by DAVID KAHN
This tells heavily against the Gallup results. If the biliteral cipher were present, all decipherers should be able to obtain the same message. That her results were at least partially subjective she conceded in her admission that “I sometimes think inspiration” is “absolutely essential.” Critics leveled objections at the use of words in her texts in senses they did not have in Bacon’s time, and at such barbarous abbreviations as adoptio’ and ciphe’ that no one but she ever used. Her “decipherments” in general have a vague pointlessness: “Seeke the keyes untill all bee found. Turne Time into an ever present, faith-full companion, friend, guide, light, and way. For he who seeks an entrance here, must be furnished in that manner aforesaid.” Was this the secret message, the primary text to which the plays, with all their soaring eloquence and profundity, were but secondary? Would Bacon have spent so much of his time enciphering this drivel, and expended so much of his cash to have a printer set it in the painstaking manner required? Would one of the most acute thinkers and most pithy of English writers (who on the Baconian theory also composed the poetry of Shakespeare) have considered such maunderings his greatest secret? No open-minded man will believe so.
Even the slight possibility that this is so is eliminated by the technical criticisms that utterly demolish the Gallup “decipherment.” Dr. Fred M. Miller, a document expert of the F.B.I. whom the Friedmans had interested in the problem, pointed out that Mrs. Gallup was notoriously inconsistent in her letter assignments. For example, some of her A-form t’s resembled B-form t’s more than they do other A-form t’s. The Friedmans, checking up, caught her frequently adding or dropping letters of the cover-text. And finally, they found that she did not consistently “decipher” the same pieces of type the same way. Printers pick up the type of the page headings from one form and use it for other signatures; the identical type should result in identical decipherments. With bibliographical help, the Friedmans identified the lifted headings and found that more often than not (in the first 21 pages of the First Folio) Mrs. Gallup’s “decipherments” of them varied.
This devastating analysis consigned her “solution” to the dustbin of enigmaduction, but it did not deny the possible existence of a biliteral cipher in the First Folio; Mrs. Gallup may simply have misread it. The Friedmans, however, assembled expert evidence to show that none does.
Experts in printing showed that the printers of the time changed authors’ spellings in order to justify lines more easily; this practice would break up the even flow of the biliteral. Poor inking made letters printed from the same piece of type look as if they had been printed from different ones. Paper was dampened before printing; it often dried unevenly and shrank identical letters to different sizes. Frequently, closed letters like a, e, o filled with ink, obscuring differences. And Dr. Charlton Hinman, who has collated dozens of First Folios letter by letter and traced hundreds of distinctive pieces of type through their pages for his massive typographical study, The Printing and Proofreading of the First Folio of Shakespeare, has found that any given copy contains “large numbers of variant readings” as a result of the customary printing practice of the time. Consequently, no biliteral message would have been transmitted with absolute fidelity. His studies “have certainly revealed nothing to encourage the idea—to put it in very moderate terms—that the book contains biliteral cipher.” All these points militate against any probability of its use in the First Folio.
Conclusive proof that it was in fact not used came from two experts. One was Frederick W. Goudy, one of America’s most distinguished typographers. Fabyan had commissioned him in 1920 to look into the possibility of a biliteral cipher in the First Folio—and then had suppressed the report. For Goudy, himself a designer of typefaces, had measured, sketched, analyzed, and compared the letters of the First Folio, and had concluded that a multiplicity of typefaces had been used, not just the two that the biliteral required. (This finding, incidentally, accorded with the then current practices of English printing, which had fallen to a low estate.) Independently, the F.B.I.’s Dr. Miller concurred: “No characteristics were found which support the classification into two fonts, such as A-font and B-font.”
