![]()
1The structures of seventeenth-century cryptography manuals
The history of cryptography made for exciting reading for both academic and popular audiences during the seventeenth century just as it does now. Writing in cipher even became a domestic pastime during the decades of and following the Wars of the Three Kingdoms. John Wallis himself remarked in 1653 that ciphering and deciphering were just as popular amongst the public as they were in the government and military, though audiences would not read this characterization until 1737: “now there is scarce a Person of Quality, but is more or lesse acquainted with it, and doth, as there is Occasion, make use of it” (10). Charles I used cipher not only for diplomatic and military correspondence but also for his private affairs. Samuel Pepys’s use of cipher in his diaries is famous, and diaries by other figures of the period also contain passages in cipher or shorthand. Personal letters also frequently include secret moments. Constantijn Huygens, Jr., Dutch scientist and secretary to William of Orange, wrote in cipher during a grand tour in 1649. In one eyebrow-raising example, Huygens confesses to a correspondent that he was unable to successfully have sex while on his tour and then continues his paragraph in cipher, apparently hiding the most scandalous (or embarrassing) highlights of his trip.1
The conventions established by seventeenth-century English cryptography manual authors have changed very little in nearly four hundred years. So enduring is the genre that even given significant technological change, the narrative structure of the seventeenth-century manual is discernible in the twenty-first. Almost all studies of and in the discipline of cryptography written since the seventeenth century, including those that became best sellers in the twentieth, share similar tropes, narrative conventions, and even examples with their seventeenth-century ancestors. The differences lie in the depth that one history may grant to a particular period, the brevity with which it treats another, and the detail that it offers when walking readers through the solution to a sample code or cipher. The author always offers a brief preface about his personal interest in or curiosity about cryptography. I can use “her” only in one instance, for Helen Fouché Gaines’s Cryptanalysis: A Study of Ciphers and Their Solution (1939), which is an exception in other ways as well that I explain.2 Authors then make defensive statements about the legitimacy of cryptography as a field of study, usually mentioning the historical fetters it never seems to shake off regardless of how many times the argument is countered: its association with the occult, which I discuss at length in Chapter 3. Each manual then mentions its current historical moment and the urgency of advancing knowledge in the discipline. Laurence D. Smith, for example, writing Cryptography: The Science of Writing in 1955, was most concerned about Russian, Asian and Middle Eastern threats.3 Next, authors lament the limitations in human communication given the sheer number of languages that are spoken around the world. Earlier manuals often begin with the curse of Adam, while twentieth-century manuals focus on those languages their authors perceive to be of most immediate danger. Regardless, the polyglot nature of the world is threatening, an obstacle to be overcome. All manuals also stress the importance of both secrecy and speed.
Following a brief mention of their contemporary situation, most authors of cryptography manuals even into the late twentieth century begin with a broad historical overview of the discipline, reaching back to Egyptian hieroglyphs and scripture. One of the most popular early examples is of Histiæus, who wanted to send a secret message to his son-in-law, Aristagoras, to leave Greece. Histiæus ordered that his slave’s head be shaved and the message tattooed onto his skull. Once the hair had grown back, the slave was sent off and required to request that his head be shaven. Wilkins is particularly excited about this story: “But amongst all the ancient practices in this kind, there is none for the strangenesse, to be compared unto that of Histiæus, mentioned by Herodotus” (Mercury 30). “By which relation you may see,” Wilkins concludes, “what strange shifts the antients were put unto, for want of skill, in this subject, that is here discoursed of” (31). Edward F. Hulme would joke about this same story in 1898 when he marveled at how much more slowly life must have proceeded in those golden ages before the hustle and bustle of the late nineteenth century.4 Smith, like Wilkins, seems to smirk in prose about the example, but he notes that although this method may “appear somewhat inefficient,” in fact “during the First World War it was the practice to send spies across the enemy lines with messages written on their bodies in invisible ink.”5
Codework usually begins just after the early historical foundation is established, and it often commences with the simple substitution method and its use for early Romans like Julius Caesar. Diagrammatic ciphers are handled next, with brief discussion of the fact that they were not secret enough because they were obviously trying to be secret. Several of the examples are identical across centuries and nations. Smith includes images directly from Mercury. He then, as is typical, moves to Bacon’s bilateral cipher. From there he merely notes that “the story of the development of military cryptography through the eighteenth, nineteenth, and twentieth centuries would fill volumes”; yet, like all other historians, he does not go on to do so.6 Other manuals discuss the sixteenth and seventeenth centuries in more depth than Smith’s but, overall, mention the same basic figures. Most discuss Wallis, note Wilkins briefly, and give a moment of attention to Falconer, but none discuss Morland. Smith offers one observation that is not commonly discussed, which is the frequency with which English citizens sent coded messages through magazines and journals, by placing dots over letters, in order to avoid paying the exorbitant postage costs.
