Military cryptanalysts: Friedman and his questions

Header Illustration: detail of plate on p.11 of the Supplemental Volume of De Re Diplomatica. Issued in 1707. following Mabillon’s De re diplomatica libri VI, published Paris, Louis Billaine, (1681). Note – the example is chosen at random; no argument should be inferred.
Previous two posts


Friedman’s questions:

courtesy of the artist.

I’ve spent the last three posts explaining the background to Friedman’s questions and Panofsky’s responses. The aim is to understand  why Panofsky says so little about the manuscript’s pictures and why his responses lack his usual warmth and erudition.

In the next post, when we look at Panofsky’s replies in full, we see that the questions caused offence: some by ignorance of good manners; others of art, of manuscript studies and of Panofsky’s work.  (Remember, everyone had two years to think about the meeting).

A number of the questions have nothing to do with Panofsky’s interests, but are just about Friedman and his theories. Some assume Wilfrid’s narrative as ‘given’.  Others make clear that Friedman had scarcely attended to what Panofsky had already told him.  And others show extraordinary lack of awareness –  as e.g. Q.3, Q.7, Q.10,  Q.13 (!!!)  and Q,15.

Take Q.13 for example:   d’Imperio says (Elegant Enigma p.42) that William Friedman was “a devoted student of the Voynich manuscript from the early 1920s on”, yet Q.13 shows that thirty years later the ‘devoted student’ had not even heard of  the (then-) fundamental texts in European palaeography:  Mabillon’s De re diplomatica and Capelli’s Dizionario di Abbreviature Latini ed Italiani (Milan, 1912).

No matter that neither includes any simple comparison alphabet; the point is that in thirty years Friedman had not advanced his study as far as the introduction to ‘manuscript studies 101’.  Nor does he seem to have realised, to that time, that dating and (if possible) placing the  script is a vital part of provenancing any manuscript.   [see earlier post, ‘Provenance: Wilfrid-style and real-world.’ (November 24, 2018)].

Friedman’s exaggerated sense of self-importance and expectation that others should serve his needs does not seem unusual for him.  A number of comparable incidents are recorded by d’Imperio.



“On 25th May, 1944 William F. Friedman wrote a letter to the widow of Dr. Wilfrid Voynich  .. requesting a photostat copy [of the entire manuscript]. The request was granted.” (Elegant Enigma p.39)

Ethel Voynich (1864-1960) photo courtesy Kotbeber

The war had not ended; Friedman was –  according to the NSA biography –  Colonel Friedman, Director, Communications Research, Signal Intelligence Service, (later Army Security Agency).  During war-time the army has power to requisition, and one does not refuse a Colonel’s ‘request’.  The inconvenience and expense was not minor – the cost about that of a week’s wages for a man.   Mrs. Voynich first wrote to Friedman, pointing out that copies existed already, among them one in the New York Library and another with Fr. Petersen – but   Friedman clearly preferred  to have her bear the cost and trouble of providing him with his own copy; she complied. (Later we learn that Friedman  also obtained Fr. Petersen’s copy ‘on loan’ – effectively preventing that scholar from continuing his own decades’ research).

So – again in connection with ‘making sport’ of Newbold –  d’Imperio reports (p.42) that Elizebeth Friedman gave “an amusing account of the sport which she, William and Manly had together in demonstrating the ‘decipherments’ that could be had from Newbold’s texts…’

It was an insensitive thing to do to involve Manly, Newbold’s friend, in such ‘sport’  whether before, or after, Newbold’s suicide in 1926.


Newbold’s suicide:

At the time, it was not done to refer openly to suicide.  The act was considered a crime by the state, a shame upon the family, and a deadly sin by the Christian churches, so the usual practice was to add the oblique  ‘suddenly’ to an obituary’s regular formula  -such as  ‘died in hospital’;  ‘died at his home’ etc.  This I take too as the implication of Newbold’s not being recorded as buried from a church, but only that “A memorial service was held for him in College Hall on the University of Pennsylvania campus.”

Works other than d’Imperio’s Elegant Enigma  use “died suddenly.”

e.g. IN MEMORIAM, The Phi Beta Kappa Key, Vol. 6, No. 8 (May 1927), pp. 526-537. Entry for Newbold is p.535.


