The Cottingley Cutout Explanation Was Public in 1920
On this page
In July 1917 two cousins borrowed a camera in the village of Cottingley, West Yorkshire, and came back with a picture of a girl resting her chin on her hand while four winged figures danced in front of her. Elsie Wright was sixteen. Frances Griffiths was nine. The plate came from a Midg quarter-plate camera belonging to Elsie's father, Arthur Wright, and within three years the image had passed from a family joke into the pages of a national magazine, defended by one of the most famous storytellers in Britain.
What makes the case worth reopening is not the question of whether fairies were photographed. That was answered, publicly and correctly, in 1920. The interesting record is the distance between how quickly the right explanation appeared and how long the legend outlived it, and the museum that now holds the material is blunt about the mechanics: cutout drawings, held in place with hatpins.
What the Two Cousins Photographed in July 1917
The first image is usually called "Alice and the Fairies," and it was taken by Elsie in July 1917. A second followed the same year. Three more were made between 1918 and 1920 with a different camera, and two of those carry August 1920 dates: "Fairy Offering Flowers to Iris" and "Fairy Sunbath, Elves, etc."
The second camera matters more than it first appears. It was supplied by Edward Gardner, an associate of Arthur Conan Doyle, and that detail changes the character of the later pictures. The 1917 photographs were made by two girls with a borrowed camera and no audience. The 1920 photographs were made with equipment handed to them by adults who had already decided the earlier images were important and who were waiting for more.
The method behind the figures is no longer in dispute, because the people who used it described it. The fairies were drawings, cut out and secured in the ground or to trees with hatpins. Elsie had copied illustrations by Claude Shepperson from a widely circulated volume, Princess Mary's Gift Book, changing a few details and adding wings.
That last point deserves emphasis. The source of the figures was not obscure. It was a popular gift book of the period, sitting in ordinary households. Anyone who owned a copy and looked carefully at the photographs had the comparison available from the beginning.
The two sittings also differ in a way that is easy to miss and hard to unsee once noticed. In 1917 the cousins were photographing for themselves, and there is no indication that anyone expected the plates to leave the household. By 1920 the pictures had an audience with an investment in the outcome, and the request for more images came with the equipment attached. Nothing about that proves the later frames were made differently, but it does mean the two sets were produced under conditions that were not comparable, and treating all five as a single body of evidence flattens a real distinction.
The documented account of the case also records the pseudonyms that travelled with the photographs into print, and those matter for the same reason. Elsie became "Alice," Frances became "Iris," and the family was given the surname "Carpenter." A reader in 1921 who wanted to knock on a door and ask a direct question had no door to knock on.
The physical objects have survived unusually well for a case of this kind. The National Science and Media Museum in Bradford holds original prints, correspondence from both cousins, Elsie's watercolour sketches of fairies, and the cameras themselves, including the Midg that took the first two photographs in 1917.
How the Retelling Outgrew the Photographs
It is worth being precise about what those surviving cameras establish, because object provenance is often asked to carry more weight than it can. A camera in a display case shows that a particular instrument existed and was associated with the people named. It does not record what was placed in front of the lens. The Midg confirms the equipment; it says nothing about the paper figures, and the museum does not claim otherwise.
Conan Doyle did not discover the pictures. He heard of them through the editor of the spiritualist publication Light, and he was already committed to the subject before he saw them. He was then asked to write about fairies for The Strand Magazine, and the December 1920 Christmas issue carried two higher-resolution prints of the 1917 photographs. In 1922 he published The Coming of the Fairies.
He also concealed the participants. The girls appeared as "Alice" and "Iris," the family as the "Carpenters." The stated reason was privacy, and that is plausible, but the practical effect was to make verification harder for anyone outside the circle. A reader who wanted to check the account could not easily find the people in it.
It is worth stating plainly what Conan Doyle's involvement did and did not establish. He was a physician by training and an enormously effective writer, and neither qualification bears on whether a photographic plate has been staged. What he supplied was distribution and credibility, and the two compounded: the images reached hundreds of thousands of readers already framed as a serious question rather than as a curiosity, and the framing came pre-endorsed by a name those readers trusted for other reasons entirely.
The magazine context is part of the record too. A Christmas issue is not a neutral venue. It is the issue a general readership buys for pleasure, and a fairy photograph lands differently there than it would in a technical journal, where the first question would have been how the plate was exposed rather than what the picture appeared to show.
Set out that way, the shape of the case is clear. The photographs did not become famous because the evidence for them strengthened. They became famous because a celebrated author attached his name and a mass-circulation magazine gave them reach, and because the people who could have corrected the record had been made anonymous.
A Reader Named the Method in 1920
The single most under-reported fact in this file is that the correct explanation was published almost immediately, and not by an expert.
A sixteen-year-old correspondent suggested the cardboard cutout theory in 1920, the same year the photographs reached a national audience. That is a full sixty-three years before either cousin admitted anything, and it is essentially the answer: flat figures, cut from paper, placed in the scene.
What happened next is the part worth studying. A committee of expert photographers and scientific men took over the investigation from Elsie and Frances. The case moved from the hands of the two people who knew exactly how the pictures had been made into the hands of people who had to reason from the images alone, several of whom were already sympathetic to the conclusion they reached.
This is a common failure shape in contested-object cases, and it is not really about credulity. Expert examination of an artefact answers a narrow question: is there a visible seam, a shadow that does not belong, a support that shows. It cannot answer the broader question of whether someone simply made the thing. When the people who did make it are removed from the inquiry and replaced by specialists studying the output, the inquiry can be rigorous and still miss the point entirely.
