Military cryptanalysts – Prelude

Header picture: Senator Joseph McCarthy maps  subversives in 1950 America. courtesy Colorado State University Archives.

previous posts:


Wilfrid had acquired the manuscript by 1912.

In 1914, Hime published his claimed decryption of the  ‘gunpowder cipher’ from  Roger Bacon’s De secretis..

By 1917, in New York,  Wilfrid was touting his own “Bacon ciphertext” as having value for the American War department.

If he hoped in that way to get a high price for his manuscript, Wilfrid’s plan back-fired. His comments (and comments misinterpreted) alarmed some of his acquaintance and brought upon Wilfrid and his circle the attention of the American Bureau of Investigation (the BOI, which would become the FBI in 1935).   By 1917 there was a BOI file on Wilfrid, and it describes him as:

“An Austrian or a Russian Pole (Jew) & and pro-German & an anti-British naturalized British subject & a pretty slick article”

That file was not closed until 1920 by which time Wilfrid and various associates and acquaintances had been questioned.

It is not surprising, then, that by 1921 Wilfrid was concentrating on a different pitch, omitting any mention of military uses or of gunpowder ciphers in his talk, in that year, to the College of Physician of Philadelphia …  but it was a little late to change tack.

William Friedman in 1917
detail from a photograph in files held by the George C. Marshall Foundation

Already, a geneticist named William Friedman, who would be promoted through the Signals corps to become a cryptanalyst for the War Department in 1930, was curious about Wilfrid’s supposed ‘ciphertext’.  Though initially rebuffed by Wilfrid and his wife, Friedman attained the rank of Colonel and strong connections with military intelligence during the second world war, and by 1944 – before demobilisation and amid the rising ‘McCarthyist‘ atmosphere –  Colonel Friedman was able to acquire a full photostat copy of the manuscript and, eventually,  access to private correspondence and research notes generated within the Voynich family’s circle of friends and correspondents.

Friedman’s only interest in the manuscript lay in breaking the cipher which he supposed to inform the written part of the text. The investigations conducted by himself, his chosen assistants, and other members of the National Security Agency would have little impact on the study until 1978, when Mary d’Imperio summarised their (ultimately failed) efforts to ‘crack the text’.  Among the few who knew it before 2009, that book was already elevated to the status of a sort of ‘Voynich Gospel’ but its impact became ubiquitous after 2009, when it was put online as a pdf.

That event was communicated to Nick Pelling, who advertised it through a post headed. ‘d’Imperio’s “Elegant Enigma” now a downloadable pdf!‘, ciphermysteries, (March 22nd., 2009).

Since then, d’Imperio’s small book has informed every account given of the manuscript – from wiki articles to published volumes, blog- and forum-discussions and upon the basis of its text, innumerable theories, still-current, depend.

Thus, forty years after its publication, and more than half a century since Friedman’s first group began their work,  d’Imperio’s summary of their efforts is fundamental reference for the majority of Voynicheros.   Any revisionist daring to engage a root-and-branch re-evaluation of its worth for our ability to understand the manuscript must expect to run the gauntlet.

I’ll look at some of the cryptanalysts’ errors in coming posts.

  • [pdf] Mary d’Imperio, The Voynich Manuscript: an elegant Enigma, NSA. The original file, in a cleaned-up copy, was later made available. [pdf]
  • [pdf] Mary d’Imperio, ‘An Application of PTAH to the Voynich Manuscript (U)‘. Originally classified ‘Top Secret’ the paper was also released as a pdf in 2009 – minus a number of excised passages.  The illustration below is slightly edited, to make a more compact image.


William Friedman:

According to a publication by the National Security Agency, Friedman retired in 1955 “after thirty-five years of service with U.S. cryptologic activities”.

  • [pdf] ‘Friedman Legacy’, Sources in Cryptologic History, Number 3. National Security Agency.  3rd. printing, 2006.  (The pdf may be slow to load).