It may put the Baconians into perspective to recognize that they are not the only enigmatologists. Gabriele Rossetti, the 19th-century Italian nationalist who was the father of the English poets Dante Gabriele Rossetti and Christina Rossetti, found in Dante’s Divine Comedy a secret language in which a secret society, opposing political and ecclesiastical tyranny from the earliest times, expressed its aims and informed its members of its affairs. Professor David S. Margoliouth, who had served in England’s M.I. 1(b) in World War I and should have known better, rearranged the colophons of the Iliad and the Odyssey and found two incantations to the muse which neither appear to make much sense nor add anything to the epics. Anti-Baconian Ib Melchior fell victim to his opponents’ obsession and enigmalyzed from Shakespeare’s tombstone a message in alleged Elizabethan English that supposedly meant, “Elsinore laid wedge first Hamlet edition”; his enigmaduction enjoys the dubious prestige of being the first to have been laid low by Claude Shannon’s unicity-point formula. Baconian Pierre Henrion has extended his enigmaplan to Jonathan Swift; by anagramming the nonsensical names in Gulliver’s Travels, substituting letters, and then anagramming again, he “proves” that LEMUEL GULLIVER really means Jonathan Swift and that LILLIPUT is Nowhere. A British artillery colonel, H. W. L. Hime, has “deciphered” a text proving that Roger Bacon invented gunpowder; unfortunately, while the most important letters of this text appear in printed versions through some errors, they do not exist in the original manuscript.
Other enigmaductions have been elicited from the Bible, Chaucer, Aristotle. Lesser minds have enigmalyzed less exalted texts. Hans Omenitsch, an anti-Semite who lived in Jackson Heights, New York, found that the Dick Tracy comic strip of April 18, 1936, concealed the plaintext Nero mob in fog rob Leroy apt rat in it are a goy. He had equal success with the Harold Teen comic strip and ordinary news stories in various New York dailies. He explained to a Congressman: “A criminal system of codes is operated daily in the press by the real masters of the country, who not only control the press and the politicians in power but who also control and direct the so-called red movement (international) as a sham to hide their real operations.” His enigmaplan was unclear, and as for his enigmaduction, the Congressman said that “it does not mean anything to anyone.”
Just as the Baconians are not the only enigmatologists, so the enigmatologists are not the only expositors of false historical and scientific theories. Science and history have their sicknesses, too. Various groups of oddballs maintain their belief in a flat earth, in a hollow earth, in dowsing rods, in the prophecies incorporated in the measurements of the Great Pyramid of Cheops, in the existence of Atlantis (it is no coincidence that Donnelly fell for both this and the Baconian theory), in flying saucers, in such medical cults as iridiagnosis, which diagnoses illnesses from the appearance of the iris. The enigmatologists form but the literary sector of this lunatic fringe. Both they and the other pseudo scientists seize upon the possible as if it were the probable, fantasize behind a mask of rationality, multiply entities beyond necessity, and refuse to test their hypotheses.
The techniques for eliciting universal truths from the Great Pyramid resemble the Baconian enigmaplans so strikingly because of these shared attitudes. For example, one Pyramid theory holds that its internal measurements embrace practically all of man’s historical and scientific knowledge. Various multiplications of measurements of height, side, the length of the Grand Gallery, and so on, produce the dates for past and future events in world history. The Creation (4,004 B.C.), the Flood, Christ’s birth, the Great Tribulation before his Second Coming, all are indicated. Other combinations yield scientific truths—the distance from the earth to the sun, the earth’s mean density, the mean temperature of its surface, and so on. With so many lengths possible in so complex a structure, and no strict rules to go by, it is
obviously possible to juggle them to produce figures which coincide with important dates or scientific facts. Such manipulations parallel the undisciplined combining of letters in the First Folio to produce Baconian authorship proofs.