The museum's account of the period is careful to separate the two kinds of scrutiny that were applied. Its survey of how the images were argued about notes that the cutout proposal came from public correspondence rather than from the technical committee, which is an uncomfortable detail for anyone who assumes expertise reliably outperforms an attentive amateur on questions of this shape.
There is a further asymmetry that the file makes visible. Confirming a photograph is genuine requires ruling out every method by which it could have been staged, which is close to impossible. Showing that it was staged requires identifying one workable method, which a sixteen-year-old managed from a printed reproduction. The burden was never symmetrical, and the investigation of 1920 to 1922 proceeded as though it were.
Readers who have followed how an object's own custody record can complicate its story will recognise a related problem in The Lewis Chessmen Preserve Several Sets but No Certain Owner, where what survives is abundant and the question of origin still resists a clean answer.
What Elsie Wrote to Geoffrey Crawley in 1983
The cousins kept the secret for more than sixty years. The break came in 1983, when Elsie wrote to the photographic journalist Geoffrey Crawley and stated that all five photographs were faked. An article in the magazine The Unexplained carried the admissions to a wider readership the same year.
Elsie's account was specific rather than general. She named the source of the drawings, described altering details and adding wings, and explained the cardboard cutouts supported by hatpins. This is what distinguishes the Cottingley file from most disputed-object cases: there is a confession, it comes from a named participant, it is detailed enough to check against the images, and it matches what an outside observer had already proposed in 1920.
The material record of that admission survived too. When Crawley's Cottingley holdings were sold in 1998, the collection that went to Bradford included a nine-page letter from Elsie admitting the hoax. The museum has continued to assemble the objects since; it acquired the last camera in the story in April 2019, a folding quarter-plate Cameo made by W. Butcher and Sons, for three thousand four hundred pounds, described as one of two given to the cousins by Conan Doyle in 1920.
The custody trail is itself worth reading closely, because it is the part of the case where the surviving documentation is strongest. Geoffrey Crawley was the editor of a photographic trade journal, not a paranormal investigator, and his correspondence with Elsie in the early 1980s is the channel through which the admission reached print. That the letters, the watercolours, the prints and the cameras all ended up in one public institution is unusual; most disputed-object files are scattered across private hands and lose exactly the material that would settle them.
Even so, the record is not seamless, and it is better to say so than to smooth it. Accounts differ over precisely which physical items passed into which hands and when, and the museum's own listing of its holdings does not match every secondhand summary of the collection. That is a small discrepancy against a well-documented case, but a file that cannot state its own custody chain without variation is not a file whose finer details should be repeated with confidence.
One further asymmetry runs through the whole file and is worth naming. The two cousins were children when the first photographs were made, and adults with reputations attached by the time anyone asked them to account for it. Sixty years of public investment in the story sat on top of an admission that would have cost them nothing in 1918 and a great deal by 1970. That is not an excuse for the delay, but it is part of why the delay happened, and it belongs in an honest reading of why the correct answer waited so long to be accepted.
It is worth being careful about one thing here, because retellings blur it. Both women said they had genuinely seen fairies. That statement is about their experience and is not verifiable either way. It sits in a different category from the photographs, which are a manufacturing question and are settled.
The One Frame Frances Never Withdrew
The cousins did not agree, and the disagreement is narrow and specific. Elsie said all five were fabricated. Frances maintained that the final photograph, "Fairy Sunbath, Elves, etc.," was genuine, and she held that position until her death.
There is no test that settles this. The image is of the same kind as the other four, made in the same place by the same two people during the same short period, and the person who says it is real is the person who would know. So is the person who says it is not. Two participants, one object, and opposite testimony is the hardest evidentiary situation there is, because adding examination does not break the tie.
It is tempting to resolve the disagreement by weighting the two statements differently, and the usual move is to prefer the fuller confession over the narrower holdout. That instinct is reasonable and still worth resisting. Elsie's account is detailed and checkable for four frames; her claim about the fifth is a claim about what her cousin did, which she was not necessarily in a position to observe. Frances is asserting something about her own conduct. Neither has the standing to close the other's testimony, and no amount of examining the print supplies the missing witness.
The museum's own framing is careful on the surrounding material as well. Of the camera it acquired in 2019, it notes that the photograph potentially taken with it has not been conclusively proven. That is the correct register for this file: the manufacturing question is closed, and several smaller attribution questions are not.
Cases where the record answers the headline question but leaves the participants unaccounted for recur in this archive, most recently in The Kensington Runestone's Carver Has Never Been Named. Cottingley is the inverse arrangement, and the comparison is instructive: there the object is judged and the maker is unknown; here the makers spoke, and one of them declined to close the last frame.
The case has since become a standard teaching reference, and the reason is that it separates two things that ordinarily arrive fused. The manipulation itself was crude by any standard, then or now, and required no darkroom skill at all. The persuasion was sophisticated, and it operated almost entirely outside the photographs, through who vouched for them, where they were printed, and how difficult it had been made to reach the people in the frame. A file that needs a technical explanation for the trick and a social one for the belief is more useful than a file where the trick was clever.
What the Cottingley photographs finally document is not the existence of fairies but the durability of a story that was contradicted in public within months of becoming famous. The cutouts were named in 1920. The confession arrived in 1983. Sixty-three years sat between an available answer and an accepted one, and nothing in the physical evidence changed across that gap. Only the willingness to look at what had already been said did.