There, his duties are listed as:

  1. Director, Department of Ciphers, Riverbank Laboratories,1919-1920;
  2. Cryptographer, Office of the Chief Signal Officer (OCSigO), Washington, D.C., 1921;
  3. Chief Cryptographer, U.S. Signal Corps, 1922-1929;
  4. Cryptanalyst, War Department, 1930-1942;
  5. Director, Communications Research, Signal Intelligence Service, later Army Security Agency, 1942-1947;
  6. Chief, Communications Research Section, Army Security Agency, 1947-1949;
  7. Cryptologic Consultant, Army Security Agency, 1949;
  8. Research Consultant, Armed Forces Security Agency, 1949-1951;
  9. Research Consultant, National Security Agency, 1951-1954;
  10. Special Assistant to the Director, NSA, 1954-1955 (retirement);
  11. Member, NSA Scientific Advisory Board, 1954-1969;
  12. Special Consultant, National Security Agency, 1955-1969.
postscript 1: [Jan 5th., 2019].  After WW II, Friedman sought permission from the Army (NSA) to claim patent over unspecified ‘inventions’ for the Signals Corps which the need for secrecy had prevented being patented earlier.  The question went as high as the Army’s Judge Advocate General, whose request for advice from the NSA  was met by Col.Marcey’s asserting (8th. Feb 1953) that “‘from 1933-1942, Mr.(sic) Friedman was a civilian.”.   The document was posted at the Internet Archive.

John Matthews Manly

Playing a central, if quiet, role in the earlier events had been John Matthews Manly, a medievalist and a cryptographer, a friend to Newbold, and a scholar interested in the Voynich manuscript. He had been a member of MI-B (sometimes seen as MI-8) during World War I and had defended Wilfrid to the BOI (see MacKinnon, above). In 1931, shortly after Manly  published a paper about the Voynich manuscript in Speculum, a series of letters passed between Friedman and himself. for details of which see:-

  • David Kahn, The Comprehensive History of Secret Communication from Ancient Times to the Internet (1996).  Manly-Friedman correspondance for 1931 – entries numbered 361 and 362 in Khan’s ‘Notes to Text’ .

Manly died in 1940; Wilfrid had died a decade before.

  • John Matthews Manly, ‘Roger Bacon and the Voynich MS’, Speculum Vol. 6, No. 3 (Jul., 1931), pp. 345-391.
  • John Dooley and Elizabeth Anne King, ‘John Matthews Manly: The Collier’s Articles’, Cryptologia, Vol. 38 No.1 (January 2014) pp. 77-88. doi: 10.1080/01611194.2013.797049.
[postscript 2. Jan 5th., 2019] – I am grateful to a correspondent who refers to a book in the George C. Marshall Research Library, whose margins and flyleaves were filled with  William Friedman’s own comments and criticisms – written in ink.  On p. 39, Friedman disputes Yardley’s praise for  Manly’s ability as a cryptographer at MIB.   The same volume  has – a little ironically – a Mayan-design image as bookplate.  (see above, right).

and see


Rising hysteria – McCarthyism

In 1938, while Friedman worked as a cryptographer in the War Department,  the HUAC (House of UnAmerican Activites) was established “to investigate alleged disloyalty and subversive activities on the part of private citizens, public employees, and those organizations suspected of having Fascist or Communist ties”.

According to d’Imperio, the ‘first Friedman group’ worked on the Voynich ‘ciphertext’ manuscript between 1944-1946, while Friedman held the post of Director, Communications Research, Signal Intelligence Service, (later Army Security Agency).

Julius Rosenberg, 1951.

From the end of the 1940s, a latent xenophobia  had blossomed in America into ‘the second red scare’ and by the 1950s public expression was now common of sentiments similar to those reflected in the earlier BOI file on Wilfrid. Any foreign-born person – and a Jew of even the second or third generation –  was at risk of being targeted as ‘un-American’ –  something increasingly  identified with ‘Communist leanings’. Adding to  public actions by the HUAC and by McCarthy’s followers at this time were those of the  FBI ‘combating the  enemies within’. President Roosevelt gave the FBI a sweeping mandate to investigate fascism and communism in the United States, and to this end Hoover increased surveillance of those suspected, not excluding the use of  wiretapping.

It is, therefore, important to realise that when Brigadier John Tiltman‘s help was enlisted by Friedman in 1951, and Tiltman began to repair certain omissions in the earlier efforts – such as  actively seeking specialists’ opinions about the imagery –  Friedman had been thirty years in service to the US government and Senator McCarthy was the height of his public influence.  1951 was also the year  the Rosenburgs were condemned for ‘conspiracy to commit espionage”. Both man and wife were executed in 1953.

This temper of the times does much to explain why (for example) Mrs. Voynich now acquiesced to Friedman’s renewed demands.  Though a private person and not a Jew she had lived for a time in Russia and openly supported  revolutionary causes, though perhaps only until 1898. The BOI file of 1917 makes  clear that she was a better-known figure than her husband.

