The Friar and the Cipher

by Lawrence Goldstone

  While Tiltman was working on the cipher, another veteran NSA analyst, Prescott Currier, noticed that, particularly in the herbal sections, some of the pages were written tightly, the letters small and cramped, while in most of the text the script was looser and more flowing. Studying the characters separately, Captain Currier theorized that not only were there two styles in the manuscript but two separate languages, which he named “Voynich A” and “Voynich B.” He couldn't penetrate either language, but, if he was correct, Currier had eliminated a potential source of error. While not everyone has agreed with Currier's assessment, the dual-language hypothesis has been incorporated into most of the subsequent research.

  In the early 1970s, Tiltman, now nearing eighty years old, heard that another young analyst at NSA named Mary D'Imperio had taken an interest in the manuscript. He turned over coordination of the Voynich transcription effort to her. (Whether the analysts were engaged in their Voynich activities in their spare time or as part of an official effort to see if the Voynich cipher had any utility has never been made clear.) In 1972, D'Imperio chaired a seminar, and then, four years later, published a monograph—actually under NSA auspices—entitled “The Voynich Manuscript: An Elegant Enigma.” Although there is little in this work that is new, it is extraordinarily detailed, complete, and well researched, covering even the most peripheral aspects of the manuscript's scholarship. If nothing else, it provides remarkable insight into the methodology of contemporary cryptanalysis and supplies some pretty revealing hints as to how they go about things at the NSA. A quarter century later, this work is still considered definitive, but D'Imperio personally (if she was in fact a real person and not an NSA phantom) has faded back into the NSA mist, never to be publicly seen or heard from again.

  THROUGH THE LATE 1970S AND 1980S, with nothing more heard from NSA analysts (at least for attribution), the spotlight was left to maverick amateurs. One of the most inventive entries came in 1978 when a self-styled philologist named John Stojko produced a solution in which he claimed that the manuscript was a copy of correspondence written in Ukrainian. In a book he called Letters to the God's Eye, Stojko claimed that before encryption all the vowels had been removed from the plaintext. The consonants were then encrypted by use of a secret alphabet with Etruscan roots. As a result, Stojko—who is Ukrainian himself and claimed to be able to read Etruscan—came up with sentences such as “Mr. is that you who believe in your believe? Miss, your Sust will renew eye and will rule. Tell me Eye of God; will God's eye of Sust heal?”

  Stojko's solution quickly went the way of Feely's. Not only did the text of the letters bear no relation to illustrations on the pages on which they appeared, but in his transcription he had inserted vowels and word spaces arbitrarily. In addition, Stojko, who also claims to be a historian, was said by others interested in the manuscript to often deviate from “the generally accepted history of the Ukraine.”

  There was little limit to the lengths that amateur cryptologists would go to be able to claim that they had cracked the code. In 1987, Leo Levitov, a New Jersey physician, produced a detailed solution in which he claimed that the Voynich manuscript was actually the last surviving liturgical document of the great Cathar heresy of thirteenth-century France and that the Cathars themselves were an offshoot of a cult that worshipped the Egyptian goddess Isis. Dr. Levitov noted,

  There is not a single so-called botanical illustration that does not contain some Cathari symbol or Isis symbol. The astrological drawings are likewise easy to deal with. The innumerable stars are representative of the stars in Isis' mantle. The reason it has been so difficult to decipher the Voynich Manuscript is that it is not encrypted at all, but merely written in a special script, and is an adaptation of a polyglot oral tongue into a literary language which would be understandable to people who did not understand Latin and to whom this language could be read. Specifically, a highly polyglot form of medieval Flemish with a large number of Old French and Old High German loan words.

  Further, according to Dr. Levitov, the streams of naked women depicted a Cathar sacrament called endura, the cutting of a vein in order to bleed to death in a warm bath.

