THE CODEBREAKERS

by DAVID KAHN

  It was thus perfectly possible for Francis Bacon to have used steganography to simultaneously conceal and reveal his authorship of the Shakespeare works. The question is, Did he?

  The first to assert that he did was one of America’s most colorful political figures, a round-faced man of great wit, ability as a public speaker, and personal popularity. Ignatius Donnelly became lieutenant governor of Minnesota at 28, and, four years later, in 1863, began the first of three terms in the House of Representatives. A political quarrel there blocked a fourth nomination. He quit the Republican party and, his radical reforming proclivities coming to the fore, became a Granger and a Greenbacker and won repeated election to the State Senate. But in 1878, following his rejection by the voters in a contest for Congress, he espoused two theories of pseudo science—the existence of Atlantis and the devastating prehistoric collision of the earth with a comet. At the same time, having chanced across a description of a steganographic system by Francis Bacon in a book belonging to one of his children, Every Boy’s Book, and apparently having heard of the new theory that attributed Shakespeare’s plays to Bacon, he determined that in the winter of 1878-79 “I will reread the Shakespeare plays, not, as heretofore, for the delight which they would give me, but with my eyes directed to discover whether there is or is not in them any indication of a cipher…. The things to be on the lookout for in my reading were the words Francis, Bacon, Nicholas [father of Francis], and such combinations of Shake and Speare or Shakes and peer as would make the word Shakespeare.”

  This search served as a recreation during Donnelly’s more serious work of supporting his family by his pen. In 1882 there appeared Atlantis: The Antedeluvian World. With enormous and wide-ranging but undisciplined erudition, Donnelly gave Plato’s legend of the lost continent coherent form for the first time. He regarded it as the actual Garden of Eden. As evidence for its existence, he marshalled similarities in flood myths, pyramid-building, knowledge of embalming, and a 365-day calendar between the civilizations of ancient Egypt and of the Aztecs and Mayans in the New World. The book was an immense success. Within eight years, it had gone through 23 editions in the United States and 27 in England (and in the 1960s a paperback publisher issued it anew). It brought Donnelly his first secure and comfortable income, and made him probably the most discussed literary figure outside of professional and intellectual circles. The following year he offered Ragnarök: The Age of Fire and Gravel. This volume, a predecessor of Immanuel Velikov-sky’s controversial Worlds in Collision of 1950, argued that earth had collided with a mighty comet in its infancy. The stories of Sodom and Gomorrah and Joshua’s commanding the sun to stand to still, together with race-memories of similar events in other cultures, testified to this catastrophe, Donnelly said. Ragnarök (whose title was taken from an Old Norse term for “doom of the gods”) also sold well.

  Even while writing it, however, Donnelly was quoting in his diary for September 23, 1882: “I have been working … at what I think is a great discovery I have made, to wit: a cypher in Shakespeare’s Plays … asserting Francis Bacon’s authorship of the plays … I am certain there is a cypher there, and I think I have the key; all this cannot be accident.” A year later the solution of this cipher had become his consuming passion: “I think about it all day and dream about it all night; it is hideously complicated and perplexing.” By May, 1884, its calculations had wearied him, and he turned with relief to the composition of the book that was to be his magnum opus, The Great Cryptogram.

  In September of that year, when he was running again for Congress, an acquaintance broke the news that Donnelly had found a cipher in Shakespeare. It stirred up some interest, but it did not win the election for him. Nevertheless, he continued his political activities side by side with his crypto-logic. In 1887, the year that he won election to the State Legislature on the Farmers’ Alliance ticket, Joseph Pulitzer’s New York World sent a professor of mathematics to Donnelly’s home in Nininger City, Minnesota, to examine the cipher system, which involved a great deal of arithmetic. On August 28 the newspaper splashed the professor’s favorable appraisal on its front page. During the winter, Donnelly worked from 10 a.m. to 11 p.m. to complete the book, called it “a terrible task,” and finished the last page with “an infinite sense of relief.”