It also helps explain why  Erwin Panofsky should consent to reply to Friedman’s interogatory of 1954.

Neither Panofsky nor anyone else could predict, at the peak of the McCarthy era, that “only about hundred” academics would actually be dismissed from their posts, and those on the grounds of having held membership in the Communist party.  The ‘witch-hunt’ was in progress and Panofsky could not have been unware  that in the course of pursuing an ‘anti-Communist’ witch-hunt against Jewish actors, writers and film-makers, the first director of HUAC had invoked the so-called ‘Sedition Act’ of 1918 – one which expanded the  anti-Espionage Act, permitting classification of any foreign-born person – even if a naturalised American – as a ‘non-citizen.’ Panofsky had come to the US as an adult, a refugee from Nazi Germany. He had seen such things happen before.

On the number of academics dismissed, see Hooke’s hostile review of Ellen Schrecker’s  No Ivory Tower: McCarthyism and the Universities:

  • Sidney Hook,  ‘McCarthy and American Universities’, Minerva, Vol. 25, No. 3 (September 1987), pp. 331-348.

Hook’s own book was severely criticised.

  • [pdf] Mark deWolfe Howe, (review) ‘Heresy, yes – Conspiracy, no by Sidney Hook.. (1953)’, Yale Law Journal, Volume 63 (1953), Issue 1 pp.132-137.

More recent studies:

  • Melvin Rader, False Witness (re-publication with afterword by Leonard Shroeter), 1997.
    In the summer of 1948, with Cold War tensions rising, a young state legislator from Spokane, Washington, named Albert Canwell set out to combat the “communist menace” through a state version of the House Committee on Un-American Activities. University of Washington professor Melvin Rader was a victim of the Canwell Committee’s rush to judgment, but he fought back. False Witness tells of his struggle to clear his name. It is a testament of personal courage in the face of mass hysteria and a cautionary example of how basic freedoms can rapidly erode when the powers of the state are allowed to serve a rigid ideological agenda.


  • David R. Holmes, McCarthyism and Academic Freedom : Stalking the Academic Communist: Intellectual Freedom and the Firing of Alex Novikoff, (University Press of New England: 1989 and 1990)
  • Holmes’ book received a long review by Russell Jacob in the New York Times, (April 09, 1989).


Next post: Military cryptanalysts and Panofsky at Princeton.



Wheat from Chaff – Books of Secrets and the ‘Secretum secretorum’

Header picture:     (detail)  Brit.Lib. MS Harley 3719  f.31v  (1275-1540), excerpt from the Secretum Secretorum
This post follows from ‘Wheat from Chaff – ‘Bacon ciphertext’ (Dec.15th., 2018)


…  investigat secreta naturae …- Roger Bacon, Opus Maius, II

As we do what Wilfrid should have done, testing his assertions severally against the secondary  evidence and the manuscript’s own testimony  –  it becomes clear that Wilfrid’s “history” for the manuscript is anything but. His failure to establish the truth of what he said about the manuscript  before making those assertions public – and the failure of others to do so before adopting them, meant that the manuscript’s study was misdirected from the first, and researchers diverted into innumerable dead-ends simply because their first premises – their ‘givens’ – were groundless.  The habit is still widespread by which  a fictional narrative ( habitually if wrongly called ‘theory’) is first promoted and  its success thereafter evaluated  by the number of believers – but not  by what demonstrable value it has for a more accurate understanding of the manuscript – not of its materials, nor of what it was intended to convey. We’ll look at this in more depth when considering false analogies, false equivalents and ‘argument from association’. (List of logical fallacies)

His imagining the text written in Bacon’s own hand is unjustified.   To suppose   Bacon or anyone else might encipher an entire book about  ‘natural history’ in the mid-thirteenth century runs counter to the historical evidence; there is no evidence to support the idea that Bacon ever enciphered an entire book, either  – not about natural history or anything else.  And the content which Wilfrid imagined (‘natural history’) cannot be accepted at face value,  given the lack of evidence for Bacon’s ever being afraid to speak of such matter  and the wide enthusiasm which met Cantimpré’s book on that subject – despite its using Aristotle and being written in 1230-1245, during the ‘ban’ years:  (see previous post).

