THE CODEBREAKERS

by DAVID KAHN

  Poe quit Alexander’s in May of 1840. A year later, when he had become editor of Graham’s Magazine in Philadelphia, he found an opportunity to exploit the same sure vein of journalistic interest that he had struck in Alexander’s. Reviewing Sketches of Conspicuous Living Characters of France in the issue of April, 1841, he offered, in almost the same words he had used in Alexander’s, to read cryptograms that readers might send him:

  In the notice of [the lawyer Antoine] Berryer it is said that, a letter being addressed by the Duchess of Berry to the legitimists of Paris, to inform them of her arrival, it was accompanied by a long note in cypher, the key of which she had forgotten to give. “The penetrating mind of Berryer,” says our biographer, “soon discovered it. It was this phrase substituted for the twenty-four letters of the alphabet—Le gouvernement provisoire!”

  All this is very well as an anecdote; but we cannot understand the extraordinary penetration required in the matter…. anyone who will take the trouble may address us a note, in the same manner as here proposed, and the keyphrase may be in either French, Italian, Spanish, German, Latin, or Greek (or in any of the dialects of these languages), and we pledge ourselves for the solution of the riddle. The experiment may afford our readers some amusement—let them try it.

  The keyphrase cipher, which apparently has never been used in practical cryptography before or after the Duchess of Berry,* uses the keyphrase as a substitution alphabet:

  The French plaintext vraiment (“truly”), which has no repeated letters, would become OOLNEUNI. Thus, O represents both v and r, and N both i and n. The keyphrase produces a polyphonic substitution—one in which a given ciphertext letter may stand for two or more different plaintext letters, and which consequently may create some decipherment ambiguities. Its ciphertext may exhibit such dismaying peculiarities as three or four identical ciphertext letters in a row. But any difficulties that this may occasion the cryptanalyst are counterbalanced by the coherence of the keyphrase, which he reconstructs along with the ciphertext.

  While waiting for responses to his challenge, Poe wrote “A Few Words on Secret Writing,” the longest of his writings on cryptology, for the July Graham’s. It comprised a medley of cryptographic information, presented with pace and vigor but giving little new. It did include a dictum that he had promulgated in Alexander’s but that here assumed the form in which it became classic in cryptology—long considered a truism but now known to be false: “It may be roundly asserted that human ingenuity cannot concoct a cipher which human ingenuity cannot resolve.” Most of the article’s “scraps of erudition,” as critic Wimsatt well put it, came straight from the Encyclopaedia Britannica, and perhaps also from the Encyclopaedia Americana. Poe faithfully reproduced the Britannica’s error in abbreviating Giovanni Battista Porta’s name as “Cap. Porta.” Poe here gave cryptology its first discussion of skytale cryptanalysis: wrap the strip of parchment around a cone and slide it up and down until sense appears; the diameter of the cone at that point is the diameter of the skytale. But he still did not say anything about how to solve the monalphabetic substitutions; he still did not disclose the secret for which his readers had clamored. Indeed, he deepened the mystery. He concluded the article by stating (incorrectly) that one could find in writings on cryptology no rules for solving ciphers that one did not “in his own intellect possess.”

  On July 1, a friend in Washington, F. W. Thomas, a novelist who held a position in the Treasury Department and who was helping Poe in his efforts to get a government job, forwarded two keyphrase cryptograms from a friend of his who had accepted Poe’s challenge in the April Graham’s. Poe solved one of them at once; the other had the same keyphrase: BUT FIND THIS OUT AND I GIVE IT UP. It was the most difficult he had ever solved, owing to its plaintext, part of which read: Without dubiety incipient pretension is apt to terminate in final vulgarity, as parturient mountains have been fabulated to produce muscupular abortions. Elatedly, he sent the solution to Thomas “by return of post” on July 4, asking for testimonials. These he used in the August issue of Graham’s, in which he published the text of the cryptogram and offered a year’s subscription to Graham’s and The Saturday Evening Post to the first solver. A few days after sending this number to press, he received another cipher from Thomas, this one from the son of the Secretary of the Treasury, to whom Thomas had been speaking of Poe. “Decypher Mr. P. Ewing’s cryptograph in your August number if you can—Let me have it by return of mail,” Thomas wrote. But it did not yield to Poe’s analyses, and Poe, who up to this point had apparently depended solely upon rules that he already possessed in his own intellect, and who might have been helped in getting a job he wanted if he solved the Ewing cipher, sought new sources of information.