.. My point is not that Friedman had faults, but that when he commandeered the study from about 1952 or so,  all commentary from better qualified people ceased.  In fact the study of the manuscript itself ceased, and NOTHING by way of research was published for almost a decade, from 1953 until 1962 when Mrs. Voynich sold the manuscript to H.P. Kraus. What research was done was being circulated among the NSA cryptanalysts in-house or issued as very general popular articles.   As we’ve seen,  some of the NSA documents, including Tiltman’s paper, remained classified “top-secret” until the early 2000s.  In Jim Reeds’ Voynich Bibliography  publications for 1953-1962 include only these:

  • 1953   E. Westacott, Roger Bacon in Life and Legend. New York: [Publisher?],1953. [A balanced writer whose errors are flaws in his sources rather than his apprehension of them. The text is online through the internet archive. – D.]
  • ? W.F. and E.S. Friedman, ‘Acrostics, Anagrams, and Chaucer’, Philological Quarterly 38 (1959), pp.1-20.
  • 1959   Jose Ruysschaert, Codices Vaticani Latini 11414 – 11709. Bibliotheca Apostolica Vaticana, (1959). [Describes the MSs acquired by the Vatican from the Collegium Romanum, and mentions that W. Voynich bought a number of them which have been transferred to various American libraries, including the VMS.].

Note the above item, first noticed and commented on by Reeds,  has recently been brought again to notice and much emphasised by Rene Zandbergen under the rubric ‘ 1903 catalogue’ because a lost document which was not a catalogue but which listed a number of books and was – as Zandbergen describes it – dated to 1903 was photographed at some later time and t(as Zandbergen describes it), the list or that photograph was what Ruysschaert was referring in 1959.   Zandbergen has shown a certain impatience with persons trying to clarify his line of argument and evidence on this point, and I recommend any revisionist attempt the task for him/herself.   (Richard Santacoloma’s puzzled comments are perhaps a little more indignant than the confusion warrants – but you must judge that for yourself).

1962 H.P. Kraus, Catalogue 100. Thirty-five manuscripts: including the St. Blasien psalter, the Llangattock hours, the Gotha missal, the Roger Bacon (Voynich) cipher ms. New York: H.P. Kraus,1962. [Beautiful reproductions of several leaves of VMS.]

and in the same year (1962)

  • June 25th., ‘Kraus Marks Anniversary With Catalog of Treasures’, Publishers’ Weekly, 181 (25 June1962) pp. 39-40. [Kraus auction – Vms listed but didn’t sell.]
  • June 26th., David Kahn, ‘The Secret Book’, Newsday. 26 June1962.
  • July 18th., Sanka Knox, ‘700-Year-Old Book For Sale; Contents, In Code, Still Mystery’, New York Times, 18 July1962, p 27, col 2. [Kraus auction. Includes picture of 85/86r4. .]
  • August 5th., Elizebeth S. Friedman, “The Most Mysterious Manuscript” still an Enigma’, The Washington Post, 5 August 1962, sec. E, pp. 1,5.
  • 1963 Jan. Alfred Werner, ‘The Most Mysterious Manuscript”, Horizon, 5 (January,1963), pp.4-9.

… in all, nothing was published which might return the study to normal channels…

For those who believe the text is in cipher, all the above may seem fair enough.  For those who doubt it, Friedman’s involvement and the long ‘block’ on the manuscript’s research hardly helped.

His ill-informed (and historically un-balanced) assumptions infused those of the NSA, as we’ll see later, by considering d’Imperio’s work, including its Table of Contents and Index (which will highlight their assumptions, and their information-gaps, respectively).

Those privvy to the NSA groups’ efforts, and who contributed something of lasting value, were those who contented themselves with making observations that they tested rigorously before sharing them: Currier’s work is well known; some of John Tiltman’s observations were much to the point.

Friedman’s ‘teams’ looked at what his own inclinations dictated; his ignorance of, and indifference to, anything but cryptology when combined  with his arrogance alienated the more learned –  and  surely lost us the chance to have two early and expert commentaries in particular:  Panofsky on the manuscript’s imagery and codicology, and Salomon on the script. (It is also noticeable that d’Imperio’s Index lists Charles Singer but makes no mention of Dorothea.).

Lacking the weight which such scholars might have brought to the study,  Wilfrid’s first imaginative ‘history’ was soon to spiral into pure fantasy about the content.

