The Dorabella Cipher: 87 Squiggles a Great Composer Never Explained
In the summer of 1897, a thank-you note arrived at the Wolverhampton rectory of the Reverend Alfred Penny. The composer Edward Elgar and his wife Alice had just spent a few days there as guests, and the letter, dated 14 July, was ordinary politeness. What travelled with it was not. Tucked inside was a small card addressed to the reverend's daughter Dora, not yet twenty-three, covered in three fluid lines of curling, hook-like symbols: eighty-seven of them, dashed off in a single confident hand. Dora turned the card over and over, looking for a way in. She could not read a word of it. Neither, in the 129 years since, has anyone else.
At the time none of it looked like history. Elgar was forty, a provincial music teacher who had not yet written the work that would make his name; the Enigma Variations lay two years ahead, the knighthood eight. Dora Penny, born in 1874, was almost seventeen years his junior, the lively and musical daughter of a country clergyman. The two had taken to each other at once and would remain friends for the rest of the composer's life.
The friendship left other traces. Elgar took to calling Dora Dorabella, borrowing the name of the flighty young woman in Mozart's Cosi fan tutte, and when he came to write his Enigma Variations he gave that name to the tenth, a fluttering woodwind miniature long heard as a portrait of Dora's light stammer and quick laughter. She later married and became Mrs Richard Powell, and it was under that name that she eventually set the cipher before the public. She maintained to the end that she had never known what it said, and that Elgar, who teased and tested her for decades, had simply assumed she always would.
The card itself, known today as the Dorabella Cipher, carries 87 characters drawn from an apparent alphabet of about 24 symbols, each built from one, two or three semicircular arcs rotated into one of eight orientations. One further mark appears on it: a tiny dot after the fifth character of the third line. Word divider, deliberate signal, or slip of the pen, no one can say, and even that dot has generated pages of scholarly argument. The writing flows as easily as ordinary handwriting, and the fluency is itself a clue, because fluency implies practice. For an artefact so slight, a single card no larger than a postcard and three lines of ink, it has drawn an astonishing weight of attention.
He had, in fact, been using the alphabet for years. At a concert in April 1886, more than a decade before Dora's card, he annotated a programme with eighteen of the same arc-shaped characters, a scrap Elgar scholars now call the Liszt fragment. In the 1920s, an old and celebrated man by then, he was still drawing the same alphabet into a notebook as clock-face diagrams, mapping arcs to positions like hours on a dial. The symbols were a private instrument he returned to across four decades.
To understand the card you have to understand the man's appetite for concealment. Elgar salted his letters with puns, misspellings and private nicknames, kept a lifelong fascination with codes and acrostics, and built his most famous work around a theme he swore was never sounded, the hidden enigma he hinted ran counter to the music and then declined, forever, to name. When John Holt Schooling published a challenge cipher in the Pall Mall Magazine in the 1890s and pronounced it unbreakable, Elgar broke it, and was proud enough of the feat to preserve his workings. A man like that does not encipher a message by accident.
By Dora's own account she asked him about the card years later, and he seemed genuinely surprised she had never worked it out, yet still told her nothing. She kept it in a drawer for four decades and finally published it in 1937, in her memoir Edward Elgar: Memories of a Variation, three years after his death in February 1934. By the time the world was invited to try, the only man who could confirm any answer was gone.
The technical heart of the problem is deceptively simple. If the card were a plain substitution, one symbol standing consistently for one letter, the arcs should fall into the lopsided pattern any stretch of English falls into, a scatter dominated by the letter E. They do not. The commonest symbols appear too seldom, the rare ones too often, and sequences repeat where English would not. That single stubborn fact is what defeats the tidy readings and pushes researchers toward stranger explanations. Frequency analysis is the first tool any codebreaker reaches for, and on this card it fails immediately.