  Several times before, he had had good results with The Cyclopaedia of Abraham Rees, finding materials for articles on Palestine and Stonehenge. Now he turned to it again and this time discovered the superb article on “Cipher” by William Blair, an English surgeon. It ran 30 pages and about 35,000 words—the length of a small book—and for almost a century, or until Parker Hitt wrote his Manual for the Solution of Military Ciphers in 1916, it remained the finest treatise in English on cryptology. Blair had done considerable original research among manuscripts for historical material; he described thoroughly the major systems of cryptography and gave the bases of cryptanalysis and a number of examples—including attempts taken (with credit) from Falconer, to solve polyalphabetics. On the back of the envelope of Thomas’ letter, Poe copied a variety of linguistic observations—“y seldom in middle of word”—plus what appears to have been his first and only frequency count, though from his work with printers he must have already been aware that letters differ in their frequency of use.

  In giving the frequencies, Blair had divided vowels from consonants and had ranked the frequencies of the consonants in four groups, in each of which he listed the individual consonants alphabetically: “To find out one consonant from another, you must observe the frequency of d, h, n, r, s, t; and next to those, c, f, g, l, m, w; in a third rank may be placed b, k, p, and lastly, q, x, z.” As for the vowels, “you will generally find e occur the oftenest; next, o, then a, and i; but u, and y, are not so often used as some of the consonants, especially s and t.” In compressing Blair’s information, Poe ignored Blair’s distinctions, erroneously transposed a and o, and listed as the “order of frequency e a o i d h n r s t u y c f g l m w b k p q x z.” Both Blair and Poe omitted j and v.

  This help, however, did not enable Poe to solve the Ewing cipher, which was very short. Nor did he use this frequency count for notes on cryptology in the October and December issues of Graham’s, though he did make use of much of Blair’s information—without ever giving credit to the source. In fact he made it appear as if the knowledge were his. For example, on the basis of the Blair-Falconer discussion of polyalphabetic solution, he wrote in December regarding the Vigenère tableau: “Out of a thousand individuals nine hundred and ninety-nine would at once pronounce this mode inscrutable. It is yet susceptible, under peculiar circumstances, of prompt and certain solution.” Poe’s readers no doubt drew the conclusion that he wanted them to draw about just who could effect that solution.

  Cryptology had made the biggest hit of any of his journalistic writings. A story based upon the sure-fire topic of a secret message seemed a natural, and a story explaining the technique of cryptanalysis—the mystery with which he had teased his readers for two years—seemed a certain success. Half a dozen years before, he had reviewed Robert M. Bird’s Sheppard Lee, whose hero frantically searched for the legendary treasure of the pirate Captain William Kidd. Poe remembered a comical Negro servant from the novel, and this brought back his Army days in the South in 1828, when he served at Fort Moultrie on Sullivan’s Island, outside Charleston, South Carolina. Sullivan’s Island became the locale of the story and, recalling some natural history studies with an acquaintance there, Poe combined the click beetle, Alaus oculatus, with its death’s-head spots, an
d the gold beetle, the gleaming Callichroma splendidum, into the gold-bug that gave its name to the story he wrote.