  What had he discovered? He found, as he had expected he would, a narrative revealing that Skaks’t spur never writ a word of them and It is even thought here that your cousin of St. Albans [Bacon was Viscount of St. Albans] writes them. “Them” referred, of course, to the plays. Some of the decipherment was in the third person; why, Donnelly never explained. How had he discovered all this?

  He had begun by misapprehending Bacon’s cipher. Based upon this, he had sought an interrelationship of numbers that would locate the words of the hidden message in the open message of the plays by their serial position on the page or in an act. In its simplest form, the system would, for example, find that the 17th, 18th, 19th, and 20th words on pages 17, 18, 19, and 20 spelled out “I, Bacon, wrote this.” After beginning with a fruitless search in a modern edition of the plays—as if Bacon, in addition to his other talents, foresaw the exact pagination to be used 200 years after his death—Donnelly woke up and turned to a facsimile of the famous First Folio of Shakespeare’s plays, published in 1623, three years before Bacon died.

  At last he began to get results. They were not simple. Nor were they as straightforward as the fictitious example above. By some reasoning that he never set forth, Donnelly settled upon 505, 506, 513, 516, and 523 as “root-numbers.” From these he subtracted “modifiers,” or sometimes “multipliers”; from the differences he deducted the number of italic words on a page, sometimes counting stage directions, sometimes not; these results were altered by the addition or subtraction of the number of hyphenated and bracketed words—although he confesses that “we sometimes counted in the bracketed and additional hyphenated words … and sometimes we did not.” The final figure showed the position of the plaintext word on the page—then the page itself varied, and sometimes the first column was selected, sometimes the second, and occasionally counting began at the beginning of a scene instead of the page, and occasionally at the bottom of the page. Donnelly neglected to say why he chose one alternative over another, though he did set out his computations in impressive detail:

  The “b & h” means that he is counting bracketed and hyphenated words. Donnelly does not explain the mid-course changes after the periods in the computations for “spur” and “never,” nor the sudden backwards counting of “(121).” The First Folio pages the Comedies, Histories, and Tragedies separately, and Donnelly’s figures refer to the pages of the Histories in the Staunton reproduction.

  Donnelly’s book, The Great Cryptogram, consisted of “deciphered” passages accompanied by their derivations and interspersed with polemics defending the method and setting the hidden story in perspective. The publisher, R. S. Peale and Company of Chicago, brought in a special printer to set up crucial portions of the book without seeing the rest. But Peale had a premonition of what was in store when he encountered an unexpected resistance to the anti-Shakespeare idea as he sought prepublication subscriptions in England. He set the first edition at 12,000 copies. But despite the reputation of the author, the book flopped. Hostile reviews battered it. Readers found the demonstration of the cipher confusing, and the book as a whole lacked flow. Worst, the cipher itself suffered crippling blows.

  At one point, after deciphering the names Cecil (as seas-ill), Marlowe (as More-low), and Shakespeare (as Shak’st-spur), Donnelly had asked: “Are there four other columns, on three other consecutive pages, in the world, where six such significant words can be discovered? And, if there are, is it possible to combine them as in the foregoing instances, not only by the same root-number, but by the same modification of the same root-number? If you can indeed do this in a text where no cipher has been placed, then the age of miracles is not yet past.”