Amateur cryptographers, professional musicologists, statisticians and hobbyists armed with software have all taken their turn, and the card has outlasted every one of them. When the Elgar Society ran a competition in 2007 for the composer's 150th anniversary, every entry, after matching symbols to letters, ended in what the judges called an essentially arbitrary string, and no winner was declared. The most rigorous recent word belongs to Viktor Wase, whose 2023 computational study concluded that the cipher is unlikely to be a simple one-to-one substitution in English or Latin, a result that tells us only what the Dorabella Cipher is not. The card has since been photographed, digitised and fed through software its author could not have imagined, without result.
So the card sits in the small, maddening company of unbroken texts alongside the Voynich manuscript and the Zodiac killer's messages, with one crucial difference. Here the author is known, his life exhaustively documented, one of his other codes already cracked by his own hand. The setter is not an anonymous stranger but one of the most studied composers in English music, which makes the silence of his 87 marks feel less like an accident of history than a decision. He went to his grave holding two secrets: the theme behind the Enigma Variations, and whatever lives in those 87 curling marks.
Conclusions and Open Questions
The proposed solutions now fill a small library, and each collapses in its own way. In 1970, writing in The Musical Times, the musicologist Eric Sams, who had unravelled Schumann's ciphers, read the card as encoded phonetic English beginning "STARTS: LARKS! IT'S CHAOTIC, BUT A CLOAK OBSCURES MY NEW LETTERS" and closing "...WHEN YOU ARE TOO LONG GONE." His pedigree is the reading's strength; arithmetic is its weakness, for the solution extracts 109 letters from 87 symbols, and a method that elastic yields almost anything. The Australian academic Tim S. Roberts later offered a straightforward substitution backed by statistical reasoning; the rigour is admirable, but the resulting text, wandering through drooping beige weeds and "pure idiocy," is barely English.
Others reject the assumption that the card is English at all. In 2009 the cipher researcher Tony Gaffney argued that the symbols record a private idiom, the backslang, contractions and in-jokes of a real friendship, which would explain both the misbehaving letter frequencies and why only Dora could ever have unlocked it. Its weakness is fatal in a different way: it can never be checked, for a key shared by two people now dead is no key at all. Javier Atance proposed that the message is not language but music, the eight orientations mapping onto notes of a scale and the number of arcs onto natural, flat or sharp; a natural thought from a composer, though the tune has persuaded no one. In 2020 Wayne Packwood claimed a complete decryption in Musical Opinion, a hidden secondary message, "RATS," included; scholars remained unmoved.
There is also the uncomfortable possibility that the card is not really a cipher at all. A true cipher hides a fixed message behind a consistent rule, but Elgar could as easily have used his arcs as a loose personal shorthand, spelling by ear, dropping vowels, letting one squiggle stand for whatever the moment suggested. If so, no single key is waiting to be found, because there was never a single rule, only a private habit that lived in one man's hand. A related theory holds that the key was never textual but personal: some joke or phrase from those Wolverhampton evenings that Dora was expected to recognise rather than decrypt. Nothing in the record supports this directly, but it would explain the strangest detail we have, Elgar's surprise.
Some argue that the card was never meant to be read. A married composer of forty sent a coded private note to a witty young woman of twenty-two, folded inside a thank-you letter to her father, and a cipher is the perfect vessel for such a gesture: intimate if decoded, deniable if not, flattering either way. Nothing in the documented record establishes anything improper, and the reading remains speculation. Others point to a quieter possibility. Even a clean solution may prove an anticlimax, a fragment of gossip or a line about the weather at the rectory, because difficulty and content are separate things, and a puzzle-lover may encipher something small precisely because the pleasure lies in the enciphering.
What remains unexplained is short and stubborn. Why the alphabet holds roughly 24 symbols when written English needs 26. What the single dot in the third line is doing. Why a man who had beaten a professional's unsolvable challenge would expect an untrained rector's daughter to break his own. And above all why, in the thirty-seven years left to him, he never supplied the key. Every few years a new solution announces itself, confident and complete, and the little card refuses it, as it refused all the others, as it refused Dora.