Apart from individuals such as Currier, the Friedman groups early came to imagine that the manuscript must belong to the late sixteenth and seventeenth centuries and for the supposed connection to the mad emperor Rudolf II, that its content must relate to occult-alchemical ideas fashionable among the nobility in Prague at that time –   several generations after the manuscript had been made in a clearly different environment.*

* four samples of vellum taken from the top 11 quires returned an adjusted radiocarbon range of 1404-1438 and the volume has been assessed recently as being made  in northern Italy.


Lost opportunity: Richard Salomon on the script…
Michael Tangl. photo courtesy of the Bavarian Academy of Sciences.

Richard Georg Salomon (1884-1966). Source unspecified.

‘Gallows’ glyphs – .

‘Gallows’ figures proper do not occur before the sixteenth century and are set in letters to warn the carrier to make haste.  The forms which are habitually mis-called ‘gallows glyphs’ or ‘gallows’ letters in  Voynich writings have no such intent  – so far as we know.

Jim Reeds investigated Capelli’s Dictionary in 1994, sharing what he saw in Plate IV  (Mon Jun 9th 1997), and quoting its Italian caption.  Salomon and Panofsky had doubtless seen this illustration before Panofsky put it in his reading list for Friedman.

Thus Reeds: “Tavola IV … shows a letter  ‘1172, Giugno 13 — Savino abbate del monastero di S. Savino in Piacenza investe il mugnaio Gerardo Albarola per se e suoi eredi maschi in perpetuo, di un mulio di ragione del detto moasstero — Scritura carolina. — Pergamena origen., conservata nell’Archivio di Stato di Parma, monastero di S. Savino.” with glorious gallows letters all over it.”

Jorge Stolfi (Fri, 6 Oct 2000),  gave that information again to someone who’d missed it, translating  “The date is on the “letter” itself, 13 june 1172. It is actually a notarial document recording the concession by the abbey of S. Savino in Piacenza of a mill of theirs to miller Gerardo Albarola and his heirs in perpetuity etc. etc. As I remember, it is signed by the abbot, several monks as witnesses, the miller (not sure), and the public  scribe / notary who prepared it .”

Reeds’ find is now seen everywhere, though rarely with any mention of him  – which omission inevitably leads to the newcomer’s supposing the careless copyist, rather than the researcher, should be credited with a particular contribution to this research;  failing to go to the original discussion and so (not rarely) to waste their time re-researching and re-discovering things long ago discovered.  Pelling once called this the Voynich ‘ground-hog day’ phenomenon and it is due almost entirely to absent or erroneous attribution.

The same example shown above, together with other items appear on a page from Rene Zandbergen’s website, re-presenting a selection of material from the past century’s shared research.

Stolfi’s last phrase provides the key:: “public scribe/notary”. Such elongated ascenders are most often found in documents of this type i.e. deeds of gift; deeds of establishment and other property-related matters and can be traced to similarly religio-legal documents as early as the tenth century in Spain.  For a time a more ornate variant was used by scribes in the imperial scriptorium, but as I noted when treating this point and introducing the early examples from Iberia:

“The eleventh and twelfth centuries, lingering into the fourteenth, are when we see such forms in various parts of Europe, usually as part of some official decree or charter”.

  • ‘Who wrote the ‘gallows’, voynichimagery, Oct. 7th., 2015.

If the apparent similarity between some Voynich glyphs and these earlier scripts is not deceptive (something which Salomon might have told us), then it is another item indicating that the content in the Voynich manuscript predates by some time the present volume’s manufacture in the early decades of the fifteenth century.

  • Capelli, Lexicon abbreviaturarum: Dizionario di abbreviature latine ed italiane, 6th ed. (Milan: Ulrico Hoepli, 1979).
  • [pdf] Heimann and Kay (trans.),  Capelli’s ‘Dictionary… ‘ (1982). This pdf has no plates.
  • [read online or pdf] the internet archive has an edition in German to the front, and Heimann and Kay’s translation to the back.  Includes between them Capelli’s Plates.

Friedman’s  character and self-important attitude could be predicted to alienate Erwin Panofsky and others of his standing in their own fields. His errors – including uncritical acceptance of much of Wilfrid’s quasi-history and Newbold’s categories – then created error exponentially.


