Wilfrid M. Voynich, famous book collector, has given to the Library of Congress a parchment manuscript of the Liber Sextus (Decretali), written in the XIVth Century under Boniface VIII, with annotations by Giovanni d’Andrea. He also presented a printed copy of the same work, dated 1514.
Italica, Vol. 5, No. 2 (Jun., 1928), “News Notes’ column, p. 42
From about 1917 to the mid 1920s, such brief notices are found in a variety of professional and scholarly journals, usually crediting Voynich with having provided, commented upon, or donated some item of real historical importance. Such notices refer to him briefly, but uniformly convey an impression that Mr. Voynich was now an admired, or at generally respected, expert and not simply another bookseller. In more than one case the inference appears to be that his name is sufficient evidence of authenticity. I’ll add an example below.
To this general air, in those years, two notable voices were publicly raised in opposition, neither questioning Wilfrid’s integrity as a dealer: both focussed on the supposed ‘cipher manuscript of Roger Bacon’. What they are criticising in fact, although ‘Voynich studies’ could not yet be said to exist, was that its approach and method were deeply flawed by an essentially romantic and a-historical approach to the object. The attitude is still pervasive, in my opinion.
One of those voices was Lynn Thorndike’s; the other George Sarton’s – he who is regarded as founder of the western-style history of science. Lynn Thorndike’s area of research made him perfectly aware of the original texts and historical setting informing the documented works of Roger Bacon.
Their antipathy towards Wilfrid, and to the ‘cipher manuscript’, may fairly be considered informed by a combination of intellectual indignation and personal distaste, but in Thorndike’s case the personal antipathy seems, to me, more informed by his genuine regard for William Romaine Newbold, whose well-being in life, and whose reputation thereafter was perceived as badly affected by what Thorndike in one review calls “Mr. Voynich’s pet manuscript”.
From a general perception of the manuscript itself as Bacon’s work, enciphered, to one which saw it as little more than a worthless curio, the devolution of public interest occurred gradually but might fairly be said complete before the second world war.
The handful of persons whose interest remained, between 1926 and 1949 (say) apparently kept that interest, and any communications between themselves, fairly private. Public perception was, apparently, that the issue had died with Newbold, or if not with him, then with Wilfrid’s death in 1930.
What remained was the more pragmatic and solid aspect of Wilfrid’s contribution to American letters: his enabling medieval and Renaissance manuscripts and printed works to be obtained from European sources and brought to America.
In that way his personal reputation seems to have survived the first phase of the debacle that has been ‘Voynich studies’.
In illustration I reproduce part of an article published in 1954, and for three reasons: first, to show the enduring reputation which Voynich then had as an expert in medieval manuscripts; secondly, to show that his reputation remained intact until the post-war period; and thirdly because the piece itself is so charming. Its author, for all his efforts to write a formal academic notice of find, could not prevent his own joy welling up. He has found a new fragment of Cistercian-English* music from the late thirteenth or early fourteenth century:
from Richard L. Greene, ‘Two Medieval Musical Manuscripts: Egerton 3307 and Some University of Chicago Fragments’, Journal of the American Musicological Society, Vol. 7, No. 1 (Spring, 1954).
In the autumn of 1951 I first inspected, in the Rare Book Room of the University of Chicago Library, that institution’s Ms. 654, ….Obtained from W. M. Voynich (1923) and given by alumni of the University.
As soon as the book was removed from its exhibition case and placed in my hands I saw that its fly-leaves, eight at the front and eight at the back, were fragments of an older music-book, with music in several parts and Latin texts. I immediately had photostats made and sent to Professor Bukofzer, who quickly characterized the compositions as polyphonic, English, and of a date around 1300 or a little earlier. He recognized, moreover, that one of the compositions, “In excelsis gloria,” was also found among the famous Worcester Fragments: No. 101 in Dom Anselm Hughes’s Worcester Mediaeval Harmony. As bound in the Ms., some of the leaves were upside down, and of course a portion of each was obscured by the binding, each pair of conjugate leaves in the Chicago Ms. consisting of something less than half of a leaf of the original music-book. But half a leaf is better than none. (pp. 27-28)
At the time that piece was published, and certainly when that Chicago ms was first acquired, Wilfrid’s name was a trusted one. And as the same piece proves, Voynich had access to works produced around the time that Bacon lived (1214/1220 – 1294).
* The abbey known in medieval times as Meaux [‘Melsa’] Abbey lay in Yorkshire. it existed between 1151 and 1539.
