempt to avoid being told off and had escaped them entirely once Doyle became involved; they had not known how to stop it without humiliating a knight of the realm and a bereaved father.
Frances maintained to the end that the fifth photograph, of fairies in a bower, was genuine. Elsie said all five were faked. The archive records the disagreement and does not treat it as unresolved: the fifth photograph shares the technique and characteristics of the other four.
04What it cost
The case is usually told as an embarrassment for Conan Doyle, and it is one — but the interesting failure is the general one. The photographs succeeded because a mass audience in 1920 had no working intuition for what a photograph could be made to do, and because the medium's authority had run ahead of the public's ability to interrogate it. That gap has reopened on a considerably larger scale, and the Cottingley photographs are the clearest historical case of what happens inside it: not that people are stupid, but that a new evidentiary medium is trusted for a while at exactly the level of confidence it does not deserve.