  Another Minnesotan, Joseph Gilpin
Pyle, promptly demonstrated that it was not. He parodied both the title and the method of The Great Cryptogram in his own The Little Cryptogram, in which he extracted by a similar method the following message from Hamlet: Don mill he [Donnelly], the author, politician, and mountebanke, will worke out the secret of this play. The Sage [of Nininger, Donnelly’s sobriquet] is a daysie. In some respects his calculations were much simpler than his subject’s: Don was simply 523 minus 273, making the 250th word on page 273, column 2. Pyle was seconded in a devastating work by the Reverend A. Nicholson, who in a cruel coincidence happened to be the incumbent clergyman at Bacon’s own home of St. Albans. In a brilliant refutation, Nicholson used Donnelly’s own root of 516 on the very pages in which Donnelly first glimpsed his solution to produce a “decipherment” diametrically opposed to Donnelly’s: Master Will I am Shak’st spurre writ the play and was engaged at the Curtain. One computation ran: 516-167 = 349 − 22b & h = 327 − 163 = 164 − 50 = 114 − 1h = 113, pointing to the 113th word on page 76, column 2, which was “Will.” Nicholson clinched his demonstration by producing the same text four times from Donnelly’s four other root-numbers. It became evident to all but Donnelly that Pyle and Nicholson had merely done consciously what Donnelly had done subconsciously: selected the words of the hidden message, then found the figures and arithmetic that supported them.

  Some of Ignatius Donnelly’s calculations

  Donnelly, unbowed, went to Europe to lecture on his decipherment. At the University Union at Oxford, where he debated a Shakespearean, a poll of the audience routed him, 167 to 27. Criticism grew increasingly harsh, and when he returned after five months, Peale told him that the book was dead. Donnelly refused to believe it, kept the controversy alive in letters to editors, and had the Pinkerton Agency check on Peale’s records. Eventually the publisher sued him for $4,000 in advance royalties that had not been earned. To settle, Donnelly traded the publisher lots in St. Paul for the plates for the book. “I will … put them in my garden, and build a little house to cover them,” he wrote pathetically in his diary on December 22, 1892. “The little building will be my monument of colossal failure. Every time I look at it, I shall think of wrecked hopes and ruined ambitions.”

  He was unusually depressed because he had just lost an election for governor on the Populist ticket, for which he had, that summer, written a ringing platform that became the creed of this third party and foreshadowed many modern reforms. Nevertheless, he soon rebounded and continued his fight for the underdog. He had, interestingly, depicted Bacon in his decipherment somewhat as he visualized himself—a courageous, honest, struggling politician victimized by greedy, corrupt officeseekers. Donnelly never lost his own faith in his Baconian revelations, and he continued to labor at the cipher. In 1899 he privately published The Cipher in the Plays and on the Tombstone. It passed directly into limbo. His political fortunes followed it the next year, when the Populists, who had nominated him for Vice President, smothered in the McKinley victory. On January 1, 1901, the first day of the twentieth century, Donnelly died.

  Of Donnelly’s “system” it may be remarked that nothing like it has appeared in cryptology before or since. And with good reason, for the system is no system at all; there is neither rhyme nor reason to the choice of numbers that lead to the result. It may also be remarked that, in an open-code system, the hidden message controls the cover-text, which is merely a function of the hidden plaintext. Donnelly, though he worked only on a few pages of the two parts of Henry IV, therefore presupposed that the magnificent language of the plays all resulted merely from the inner workings of a cipher. Did Falstaff, marvelous Falstaff, exist so exuberantly only to make sure that Bacon would have the right words for an open code? The thought is hard to bear.

  Donnelly’s murder of logic, like the slaying of Banquo, started a line of phantoms that threatens to stretch out to the crack of doom. Among the Baconians, these apparitions are “ciphers” that are not really ciphers. Likewise, the technique of descrying them is not really cryptanalysis, and the results are not solutions or decipherments. They are the deliriums, the hallucinations of a sick cryptology. The suggestion to call this whole area “enigmatology” comes from a Baconian; and it is a good one, for it will prevent using the terms of cryptology for noncryptology, will prevent calling a “cipher” that which is not a cipher. On this basis, then, a Baconian “cipher system” would be an “enigmaplan,” the verb for its obscuration would be “enigmalyze,” and the result would be an “enigmaduction”—a term every bit as graceful, as purebred, and as well constructed as that which it denotes.