  In spite of the story’s length, George Rex Graham, the publisher of Graham’s Magazine, snapped it up. He paid only a minimum price, however, and when Poe heard that the Dollar Newspaper was offering a $100 prize for the best story, he got “The Gold-Bug” back from Graham and entered it in the competition. (Unable to return Graham’s money, he had to make it up with a series of reviews.) “The Gold-Bug” won the prize and was published in two installments in the Dollar Newspaper on June 21 and 28, 1843. It made an instant hit. So great was the demand that it was reprinted in the Saturday Courier, and then reprinted again in the Dollar Newspaper. Playwright Silas S. Steele adapted it, and on August 8 it was given as a curtain raiser at the Walnut Street Theatre in Philadelphia. The tale was—and is—by far the most popular of any of Poe’s stones. In 1845, it was published in book form for the first time, in Tales. Poe revised the plaintext slightly—changing a forty to a twenty—and corrected the ciphertext to agree; but, human as he was, he forgot to make the appropriate changes in his frequency count. The Tales count also omits one character, the ( representing r. Most editors have reproduced these slightly erroneous figures. In November of 1845, a French translation by Alphonse Borghers appeared in La revue britannique, and in 1856 the great French symbolist poet whom Poe had so influenced and who in turn influenced so much of French poetry, Charles Baudelaire, published his translation as “Le Scarabée d’or.” “The Gold-Bug” was both the climax and the conclusion of Poe’s cryptologic writings; he published nothing more on the subject, although he solved cryptograms sent him by correspondents during the next two years. He eventually stopped this, complaining in a letter to a friend that “I have lost, in time, which to me is money, more than a thousand dollars, in solving ciphers.”

  “The Gold-Bug” opens with the hero, William Legrand, living a secluded life on Sullivan’s Island with an old Negro servant, Jupiter. A sort of amateur naturalist, Legrand has found a new specimen, a gold-colored bug; he sketches it for his friend, the first-person narrator of the story, upon a scrap of parchment that he has picked up on the beach. The narrator accidentally holds the parchment near a fire. When he looks at it, he sees only a reddish death’s-head. Legrand becomes abstracted upon seeing this. During the next month or so his behavior becomes increasingly strange. Jupiter fetches the narrator, and upon Legrand’s request they set off into the woods, carrying shovels. Halting at a large tree, Legrand makes Jupiter climb it, find a skull at the end of a branch, and drop the gold-bug (which Legrand has gotten back from the friend to whom he lent it) through one eye. Following a line determined by the bug and the tree, they dig—and exhume a fabulous treasure of glittering gold coins and jewels, buried there by Kidd. The mystery of how Legrand knew how to find it is resolved when he explains how, using heat, he developed a cryptogram in invisible ink on the parchment, and then (using the frequency count that Poe had copied out of the Blair article) how he solved the cryptogram.

  The story is full of absurdities and errors. The parchment was found near “the remnants of the hull of what appeared to have been a ship’s long boat,” which “seemed to have been there for a very great while; for the resemblance to boat timbers could scarcely be traced.” This was the boat in which Kidd had brought ashore his treasure. Would the parchment have remained in the same place for generations? If it did, would it not have suffered from the elements, as the timbers did? The invisible ink, Poe specifies, is “regulus of cobalt, dissolved in spirit of nitre.” Unfortunately, this gives cobalt nitrate, which is readily soluble in water. Would any trace of ink have remained on the parchment after decades on the beach? Even if any had, it would have been obliterated when Legrand washed the parchment in warm water to remove dirt. Legrand spotted the skull from a hillside seat through a rift in the trees; as soon as he moved from the seat, the skull disappeared. Poe implies that for that very reason the pirates chose that seat, that tree, and that branch. Would that narrow rift have remained unchanged through 150 years of arboreal growth? The first traces of the ink were accidentally brought out when the narrator’s hand, holding the parchment, dropped close to a fire; but heat strong enough to bring out the ink would probably have scorched the narrator’s hand. The frequency table was of course absurd, though Poe paid no attention whatever to it after using it to identify e. Finally, one may wonder whether Kidd would have employed so ridiculously melodramatic a manner of recording the treasure location, and have been so careless as to lose the note.