What a more focused historical consideration of the textual tradition of cryptography reveals is that, despite the scholarly lack of attention given to the role of the seventeenth century in the discipline’s past, the cryptography manual emerged as a popular genre between the reigns of Charles I and Charles II, with a set of conventions that would survive for hundreds of years. One need look no further than a curious World War I textbook for kindergarten students, produced in the United States cryptanalyst headquarters at Riverbank Laboratories in Geneva, Illinois, to see that seventeenth-century cryptography manuals defined the structure of cryptography instruction for centuries. Riverbank Kindergarten Director Dorothy Crain issued a series of textbooks around 1916, including Ciphers For the Little Folks: A Method of Teaching The Greatest Work of Sir Francis Bacon (1916) by Helen Louise Ricketts.7 This textbook not only argues that childhood understanding of seventeenth-century bilateral cipher is integral in the development of reading and composition skills, but it also replicates the lessons (from simple to complex) that one sees in early modern manuals. It also includes a history of hieroglyphs and the alphabet that is much in line with the history of hieroglyphs and alphabetic language and writing that Wilkins includes in Mercury to argue that cryptography is the inevitable next step in human communication. The influence of this line of thinking—that cryptography was the foundational skill for child reading and writing development—is beyond the scope of this study, but future scholarship on the influences of cryptography on early childhood education is needed.
Tracing this history requires at least cursory awareness of the textual precedents for the genre, particularly of Johannes Trithemius’s (pseudonym for Johann Heidenberg) Polygraphiae libri sex, Ioannis Trithemii abbatis Peapolitani, quondam Spanheimensis, ad Maximilianum Caesarem (1518) as a foundational text and of successors who either built from or countered Trithemius’s work. These include Girolamo Cardano’s De Subtilitate rerum (1550), Giovanni Battista Porta’s De Furtivis Literarum Notis (1563), Blaise de Vigenère’s Traicté des Chiffres (1586), Augustus of Brunswick-Lüneburg’s (also August, Duke of Braunschweig-Lüneburg, alias Gustavus Selenus) Cryptomenytices et Cryptographiae libri IX (1624), and Antonio Maria Cospi’s La interpretazione delle cifre (1639).8 I will contextualize some of these writings in Chapter 2. Not all of them share the same conventions as the English versions during and after the Wars of the Three Kingdoms, but most are highly self-reflexive and intertextual, are written by expert practitioners, and often begin with tropes of accidental education in and mere curiosity about the field, the setting up of historical precedent, and then the technical working through of a number of examples, from simple to complex. And like Mercury, New Method, and Cryptomenysis, some appear to contain errors and misleading explanations, which I examine in Chapter 7.
Many of the manuals across the sixteenth and seventeenth centuries copy in the examples of previous publications, sometimes duplicating the errors and sometimes adding them. The manuals are conscious and manipulative of their own physical structures. Wilkins, Morland, Bridges, and Falconer are also aware of the textual history and contemporary trends in a range of genres, particularly the experimental essay, which they use to help define their pedagogy. Printing quirks, faint invisible writing, and paratextual playfulness and cross-referencing are also common. So, in short, these manuals are not simply technical textbooks teaching a rigorous discipline but creative experiments with language, the physical printed object, and the limits of reader-writer trust and reader observation. Where Mercury diverges is in its conclusion, when it doubts the usefulness of the written contributions it has just examined at length and advocates for a multimodal model of communication that can better drive the British economy.