  Like Newbold, Brumbaugh, and even Stojko, Dr. Levitov, after years of work, had created a tightly knit logical structure and transcription alphabet to support his case. Although his solution was at first met with some tentatively favorable comments, his methods and conclusions soon came under more harsh scrutiny. His transcribed text did not match details in surviving Cathari records (the existence of which he was seemingly unaware), and the endura was almost always a ritual fast at the end of life and not a form of suicide. Dr. Levitov's linguistics was also called into question.

  In 1991, in order to promote more serious research, Jim Gillogly, a computer scientist at the Rand Corporation, the California think tank, and Jim Reeds at AT&T Laboratories used the fledgling Internet to form a loosely knit colloquium of other Voynich devotees around the world. Some, like Reeds and Gillogly, had worked with encryption techniques professionally, whereas others were merely dedicated amateurs. The group soon grew into a cyber-study network with members from the United States, Europe, and Asia. “It's very orderly,” observed Dr. Reeds. “Everyone is listened to politely, even the crackpots.”

  Today, while there are dozens, perhaps hundreds, of Web sites that present one theory or another as to the manuscript's origins, the Reeds-Gillogly network (although both of them would deny that the network is “theirs”) continues to dominate the field. With its formation, study of the actual characters has drifted beyond the layman's grasp into the rarefied world of higher mathematics, statistics, and computer science. This is sad in a way, since, historically, it has been the Levitovs, Stojkos, Feelys, and Strongs who have provided most of the fun.

  And technology has done no better. A number of alphabets and transcriptions have been proposed, but despite subjecting the manuscript to the most sophisticated tests—Zipf's Law, for example—current researchers are not really any closer to a solution.*8 Someone even completed a master's thesis in library science with a statistical analysis that demonstrated that the “gallows characters” in the manuscript were not nulls. Jim Reeds noted, “No real progress has been made in reading the Voynich manuscript, except in the purely negative sense of shooting down screwball theories.”

  Any progress that has been made is in tracing provenance—Baresch and Beckx have recently been added to the list of owners—and fleshing out the sometimes lurid details of Wilfrid Voynich's life and that of his family. But none of this has aided decryption at all. Most of the experts and professional cryptologists who have studied the manuscript are convinced that it is not a hoax, but even that is only supposition. The absence of any attempt to scientifically date the paper or ink—Yale refuses to do either—renders the general conclusion that the manuscript dates from the fifteenth or sixteenth century rather than the thirteenth also merely a guess.†9

  CRYPTOLOGY HAS COME A LONG WAY since Bacon wrote his “Seven Ways of Concealing Secrets” in 1248, transformed, as most other forms of communication, by the vast capacity of computers. The art—or science—is now almost entirely computer based, and encrypted messages are a regular part of the life of anyone who uses the Internet. Interestingly, technology has not made cryptology more complex but, at least in its basic approach, more simple. Cryptographics is no longer keyed by mysterious symbols, page templates with cutouts, or Kaballist numerology—these are all considered just grist for the supercomputer's mill—but instead in complex algorithms or numbers so large that only computers can manipulate them.

  Still, in some ways, there has been no real progress at all. From Roger Bacon's time up through the Friedmans', any code or cipher fashioned by human ingenuity was susceptible to decipherment by the same means. No matter how brilliant the mind that fashioned a code, an equally brilliant mind might break it. In the technological age, we have simply transferred that same principle to machines—no matter how powerful a computer is thought to be
or how impenetrable an encryption, another more powerful computer might always come along and decrypt it.

  The search for the unbreakable code has led cryptology to the horizons of particle physics and a system called quantum cryptography. In quantum cryptography, a key to a computer encryption, changed at rate of up to 1,000 times per second, is transmitted over a fiber-optic link in a stream of photons, indivisible particles of light that have no mass but carry electromagnetism, one of the fundamental forces of nature. Photons, of course, travel at the speed of light. In quantum cryptography, these photons are polarized (the same technique is used in sunglasses, laptop screens, and liquid crystal displays), then sent in two streams. Because the electromagnetic energy in the system can be precisely measured, anyone trying to eavesdrop or break into the stream will cause an immediately detectable fluctuation that will disrupt the stream and render the message meaningless. When the intended receiver cannot decode the message, he or she will realize that someone has interfered.