To believe that Voynichese is unreadable because written in a Baconian cipher demands, today, not only suspension of disbelief but its elevation beyond this mortal world –  for if Bacon’s methods of encryption were unsophisticated, modern decryption algorithms are not.  And there is the other point – that whether or not our present manuscript was made in Latin Europe, there is no way to know yet where the content was first enunciated, nor whether the text’s underlying language (if any) is European.  Such things are also treated as ‘givens’ – maintained without critical scrutiny since 1921 and despite the very obvious fact that if the text or the imagery conformed to Latin European practice, the latter, in particular, would not have proven illegible in those terms – for a century.

And finally, as we’ve seen,  Bacon’s sentence from De secretis..  (‘he’d be a fool….’ ) is neither a text’s justification nor the manifesto which it is so often portrayed as being.  It is, rather, part of an introduction to a subject  which Bacon thought might prove useful at some later time.

On the other side of the scale, one item is potentially in favour. Bacon was much impressed by the Secretum Secretorum. It can be called a book of secrets, was attributed to Aristotle and its title might suggest a need to keep its content secret, even though it is just as much a text in the tradition of the ‘Mirrors of Princes’.


‘Books of Secrets’ genre.


Eamon describes it:

… To the modern reader expecting to encounter some mysterious, arcane wisdom, these works are bound to be disappointing. What was revealed, typically, was not the lore of ancient sages or magi, but recipes, formulae, and “experiments”, often of a fairly conventional sort, associated with one of the crafts or with medicine: e. g., quenching waters for hardening steel, recipes for dyes and pigments, instructions for making drugs, and “practical alchemical” formulae such as a jeweller or tinsmith might use. When a medieval or sixteenth century writer claimed to have discovered a “secret,” he often had this meaning in mind; and when a contemporary library catalogue referred to a “book of secrets,” it usually indicated a compilation of such recipes…

….They exist in countless medieval Latin and vernacular manuscripts, and in printed books of almost every European language. Nor did these writings disappear with the “triumph of modern science” in the seventeenth century. Despite [Thomas] Browne’s warning, books of secrets continued to be written, copied, published, and read by a sizeable portion of the reading public, and by some whom all would agree were among the leading scientific personalities of the day….

  • William Eamon, ‘Books of Secrets in Medieval and Early Modern Science’, Sudhoffs Archiv, Bd. 69, H. 1 (1985), pp. 26-49.


That’s exactly how Roger Bacon understood the matter, too.  His ‘scientist’ is master of practical techniques and know-how –  as we see in Bacon’s letter to Pope Clement (1267):

Hence he  has peered into all the processes of smelters, of goldsmiths, of silversmiths, and of other workers in metals and minerals; he knows everything pertaining to war, weapons and hunting; he has examined everything pertaining to agriculture, surveying and other occupations of the countryman; he has even taken into consideration the experiments of witches and their fortune-telling and charms and those of magicians in general, likewise the tricks and illusions of legerdemain — so that nothing worth knowing might remain unknown to him and that he might know what to condemn as due to sorcery or magic.

Nick Pelling is among those who have  believed  that the Voynich manuscript may be a ‘book of secrets’.  For details see (e.g.)


“Confounding demons”       detail from folio 51r,   Brit.Lib. Yates Thompson MS 28

Aristotle’s  ‘Secretum Secretorum’:

Today we deny Aristotle’s authorship, but Bacon and his successors did not, so I have omitted ‘pseudo-‘ from the heading.

One reason Bacon expended so much energy on his project  [an edition and commentary on the ‘Secretum Secretorum’] is that he valued [the text] as possibly Aristotle’s greatest production. Not only did it impart sage counsel on issues of morality, politics, and health: Bacon, for example, ate rhubarb on the Secretum‘s advice, and felt much the better for it!   In addition, Bacon believed the Secretum treated important subjects deliberately omitted in the books Aristotle wrote for a general audience.

  • quoted from Steven J. Williams, ‘Roger Bacon and His Edition of the Pseudo-Aristotelian Secretum secretorum’, Speculum, Vol. 69, No. 1 (Jan., 1994), pp. 57-73.. the passage  is pp. 64-65.


Origin of the Latin versions:

Salim Abu l-ʿAla’, secretary to the caliph Hisham ibn ʿAbd al-Malik (r. 724–743), initiated the translation of the pseudo-Aristotelian letters on government to Alexander the Great.  This collection forms the nucleus of the most famous among the “mirrors for princes”, the Sirr al-asrar…, known in the Latin Middle Ages and early modern times as the Secretum secretorum. One of the Arabic translations of the pseudo-Aristotelian De Mundo also traces back to this period… However, it was under the ʿAbbasids (750–1258), and in particular in the first two centuries of their caliphate, that the translations blossomed.