  All of these criticisms are valid. They show that Poe was less concerned with accuracy than with the appearance of accuracy, that he affected a learning which he did not have and to this extent was not intellectually honest. Beyond that, and to the reader, none of them matters. For no one thinks of these problems when caught up in the powerful narrative current of the story. The tale perhaps owes some of its force to Poe’s using it to vent some of his frustrated desires. “I cannot keep from thinking with sadness how the unfortunate E. Poe must have dreamed more than once about how to discover such treasures,” wrote Baudelaire. The plot, with its finding of buried treasure, may seem trite to readers of today. But that is for the same reason that Hamlet seems loaded with clichés: its virtues have made its lines so well known. Something of how it struck contemporaries, however, may be felt in what Baudelaire wrote: “How beautiful is the description of the treasure, and how good a feeling of warmth and dazzlement one gets from it! For they find the treasure! it is not a dream, as generally happens in all these novels, where the author awakens us brutally after having excited our minds with appetizing hopes; this time, it is a real treasure, and the decipherer has really won it.” The construction of the story, which ends, not with the finding of the treasure, as lesser writers might have done, but with the resolution of the cryptogram, is exactly right. Moreover, the exposition of that crucial part of the story, on which all turns, is a masterpiece of lucidity. “As we follow the steps of the argument,” wrote critic Wimsatt, “we have the impression of intricacy and precision, of Legrand’s shrewdness and patience—each detail receives attention—and yet we are never lost, the main outlines remain clear, the reasoning turns where it should, the momentum, or rhythm of the whole is sustained. The writing of this kind of prose was, as 1 see it, one of Poe’s most impressive gifts.”

  The story’s excellence wells up from deeper springs, however. The reader watches fascinatedly as a chain of logic appears link by link and ends in the disclosure of the answer to the central problem. “The Gold-Bug” and one or two of Poe’s other stories are the first to employ this intellectual operation as their theme, and so they are called the first detective stories. But “The Gold-Bug” has something that other detective stories do not. All mystery stories give the reader, who identifies with the hero, the mental satisfaction of resolving a puzzle. Beyond that, the reader gets nothing out of it, for in most cases the conclusion reaches out to a third party—it merely punishes the murderer. “The Gold-Bug,” however, rewards the hero-reader. Unlike other detective fiction, “The Gold-Bug” purges, in an Aristotelian sense, man’s craving for wealth and power. The story satisfies the emotions. It adds that extra dimension to the intellectual one. In that may lie the story’s special merit, its perennially engaging quality, while the breadth of its appeal may lie in the universality of the desire it brings to catharsis.

  At the same time, the cryptological element in “The Gold-Bug” exists on the same two levels: the levels of reason and desire that summoned up so intense a response in Poe. On the surface, the story deals only with the scientific aspect of cryptology: the text handles it strictly as a subject for “ratiocination,” for logical investigation. But the structure of the story utilizes cryptology as a form of divination. The mysterious symbols of the message contain the secrets of great wealth, and the man who reads them compels the earth to yield up these hidden treasures. These are precisely the operations of divinatio
n and magic, which seek to fulfill human desires by foretelling and so controlling nature, and these are the operations that the plot executes and so confirms. On this deeper, almost subliminal level, then, Poe set up anew a sympathetic vibration between cryptology and magic. Poe, in other words, glamorized cryptology.

  The effect was to popularize it. He was the first to do so. The immense success of his story gave far wider currency to valid information about cryptology than any mere textbook could have given. “Popular interest in this country in the subject of cryptography received its first stimulus from Edgar Allan Poe,” wrote Friedman. Interest extended to England, where the mathematician Charles Lutwidge Dodgson, better known as the writer Lewis Carroll, copied Poe’s table of letter frequencies into a notebook, along with a few other cryptologic odds and ends. “The Gold-Bug” was the first widely available and easily palatable lesson on the subject, and it continues as such. People still read it, learn from it—and are inspired by it. The contributions of those thereby attracted to cryptology probably cannot be measured, but whether they be few or many, cryptology owes them to Poe.

  It also owes him much of its literature. Since Poe showed the way, other writers have used secret writing in their stories. And not a few of their tales would be more honestly entitled “Return of the Gold-Bug” or “Son of Gold-Bug,” since they utilize the same hidden-treasure motif. (It is in reading many of these stories and then rereading “The Gold-Bug” that Poe’s genius becomes apparent.) Their cryptologic episodes are often well done, however, and the quality of the writer often shows itself in the way he handles the cryptology. The ciphers may be more plausible, the exposition of their solution clearer, their connection with the plot—where they are not merely a flourish—more intimate. Nearly always the ciphers are simple ones, for to explain or solve a complicated one slows the narrative pace too drastically and involves more explanation than the reader of fiction will put up with. Sometimes, of course, the author himself does not know enough to write effectively about the subject.