However, despite the range of languages at his disposal, and the considerable advantage he had by reason of his broad, daily and practical contact with medieval manuscripts, Wilfrid’s inability to think critically about his own instinctive response mislead investigations into the supposed ‘cipher manuscript’ from the start. And it was that carelessness which best explains the maddened and alienated comments from Lynn Thorndike. The historian could not abide histories formed chiefly by an imposition of contemporary prejudices on the past, even though that was the norm at that time, and one which more recent historians have argued is impossible to avoid. But Thorndike in this case particularly abhorred constant repetition, in relation to the supposed ‘Bacon cipher manuscript’, of ill-informed ideas of Roger Bacon as ‘oppressed, visionary scientist’. On the scientific front an equally weighty voice – perhaps somewhat more affected by nationalism and class consciousness – was that of the renowned George Sarton, best remembered today as founder of the western construct of the “history of science”.
One suspects that in the usual way, the growing influence of these men among others would have seen the ‘cipher manuscript’ fall from public interest after Wilfrid’s death, to become thereafter another minor curiosity known only to some few bibliophiles. By 1930 it had become clear that the process of the manuscript’s investigation had been deeply flawed by presumptions, a-historical attitudes, and hunts for the key to some possibly chimerical cipher.
By the sixties, if one projects from the state of affairs as they were in the 1930s, one would have expected that this manuscript, once termed an ‘ugly duckling’ and demonstrated as a kind of albatross for Newbold and others, would have been widely perceived (as indeed Kraus may have considered it) to be no more than a white elephant: a curio which no collector cared to buy.
“Voynich studies” had not progressed between 1910 and 1930. The few items called ‘certain facts’ – even today – had been published by Voynich himself before 1922.
It was, in one sense, an accident of history that the manuscript appealed to a couple of people with which its supposed origin and content struck a chord for its apparently intersecting some pre-existing professional or private interest, and so promising an avenue for greater expression and recognition of those interests if interest in the manuscript were revived.
There have been written a good number of surveys of who has been involved, and when, in ‘Voynich studies’. My purpose here is not to repeat such matter, but instead to highlight the points and reasons for divergence in study of this manuscript from the usual process of investigating an old manuscript of unknown origins, in an unread script.
Investigating the record in that way, one finds the distortions already established and proliferating by 1912, in an account which Voynich gives, nine years afterwards, of his discovery and reactions. The talk was delivered to the Philadelphia College of Surgeons in concert with a discussion by Newbold of his ideas, of Bacon’s life, and of his own discoveries as he perceived them. Both papers were subsequently published in a volume of the Transactions.
Wilfrid’s first remarks are a normal description of initial observations made by a competent appraiser – though his ideas about the imagery seem somewhat ambitious, given the paucity of evidence for any similar imagery in western Latin manuscripts, even now. He said:
“Even a necessarily brief examination of the vellum upon which it was written, the calligraphy, the drawings and the pigments suggested to me as the date of its origin the latter part of the thirteenth century.”
Had he been more specific, or compared each point to other known works of secure provenance, his talk might have been duller, but of greater value in the long term. In particular, one wishes that he had commented on what it was about the vellum, calligraphy, pigments or drawings which instantly suggested an origin in England.
From the first, however, Wilfrid’s inability to distinguish between the bibliopole’s system for providing a text with narrative-and-genealogy as provenance, and that of the historian’s attitude to establishing an object’s place in the historical narrative, meant that Wilfrid failed to examine the limits of his own skills, instincts and expertise.
Specifically: while it is clear that he cannot read the written part of the inscribed text, he never stopped to determine just why it was obscure. He did not wonder – just as example – whether it might be pure nonsense, or perhaps a form of ‘Greeking’ (though the thirteenth century is perhaps a bit early for that). He did not ask whether or not the language here might be other than European, or whether it might be a type of notation not directly linked to standard periods of prose, or verse. He appears never to have paused to reflect on the manuscript traditions which exist outside those of western Latin monasticism. He did not ask himself whether, overall, he had yet the skills needed to perfectly provenance the artefact, let alone attribute its text’s composition to a particular author.
These blind spots are just as plainly present in 1921, when he relates the process of his attribution of it to Roger Bacon, even before reading the seventeenth century letter within it:
“I hastily considered the question of the possible authorship of the work, and the names of the only two thirteenth-century scholars who could have written on such a variety of subjects [as Voynich imagined he saw] were, first Albertus Magnus… and secondly, the Franciscan friar, Roger Bacon, who had been persecuted…and whose scientific discoveries had been misrepresented as black magic…”
His definition of ‘scholars’ is limited to the Latin monastic type. He presumes ‘authorship’ for the work. He never questions whether there might not be a problem in the fact that he himself can ‘think of only two…’.
