Poe the Codebreaker: 7 Facts About Edgar Allan Poe's Obsession With Secret Writing
On December 18, 1839, Edgar Allan Poe challenged the readers of a Philadelphia weekly to mail him ciphers, and he promised to solve every one. He solved most of what arrived and turned the hobby into "The Gold-Bug" in 1843. A cipher lesson built on his letter table later appeared in a U.S. Army signal manual, and one of his readers broke codes in two world wars.
1. He Dared a Newspaper's Readers to Stump Him
The challenge ran in Alexander's Weekly Messenger. Letters came in from around the country, and from January through April 1840 Poe printed solutions in nearly every issue.
Poe later said he had received about a hundred ciphers. Clarence Brigham's 1943 edition of the newspaper pieces counts thirty-six. Poe printed solutions to about two dozen, claimed three more without printing them, and admitted he had not solved six.
2. He Called a Real Cipher a Fake
One of the six came from a reader named Kulp. On February 26, 1840, Poe declared it an "imposition." In his 1841 account he described an unsolved cipher as "a jargon of random characters, having no meaning whatever."
The verdict protected his record. In Poe's retelling, one cipher in a hundred went unsolved, and that one was a hoax.
Poe was wrong. Kulp had used a real cipher with a shifting alphabet, keyed to the words UNITED STATES, and Brian Winkel published the solution in Cryptologia in 1977. The message asked why the Messenger reached Lewistown three days late. Kulp had broken Poe's rule of one symbol per letter, so the cipher fell outside the challenge, but it was no hoax.
3. He Claimed No Cipher Could Beat Human Ingenuity
In July 1841, as an editor at Graham's Magazine, Poe published "A Few Words on Secret Writing." Its central claim: "human ingenuity cannot concoct a cipher which human ingenuity cannot resolve."
The claim failed. One system cannot be broken: a truly random key, as long as the message, used a single time. It is called the one-time pad. Claude Shannon published the proof in the Bell System Technical Journal in 1949.
4. A Washington Doctor Built a Cipher to Beat Him
In the summer of 1841, Poe's friend Frederick W. Thomas forwarded a cipher built by Dr. Charles S. Frailey of Washington. Poe mailed back the answer, and Frailey confirmed it in a letter dated July 6.
Poe printed the cipher in the August issue. He offered a year's subscription to Graham's and to the Saturday Evening Post to the first reader who solved it. He published the solution himself in October, and that December he credited a reader, Richard Bolton of Pontotoc, Mississippi, with a correct answer. Poe noted in print that the cipher went well beyond the limits of his challenge.
5. Two Ciphers He Printed Took 150 Years to Solve
In December 1841, Poe printed two cryptograms from a "Mr. W. B. Tyler" and left them to his readers. Nobody solved them. The scholar Terence Whalen solved the first in the early 1990s and published the solution in 1994. John Hodgson solved it independently in 1993. It hid a passage from Cato, a 1713 play by Joseph Addison.
The second lasted until 2000. Gil Broza, a 27-year-old software engineer living in Toronto, won a $2,500 prize offered by Shawn Rosenheim of Williams College, according to the college's announcement. That cipher used several different symbols for each letter and carried more than two dozen errors made by the typesetters or by the person who wrote it.
Its message begins, "It was early spring, warm and sultry glowed the afternoon." Nobody has found the exact source. The Poe Society of Baltimore has traced the passage to a joke that ran in newspapers in 1840. Louis Renza of Dartmouth proposed in 1985 that "Tyler" was Poe himself. No evidence settles it.
6. "The Gold-Bug" Taught the Public, and Its Letter Table Reached the Army
Poe turned the hobby into fiction in 1843. "The Gold-Bug" follows a man who finds a scrap of parchment, solves the cipher on it step by step, and digs up a pirate's treasure. The story won a $100 prize from Philadelphia's Dollar Newspaper and ran that June. In May 1844 Poe claimed that more than 300,000 copies had been circulated.
Albert J. Myer founded the U.S. Army Signal Corps. In the 1868 edition of his Manual of Signals, he reprinted an unsigned Harper's Weekly article, "Curiosities of Cipher," to show officers how a cipher is broken. Its text matches an essay by the English clergyman Sabine Baring-Gould, later collected in his Curiosities of Olden Times (1869).
The article's worked example appears to be modeled on "The Gold-Bug." A man finds a slip of paper he believes will lead him to a hidden fortune. A friend solves it by counting symbols. The figure 8 stands for e, as it does in Poe's story.
The letter table gives the link away. Poe ranks the most common letters in English as e, a, o, i, d, h, n, r, s, t. Standard counts of English text put t second, and Myer says so himself in a note beside the table. The writer moved t up and kept the rest of Poe's order, including his omission of j and v. Poe's name appears nowhere in the article or in Myer's manual.
7. One of His Readers Broke Codes in Two World Wars
William F. Friedman read "The Gold-Bug" as a boy. In early 1918 he taught codebreaking to a class of 71 Army officers bound for France, according to the historian William Sherman. Friedman then served in France himself. He coined the word "cryptanalysis."
The Signal Corps created its codebreaking bureau, the Signal Intelligence Service, in 1929 and put him in charge. By the end of 1940 his team, led day to day by Frank Rowlett, had rebuilt Japan's PURPLE cipher machine without seeing one. That work let the United States read Japanese diplomatic messages during the Second World War. When the National Security Agency formed in 1952, he joined it as a senior consultant, and he retired in 1955.
Friedman credited Poe. Popular interest in cryptography in the United States, he wrote in 1936, "received its first stimulus from Edgar Allan Poe." In the same article he judged Poe a dabbler who still knew more than most of his contemporaries.
How Good Was Poe at Breaking Codes?
Poe was a fast solver of simple ciphers. The scholar W. K. Wimsatt studied the newspaper files and wrote that Poe's correspondents were "extremely naive." Wimsatt found no sign that Poe could attack the harder systems of his day. His letter is printed in Brigham's edition.
Readers can test the method on Poe's own cryptogram. The full text of "The Gold-Bug" is free at the Edgar Allan Poe Society of Baltimore.
Ciphers run all through Prophet, Bird or Devil, a novel by Nathaniel Free.
Sources
Edgar Allan Poe, "Enigmatical and Conundrum-ical," Alexander's Weekly Messenger, December 18, 1839.
Edgar Allan Poe, "A Few Words on Secret Writing," Graham's Magazine, July 1841, with the August 1841 follow-up and the December 1841 follow-up.
Edgar Allan Poe, "The Gold-Bug," Dollar Newspaper, June 1843.
Albert J. Myer, A Manual of Signals (new edition, 1868), pp. 322-336, reprinting "Curiosities of Cipher," Harper's Weekly, December 19, 1863.
Sabine Baring-Gould, "Curiosities of Cypher," in Curiosities of Olden Times (1869).
Clarence S. Brigham, Edgar Allan Poe's Contributions to Alexander's Weekly Messenger (1943).
W. K. Wimsatt Jr., "What Poe Knew About Cryptography," PMLA, September 1943.
Brian J. Winkel, "Poe Challenge Cipher Finally Broken," Cryptologia, January 1977.
Claude E. Shannon, "Communication Theory of Secrecy Systems," Bell System Technical Journal (1949).
William H. Sherman, "How to Make Anything Signify Anything," Cabinet no. 40 (2010-2011).
Williams College, announcement of the Tyler cipher solution, October 31, 2000.
Comments