  The more important of these enigmaplans have come under the cool scrutiny of the Friedmans in The Shakespearean Ciphers Examined (a nice anapestic trimeter of a publisher’s title which they dislike because it implies that ciphers exist in Shakespeare). In its uncondensed form, this book won the Folger Shakespeare Library Literature Prize of $1,000 in 1955. The Friedmans pointed out that, unlike, say, a professor of English, they had “no professional or emotional stake in any particular claim to the authorship of Shakespeare’s plays” and that the anti-Shakespearean “claims based on cryptography can be scientifically examined, and proved or disproved.” They say that they will accept as valid any cipher that fulfills two conditions: that its plaintext make sense, and that this plaintext be unique and unambiguous—that, in other words, it not be one of several possible results. So saying, they set out to see if anyone had discovered valid cryptologic proof that Non-Shakespeare wrote Shakespeare.

  They find no such proof. But they have a fascinating trip. They pass through a surrealist landscape where logic and the events of history both resemble and do not resemble the real things, like the oozing watch of Salvador Dali, where supermen of literature outperform the most harried of hacks in volume and the most thoughtful of philosophers in profundity—and then sit up nights enciphering secret messages to tell about it, where enigmatologists frantically nail together wild tottering structures upon the quicksands of conjecture. Though sometimes reviled by the natives, the Friedmans never lose their composure. As guides, they are wise, courteous, and quite entertaining.

  They introduce their readers to Orville Ward Owen, a Detroit physician. His basic tool was a “cipher wheel,” which consisted of 1,000 pages of Elizabethan writing glued onto 1,000 feet of canvas wound on two giant spools. With its help, he enigmalyzed from those pages an autobiography in which Francis Bacon revealed that he was the natural son of Queen Elizabeth and Robert Dudley, Earl of Leicester, and had written not only the works of Shakespeare but those of Christopher Marlowe, Edmund Spenser, Robert Burton {Anatomy of Melancholy only), George Peele, and Robert Greene. Indeed, he had done so primarily to conceal the story that Owen laid bare. Owen’s enigmaplan depended upon four key words, FORTUNE, HONOR, NATURE, and REPUTATION. The Friedmans summarized its rules as: “first find one of your key words (or one of its various derivatives); then look for a suitable text somewhere near the place where it occurs; and if you find one which fits into the story as you want it to be, there you are—another triumph of decipherment.” That Owen drew one enigmaduction from a translation prepared 22 years after Bacon’s death only confirmed the immortality of the Lord Chancellor’s genius. Owen learned that Bacon left the original manuscripts of the plays in a set of iron boxes on the grounds of Chepstow Castle in England. He went there and excavated, shifting his spot several times as the enigmaplan issued varying instructions. No manuscripts were found.

  Some Baconians claim to have discovered “cipher signatures” of their hero in the Shakespeare plays. Walter Conrad Arensberg, a wealthy Philadelphian, showed that for hundreds of years millions of readers had been blind to Bacon’s authorship when he found such a signature in Polonius’ famous advice to Laertes (Hamlet I. iii. 70-73):

  Costly thy habit as thy purse can buy;

  But not exprest in fancie; rich, not gawdie:

  For the Apparell oft proclaimes the man.

  And they in France of the best ranc
k and station,…

  “Consider in these lines,” wrote Arensberg, “the following acrostic letters:

  Co

  B

  F

  An

  Read: F. Bacon.”

  The Friedmans refuted this sort of nonsense by painstakingly counting the initial letters of 20,000 lines of the First Folio. They calculated that chance would assemble the letters b, a, c, o, and n in that order only 0.0244 times in the approximately 100,000 lines of the First Folio. Significantly, Arensberg did not find any such straight acrostic. Instead, he had to widen his field to include second letters, such variants as “Baco” and “F. Baco,” and anagrammed forms of these. This promptly brought the pure-chance probabilities well within the range of the First Folio—and it is these that Arensberg “discovered.” If a die turns up a deuce 1,000 times out of 6,000 throws, can this prove more than that what will probably happen has indeed happened?