Both the Baconian and the Pyramid assertions claim to be legitimate scientific theories, but their behavior in the face of criticism or embarrassing fact strips them of such pretensions. Their proponents do not reconsider, request new tests, submit to verification. Rather, they vilify their critics, dodge, equivocate, explain away. Never do they concede that they might be wrong. When Mrs. Gallup, in a tight corner, had to read a B-form as an A-form to make sense, she summoned up a phony explanation that she never had to use in any of her previous determinations: the wrong form was deliberately inserted to confuse because “ciphers are made to hide things, not discover them.” When the Friedmans knocked enigmaplans wholesale into a cocked hat, the Baconians suddenly came up with something that not one had ever mentioned: “while an Elizabethan cipher may be considered invalid by modern standards of rigid cryptography, it may well have provided its institutors with a fairly safe method of recording historical facts, or personal opinions, which they could not express without grave risk. An Elizabethan cryptogram may be suggestive rather than conclusive, and these suggestions may ultimately, by their very frequency, command assent.” Not an iota of evidence exists for this, outside of the Baconian claims themselves, and it is safe to say that for every Baconian “suggestion” drawn forth, someone could, if he wanted to, extract an equivalent Shakespearean one. But it would be as idle to do so as to talk back to a phonograph. For the Baconians do not seek knowledge—they have the faith; they are not scholars, but advocates.
The situation which the enigmatologists and the other pseudo scientists exemplify almost ideally is not at all uncommon, and philosophers have examined it often. A. J. Ayer of Oxford is especially clear: “A man can always sustain his convictions in the face of apparently hostile evidence if he is prepared to make the necessary ad hoc assumptions. But although any particular instance in which a cherished hypothesis appears to be refuted can always be explained away, there must still remain the possibility that the hypothesis will ultimately be abandoned. Otherwise it is not a genuine hypothesis. For a proposition whose validity we are resolved to maintain in the face of any experience is not a hypothesis at all, but a definition.”
The Baconians so maintain their view. They insist that their theory is true, but if it may be true it may also be false, and this they will not concede. For upon what evidence would they abandon their assertions? The finding of a holograph of Hamlet? If experience is any guide, they would say that Shakespeare copied it out at Bacon’s orders. The discovery of a note by Bacon that he hated Shakespeare’s plays and would have nothing to do with such claptrap? Obviously a clever trick to throw contemporaries off the trail. The Baconians cannot lose. But then they cannot claim to have won, either.
People ask, “Does it matter who wrote the Shakespeare plays? After all, it is the plays themselves that count, not who wrote them.” It matters because truth matters. The Baconian error has implications far beyond the Bacon-Shakespeare question. “If one can argue that the evidence in Shakespeare’s case does not mean what it says,” a scholar has written, “that it has been falsified to sustain a gigantic hoax that has remained undetected for centuries, then one can just as surely argue that other evidence is not to be trusted and that, as Henry Ford said, ‘history is bunk.’ ”
It is as pointless to try to convince the Baconians of this on rational grounds as it would be to demonstrate to an inmate of a mental hospital, with pictures of Napoleon’s funeral and tomb and attested documents of Napoleon’s death, that he is not Napoleon. For neither he nor the Baconians hold their views rationally. They hold them emotionally. The problem of enigmatology is, at heart, not logical but psychological. This is not to say that Baconians are psychotics—on the contrary, in non-Baconian spheres they function adequately, perhaps even outstandingly. But as Baconians they live in a fantasy world. Enigmaductions are classic instances of wishful thinking, of unconscious projection, of figments of the imagination. These results of an overactive cryptanalytic gland, these bloated growths of a chaotic imagination are like cancers on the corpus of normal codemaking and-breaking. They constitute the pathology of cryptology.
Paracryptology
25
ANCESTRAL VOICES
IN THE FALL OF THE YEAR, the staid Journal of Hellenic Studies issues its annual volume. Its articles examine in exhaustive detail some lesser points of Greek philology or literature—the individual bricks of scholarship that raise mankind’s house of knowledge. Their prose is flat and unemotional, their titles restrained.