  Quantum cryptography is, in theory, unbreakable because the very act of trying to break the code destroys it. If history has taught us anything, however, it is that there is a technology just over the horizon that will trump the technology on the horizon. And, of course, even with supercomputers, giant streams of numbers, and polarized photon cryptology, the centuries-old Voynich manuscript remains unread.

  With all the failures and dead ends, it becomes tempting to wonder if Newbold might have been correct back in 1921. More than eighty years later, no one has really done any better. And there are indeed those who insist that Newbold was unfairly attacked by his detractors and his version is largely correct, which would also mean that Roger Bacon is the manuscript's author after all.

  In a 1997 article in the New Agey Journal of Borderland Research, a computer cryptographer named Aldrich Daimler argued that when people like the Friedmans turned the art of cryptography into a mathematically based science, they may have made one type of cryptanalysis more efficient, but they also effectively closed off an important area of inquiry.

  According to Daimler, a cryptology consisting almost entirely of statistical analysis, complex algorithms, and the generation of immense streams of numbers completely ignores what in the Middle Ages was actually a creative art. Medieval cryptographers such as Roger Bacon intentionally employed steganography in a complex manner, rife with ideographs, anagrams, obscure and varying keys, puns, biliteral ciphers, and anything else they could think of, often in the same text. It is perfectly reasonable, therefore, Daimler argues, that Roger Bacon would employ all seven methods for hiding secrets in one document. What is certainly true is that Roger Bacon had a sense of humor. It was in his nature to have fun with his ciphers. He wrote them as much to playfully challenge his friends as he did to hide information. The tortuousness of Newbold's solution, then, rather than detracting from its veracity, actually reinforces it, and the Voynich manuscript becomes a prime example of a thirteenth-century cryptographer's art.

  In this, Daimler may well be correct. Even Jim Reeds admits that the Friedman study groups, since they were so mathematically oriented, never considered the illustrations as being integral to the cipher instead of merely objects of it.

  Daimler is most certainly writing from the fringe. Borderland Sciences was founded “for the purpose of investigating into realms normally beyond the range of basic human perception and physical measurement.” According to its Web site, Borderland currently specializes in alternative medicine, radionics, psychotronics, water technology, dowsing and radiesthesia, Tesla technology, and new energy. Daimler himself, according to Borderlands, seems to have disappeared. Regardless of the source, however, the article contains some persuasive specifics.

  He noted, for example, that while Manly claimed that the “pen strokes” in the o in multos in the key were due to the ink cracking, the ink in many other manuscripts of the same period had held up just fine. In another section, he takes issue with the most serious charge in Manly's article, that anagramming is incurably subjective and that Newbold strayed all over the text to force-fit the solutions to anagrams into a predetermined transcription. But even if Daimler is correct and Newbold never went further than a few characters in any direction, there is no saying that Newbold came up with the correct meaning.

  Other than the spectral Daimler, there are few today who are willing to say that Roger Bacon had anything to do with the Voynich manuscript at all. Bacon may well not have been involved, but, with all the conjecture and inconsistency in every other theory, Newbold's hypothesis makes the most sense. If the key does decrypt into Bacon's name—either by Newbold's reasoning or Brumbaugh's—and the probability that Edward Kelley would go to all that needless effort to fool Rudolph is small, why not accept Newbold's decryption, at least in part? None of the artificial language arguments has gotten anyone anywhere. Newbold's solution at least matched the illustrations, and he may not have subconsciously known of the Andromeda Nebula or riots at Oxford.