A different account is given by Robert Steele,

In the introduction to the work as we now have it we are told that it was translated from Greek into Rumi, and from Rumi into Arabic, by Yuhanna ibn el-Batrik (or Ibn Yahya al-Batrik). Rumi is the common word for Syriac, when it does not mean Greek, and Yuhanna, who died a.d. 815, was a well-known translator, physician of Al- Ma’mun, who is said to have rendered the Politics and the Historia Animalium into Syriac, and the De caelo et mundo and the De anima in epitome, with other works, into Arabic. There does not seem anything obviously unlikely about the statements that a Syrian text has existed, and that it was translated into Arabic about the beginning of the ninth century by Ibn al-Batrik, while it is to be hoped that English scholars, at any rate, have dropped the pose that a manuscript attribution is a decisive argument against the supposed author or translator having any connexion with the work.

A curious confirmation of the possible existence of a Syriac version has lately turned up in the publication by Dr. Budge of a thirteenth-century collection of medical treatises and receipts in Syriac (Syrian Anatomy, Pathology, and Therapeutics, 2 vols., London, 1913). Among them (ii. 540) is the formula for calculating victory by taking the numerical value of the names of the generals and casting out the nines (see pp. Ix, 250). This formula is identical with one which exists in both forms of the Arabic text, though it is omitted in the Vulgate Latin version.

It is unlikely that the Syriac text, if it should ever be found, will bear the name of the Secret of Secrets. Perhaps the traditional name preserved by Al-Makin, The Book of the knowledge of the ‘ Laws of Destiny, or the Kitab-al-siyasa of Ibn Khaldun, its alternative title in Arabic, may afford some clue.

  • Roger Bacon, Robert Steele, Opera hactenus inedita Rogeri Baconi V: Secretum Secretorum (1909) Latin text; Steele’s notes in Eng.
  • Roger Bacon, Opera hactenus inedita Rogeri Baconi, 5: Secretum secretorum cum glossis et notulis, ed. Robert Steele (Oxford, 1920).

See also:

  • Steven J. Williams, The Secret of Secrets: The Scholarly Career of a Pseudo-Aristotelian Text in the Latin Middle Ages. (2003).
  • Yela Schauwecker, Die Diätetik nach dem ‘ Secretum secretorum * in der Version von Jofroi de Waterford Teiledition und lexikalische Untersuchung, Würzburger medizinhistorische Forschungen 92   (Würzburg: Königshausen & Neumann, 2007).

Yela Schauwecker here presents us with an Old French translation of those sections of the pseudo-Aristotelian ‘Secretum secretorum’ pertaining to diatetics, detailing which items are beneficial for each part of the body, and the qualities of various foodstuffs.  The edited text also features the much briefer passages which precede these and which relate to the benefits of doctors, prayer and astrology.   The translation is of particular philological interest because it involves the collaboration of an Irish Dominican author (Geoffrey of Waterford), composing in Anglo-Norman, and his Walloon scribe, Servais Copale. Until recently, Schauwecker’s base manuscript BNF, fr. 1822 was believed to be the only extant manuscript preserving Geoffrey’s translation.

  • quoted from the review by Alex Stuart in  Medium Aevum, Vol. 79, No. 1 (2010) EDITIONS OF TEXTS, pp. 160-169. passage occurs on p.167.


  • L. Saif,  ‘Textual and Intellectual Reception of Arabic Astral Theories in the Twelfth Century’, in  The Arabic Influences on Early Modern Occult Philosophy. (2015)
  • Lynn Thorndike, ‘The Latin Pseudo-Aristotle and Medieval Occult Science‘,  The Journal of English and Germanic Philology, Volume 21 No.4 (April 1922) pp.229-258.
  • M. Gaster, ‘The Hebrew Version of the “Secretum Secretorum,” a Mediæval Treatise Ascribed to Aristotle’.  Journal of the Royal Asiatic Society of Great Britain and Ireland (Oct., 1908), pp. 1065-1084 .
  • [pdf] Antònia Carré and Lluís Cifuentes, ‘Girolamo Manfredi’s Il Perché: II. The Secretum secretorum and the book’s publishing success’Medicina & Storia, X, 2010, 19-20, n.s., pp. 39-58.
  • Linda T. Darling, ‘Mirrors for Princes in Europe and the Middle East: A Case of Historiographical Incommensurability’ (no publication details given; paper available through


Next post: Bacon’s De Secretis in re  Hime’s “gunpowder cipher”