The volume for 1953 contained an article that in all outward forms resembled the others. Its text was a thicket of Greek verb endings and grammatical forms, its title as carefully circumscribed as the others: “Evidence for Greek Dialect in the Mycenaean Archives.” Yet for many readers it evoked the ringing plains of windy Troy, the coruscating helmets and dancing horsehair plumes of Homeric heroes crowding beneath the walls and topless towers of Ilium, and ancient Crete, with its bull-dancer frescoes, its sinister labyrinth, its Minotaur, Theseus, and the House of the Double Ax.
For the article reported the decipherment of a lost writing, a script called Linear B scratched on clay tablets when Achilles and Agamemnon, Helen and Menelaus walked the earth. It was the most recent in the long series of decipherments that have given tongue to mute stones, brought to life Pharaoh and Nineveh and the panoply of ancient civilizations, allowed ancestral voices silenced for millennia to whisper across gulfs of time and space to the men of today. Some of these decipherments must rank among the noblest achievements of the human mind. For how to read the unknown writing of those long dead? How to speak the words of those whose voices murmur only in the sighing of the wind?
The solution of the problem shares some techniques with cryptanalysis. In one way the linguistic problem is easier, for there has been no deliberate attempt to conceal, but in another way it is harder, for sometimes an entire language must be reconstructed. In general, the linguistic problem involves two factors—the writing and the language. Either may be known or unknown. Four cases therefore arise.
In Case 0, both script and language are known. No problem exists: an Englishman can read English in its customary alphabet. In Case I, the language is known but not the writing. This is the simplest problem, equivalent to that of a substitution cipher. If the writing is alphabetic, the solution resembles that of a monalphabetic substitution; if syllabic, like kata kana, that of a nomenclator; if logographic, like Chinese, that of a code. In Case II, the script is known but not the language. An American who does not know Italian may be able to read aloud a newspaper article in an approximation of that language—but he will not understand what he is saying. The problem that faces the decipherer in Case II parallels the one that would face an American who wanted to teach himself Italian without any grammars or dictionaries, helped only by any pictures that might accompany the texts, by an English translation, or by a knowledge of related languages, such as French, Spanish, and Latin. Without any of these external elements, solution is probably not possible. In Case III, neither writing nor language is known. If this occurs in cultural isolation, so that neither can ever be known, no solution can ever be reached. But it has happened that although neither is known at the beginning of a study, outside information—usually a proper name known from another culture—determines the sound-values of the writing system and this, with the help of a translation or cognate languages, leads to the reconstruction of the language as well. The problem resolves into a succession of Cases I and II.
Case I solutions sometimes read like textbook explanations of elementary cryptanalysis.
In the summer of 1946, Édouard Dhorme, an eminent Orientalist who was a dean at the Sorbonne’s School of Advanced Studies, undertook the study of some inscriptions dug up in Syria at B
yblos, the town that gave its name to the Bible. The writing resembled hieroglyphics, but it made no sense read that way. From the location of the inscriptions and from their approximate age, Dhorme felt confident that the language underlying the approximately 100 sign-images was Phoenician. The number of signs suggested a syllabary, in which each sign would represent a syllable, like /ta/. But the usual Phoenician writing followed the Semitic trait of writing only the consonantal skeleton—“mister,” “master,” “muster,” “mystery,” and “mastery” would all be written mstr in the Semitic convention. Since no one knew how the words sounded, Dhorme could not hope to recover the vowels of his syllabary. But since the Phoenicians who wrote the inscriptions knew the vowels, and presumably used different signs for, say, ta, ti, and tu, Dhorme had to expect to find several signs for what he would know only as t.
These difficulties did not daunt the 65-year-old scholar, who had been made an officer of the Légion d’Honneur for cryptanalytical work in World War I and who had been one of three independent decipherers of another script in 1929. He boldly attacked the Byblos pseudo-hieroglyphics, beginning with the assumption that seven vertical marks in the lower left corner of a tablet represented a regnal date.