…. but to return to 1954 – all things considered (and though you are free to differ) –  it seems to me that Panofsky had reason enough to give Friedman responses which said as little as possible, being restrained by caution; by awareness of the temper of the times; and by   knowledge of by whom, and to what end, his statements might be used. Whether Friedman already had access to Panofsky’s assessment of 1931, or whether Panofsky knew he did, if so, are other questions still undetermined and unaddressed.

The list of Friedman’s ‘Questions for Professor Panofsky’ (below) comes from Jim Reeds’ original paraphrase  (Reeds’mailing list, Friday April 15th., 1994).

The Questions

  1. Have you examined the VMS itself?
  2. What is it written on; with what writing tool?
  3. What’s the date?
  4. Why do you think so?
  5. What’s it about?
  6. Are there any plain text books sort of like the VMS?
  7. What plain text have you found in the VMS?
  8.  What plants, astronomical, etc, things have you recognized?
  9.  Is it all in the same hand?
  10. Why was it written?
  11.  Where & when?
  12.  What do you think of the Roger Bacon theory?
  13.  [provide Friedman with…] Full title of the Dictionary of Abbreviations. Title of Hans Titze’s book on forgeries, & of Mibillon’s history of diplomatics.
  14. What other scholars are interested in the VMS?
  15. What do you think of the artificial language theory?

Afterword: What’s Wrong with that?

It is understandable that a reader with little prior background might wonder if there’s really very much wrong with those questions. For those not asking this rhetorically, I provide more detail. (click the small black arrow).

As always, the things not understood manifest in absence, and silence, so let me illustrate Panofsky’s capacity for analytical-critical commentary, and then consider what we might have had from him if Friedman had better understood the discipline of iconographic analysis, or the calibre of the man to whom he had been introduced.

Consider, for example, the “ladies” pages in the manuscript, and their curious gestures.  Now, here’s Panofsky’s commentary on one, simple, everyday gesture – a ‘snapshot’ from daily life: a man lifts his hat.

  • [Introduction] Studies in Iconology: Humanistic themes in the art of the Renaissance.

It makes no difference for our needs that modern scholars differ about the relative value of Panofsky’s analytical system, or debate his preference for ‘authorial’ art, nor even debate the relative value of scholarship he produced during his German period as against his time in America.  It should be obvious enough from that one example what a depth of commentary he might have made had he been simply asked to share his thoughts on the manuscript’s imagery, or even just on the figures of the ‘ladies’.   Had he not been approached in the way he was, or sent that prescriptive ‘quiz’, the manuscript’s study might have advanced far more rapidly, and along very different lines, than it did after 1954.

Friedman’s single-minded focus on the written text; his implicit belief that he was the most important person to study the manuscript;  his belief that it was ‘enciphered’ or ‘encoded’ reflect habits of mind which made him such an effective code-breaker (self-confidence; self-reference; self-sufficiency;  single-mindedness; unswerving determination and a habit of organising information into neat categories for cross-reference) also made him utterly unsuited to conceiving of the range and depth of learning which might be needed to understand so problematic a manuscript –   or even to have Panofsky open up on the subject.

I find it telling that even Brigadier Tiltman’s paper of 1968 misspells Panofsky’s name and that, despite the amount of time Tiltman spends talking about the imagery, he refers in that paper more often to Charles Singer – a writer of popular histories of medicine and science- than to Panofsky.[note]   One remark – unattributed – may be Panofsky’s, because it is  the first instance I’ve seen so far of any cryptanalyst recognising the fundamental distinction  between provenancing manufacture and provenancing content. (on which see  ‘Provenance: Wilfrid-style and real-world’. ().

Professor Panoffsky [in the questionnaire] and the keeper of the manuscripts at the Cambridge Library both independently insisted on a date within 20 years of 1500 A.D., and the manuscript as we have it may be a copy of a much earlier document.

In fact this mis-represents the case. What Panofsky said is that if it hadn’t been for [O’Neill’s claim to have identified] the sunflower as the subject of one image, he would have dated it to no later than 1470.

  • [pdf] John Tiltman, ‘The Voynich manuscript: “the most mysterious manuscript in the world”‘(1968). Paper released by the NSA under the Freedom of Information Act, 23rd April 2002.

Note added Jan.17th., 2009.  I’ll come back to Charles Singer, in connection with the ‘S.E.P.’ phenomenon, and do him more justice than the brief mention above. Since my first degree was a double major in art and in the archaeology of industry, Singer (editor of the first encyclopaedic ‘History of Technology’) happens to be one of my early heroes.