  But whether or not Roger Bacon had any impact on the Voynich manuscript, the Voynich manuscript had a very definite impact on Roger Bacon.

  CHAPTER NINETEEN

  The Unfinished Legacy

  of Roger Bacon

  • • •

  INSTEAD OF BEING THE MAKING OF BACON'S REPUTATION, the cipher manuscript turned out to be the ruin of it.

  Before Voynich and Newbold, Roger Bacon's scientific work had been undergoing a slow but solid renaissance. In 1861, Emile Charles had published Roger Bacon, sa Vie, ses Ouvrages, et ses Doctrines. This detailed, highly researched four-hundred-page biography was the most ambitious work on Bacon ever undertaken. Without making any outrageous claims as to Bacon's achievements, Charles lauded him as a martyr to the advancement of knowledge. Two years before, J. S. Brewer, a professor at King's College, London, had edited the Opus Minus, Opus Tertium, Compendium Studii Philosophii, and the 1248 letter and included a long introduction that came to much the same conclusion as Charles's work.

  The Brewer edition and the Charles biography sparked a number of articles and studies that denoted Bacon as a serious student of science. Then, in 1896, Andrew Dickson White published A History of the Warfare Between Science and Theology in Christendom. White, trained as a historian, had been a New York state senator, diplomat, and cofounder and first president of Cornell University. He was also a noted bibliophile, and, in 1891, he donated the 30,000 books in his collection to the university. His medievaliana still represents the cornerstone of Cornell's collection.

  Andrew Dickson White CORNELL UNIVERSITY

  Although a devout Christian (he was Episcopalian), White detested dogma, and Cornell had been established as a nonsectarian university, to the fury of many conservative religious leaders. Accusations of godlessness were not about to make Andrew Dickson White back down, however. In his two-volume work (every bit as well researched as Thorndike's), White traced scientific discipline after scientific discipline—chemistry, geology, astronomy, botany, physics—showing the progression from science based solely in scripture through the period when empiricism began to tip the scales to when scientific inquiry could finally proceed unencumbered by dogma. A History of the Warfare Between Science and Theology in Christendom caused an enormous sensation when it was published and is today, more than a century later, still in print.

  White saved his harshest condemnations for ecclesiastic zealots, such as Thomas and even Albert, who had “been drawn or driven from the paths of science into the dark, tortuous paths of theology,” and his highest praise for men who not only were great scientists, but also showed moral courage in maintaining their belief in the truth against the repression of theocrats. For this latter category, his epitome was Roger Bacon. White wrote, “His method of investigation was even greater than its results. In an age when theological subtlizing was alone enough to give the title of scholar, he insisted on real reasoning and the aid of natural science by mathematics; in an age when experimenting was sure to cost a man his re
putation, and was likely to cost him his life, he insisted on experimenting and braved all the risks. Few greater men have lived.”

  Interest in Roger Bacon increased. The next year, an English translation of the complete Opus Majus was published, followed by many of Bacon's Oxford lectures and earlier works. In 1914, to mark the seven hundredth anniversary of his birth, a group of international scholars formed a committee (Colonel Hime of the gunpowder article was vice chairman) to produce a series of new authoritative essays detailing all aspects of Bacon's achievements.

  So, when Thorndike first published his articles, he was attacking a far less sensational, more solidly based Roger Bacon than that which emerged after Newbold, and so his smugness, sarcasm, and obvious bias did not win many over.

  After Manly's 1931 Speculum article, that all changed. Scholars, embarrassed at having been duped, fled from the fanciful theories of the discredited Newbold and rushed to the comfortable academic meticulousness of Thorndike, tending to overlook that in his subjectivity and fussiness Thorndike was just as unreliable. Detail without context became the touchstone of Bacon scholarship, and the pendulum swung, as pendula will, past dead center and on toward the other extreme. The lingering need to debunk Bacon, to demonstrate that he wasn't the great scientific figure that he had been made out to be, has dominated the research ever since.