At the other end of my historical spectrum is Davys’s 1737 An Essay on the Art of Decyphering, which indicates that while the structure of the manual, as genre, survives, there will in fact be a rhetorical shift in the way in which cryptography is presented (see Figure 1.1). The eighteenth century’s writings on cryptography will negate some of the work that the seventeenth-century manuals did to distance cryptography from the occult and promote it as accessible to, and necessary for, the everyday citizen. Ironically, Davys claims to publish his Art of Decyphering “to inform the Publick,” yet his edition will argue that only specialists can truly succeed as decipherers, and Wilkins declares that he is publishing Mercury solely because it is “salable” and not because it will be “usefull,” yet his manual does teach a number of methods in an accessible style (“Preface,” “To the Reader”). Both publications, however, suggest that books about ciphers and secret communication were popular reading matter for the public whether or not they were trained mathematicians or linguists. That Wilkins, Bridges, Falconer, and Davys wrote for a general audience is evident in the narrative structure of their manuals, which proceed methodically through contextual explanations and then examples arranged from the simple to the complex. Their intertextualities, too, connect them to traditions, inviting readers into a specialized discourse.
Figure 1.1The title page of John Davys’s An Essay on the Art of Decyphering (1737). Courtesy of the National Cryptologic Museum.
Morland’s shorter pamphlet, explicitly addressed to Charles II and a court audience, is obviously different in its structure, dispensing with dedications, epistles to the reader, historical contexts, and slow-moving technical examples to begin immediately with protocols and quickly complex illustrations. And perhaps not coincidentally, Morland’s New Method, if it has been at all noted, has not been read closely as part of the genre. Its technical language contrasts with the others. Even Davys, as he limits the discipline of deciphering to only those with training and the rare gift of genius, still uses a prose style that will attract a nonspecialist reader; for example, he consciously chooses (and states that he chooses) words like “guesses” and “steps” over the scientific language of “rules” and “methods” (“Preface”). He claims that although the reader will not “immediately commence Decypherer, yet I venture to promise him, that he will find several Things in them worth his Observation on so curious a Subject” (2).
Manuals of the 1640s–1680s
Wilkins’s Mercury was published in 1641, on the precipice of Charles I’s decline. Karl de Leeuw and Jan Bergstra, authors of a history of information security, describe Mercury as a popular science book that “attempted to inform a general readership of difficult subject matter but was also a timely gift to the diplomats and leaders of the Civil War.”9 Contrary to Kahn’s assessment that Wilkins was a theoretician but not a practitioner, de Leeuw and Bergstra believe he showed “solid mastery” of past and contemporary methods, pointing out that he corrects a mistakenly deciphered musical code from Godwin’s Nuncius Inanimatus, which Wilkins claims was the inspiration for his interest in cryptography. Mercury also closely follows the 1640 English translation of Francis Bacon’s Of the Advancement and Proficience of Learning; or, The Partitions of the Sciences.10 Wilkins discusses Bacon’s Advancement only briefly in Mercury, though the moment is worth noting. Calling Bacon the “English Aristotle,” Wilkins explains that Bacon treats cryptography as an art of grammar, “noting it as a deficient part,” and he seems to critique other writers on cryptography as referencing only Bacon and no one else, as if Bacon is the last word (Mercury 10). Wilkins is in the middle of a thorough historical overview of the role that cryptography has played in human communication since the Egyptians and in biblical passages, and though he may critique certain methods as inferior in wartime or political correspondence, he certainly does not present cryptography as “deficient,” but rather as underestimated.
Wilkins was fickle in his allegiances. He was at first a Royalist, part of Charles I’s intelligence circle and subscribed to the Solemn League and Covenant in 1643. He then publically changed his allegiance to serve Oliver Cromwell, vowing loyalty to the English Commonwealth in 1649 and even marrying Cromwell’s sister, Robina French née Cromwell, in 1656.11 Later, Wilkins would form new ties with Charles II, submitting to the Act of Uniformity. This fluidity of identity across regimes was not accidental or lucky chance. Shapiro notes that “behind Wilkins’s dynamism lay a remarkable ability to be in the ri...