The same habits are endemic, today, in Voynich studies, but there is perhaps as much justification for them now as there was for Wilfrid. His interest was in documents of historical and pecuniary value; most modern researchers have little interest in anything but hitting the ‘pay-dirt’ which they associate with solving the presumed cipher.
In all this, the manuscript itself is permitted only the most minor role. In effect, “Voynich studies” has ceased to have any more than the most tangential relationship to the object – of vellum, ink, pigments, binding materials and inscribed forms – which in the usual way are the chief focus of a manuscript’s first assessment.
Then, again, there is the custom of writing about the manuscript in terms more appropriate to a popular novella. Wilfrid and Newbold both explained the text’s assumed “encipherment” by reference to the then-current idea of Bacon as “suffering scientist and quasi-modern”. It was a habit which infuriated Lynn Thorndike and which he hoped to dispel by repeated terse dismissals. These did have their effect, but by a slow and accumulating erosion, the effect of which was not obviously apparent as yet.
Thorndike – and doubtless others – had tried to stem this romantic attitude to Bacon but it had clearly gained in popularity, and evidently all the more after the supposed ‘cipher manuscript’ had been known in America. In a more general context, and as early as 1915, Thorndike had written:
“…it can not be shown that in the thirteenth century the church persecuted men of science. Rather, popes and prelates were their patrons.”
Lynn Thorndike, ‘Roger Bacon and Gunpowder’, Science, New Series, Vol. 42, No. 1092 (Dec. 3, 1915), pp. 799-800.
His objection affected popular attitudes not at all, so far as one can tell, for at least the next decade, and even today the legend persists in some writings. Popular legends are popular for a reason: they are easier to remember and need no real evidence to be given which might slow the narrative.
But to be fair, I think that the persistence of the belief (which is not disproven) that the content of the Voynich manuscript is a “Bacon cipher” is due in large part to the fact, which appears indeed to be fact, that Wilfrid Voynich was a highly regarded expert in appraisals for western medieval manuscripts.
In illustration, here is part of the ‘Notes and Quotes’ column published in July 1921, in the Catholic Historical Review. It appears to be based on the talks given by Wilfrid and Newbold, perhaps supplemented by personal interviews. It was published before the full text of those papers, which the College included the following year in its Transactions.* This author clearly subscribes to the ideas which at the time were considered thoroughly modern and ‘revisionist’ in regard Roger Bacon:
Bacon paid the penalty for being in advance of his time, for the trend of his studies earned for him a reputation of dealing in magic and the black arts and even threw suspicion on his orthodoxy. Although his life was a long one, the ten years he spent under strict supervision at Paris and with an inhibition against writing anything for circulation were necessarily great checks on his productivity.
* Transactions of the College of Physicians of Philadelphia (Third Series, Volume 3. Online – see Table of Contents here).
Newbold presented papers, or the same paper, more than once. A ‘Notice’ in the journal of the Classical Club of Philadelphia records that Newbold addressed them on April 7th., 1922, to an audience twice the size of the previous Friday meeting’s: fifty over twenty-two. (Classical Weekly, Vol. XV, No.24 (Whole No.240) p.192.
His flaws were elaborations on those flaws in Wilfrid’s initial assessment of the manuscript, and to Wilfrid far more than to Newbold, we must attribute the very early de-railment of this manuscript’s study. Wilfrid’s habit of confusing the construction of a ‘logical chain’ with establishment of verifiable historical data began then, continued with the study’s revival under Brambaugh, and still besets the subject to the time of writing.
I should still be willing to grant that the general ambiance… the ‘heft’ of the volume … conveys an impression that it derives from the fringes of the Latin world during the late thirteenth and earlier part of the fourteenth century. To this, the dating of our present manuscript’s vellum offers no real contradiction: it merely suggests that the fifteenth century copy was faithful to one or more precedent works.
It is notable that among the few English specialists of whose opinion we have record, most did not disagree with Wilfrid’s evaluation on that point. Hesitation and caution are expressed, but not outright disagreement of his proposed thirteenth or early fourteenth century date.
What there is clear and general refusal to confirm is Wilfrid’s assertion that, because he could think of “only two” persons as possible “authors” for what he imagined the manuscript to contain, an attribution to Roger Bacon was a reasonable conclusion. It wasn’t, as they knew.
Wilfrid’s approach to ‘history’ and ‘art analysis’ was to hunt for records which might support his own, pre-determined, expectations and aid his efforts to persuade others in the absence of fact or appropriate, validated, comparisons. From the first, that practice became a kind of template for the self-defeating habits still characteristic of most efforts in “Voynich studies”. The usual pattern has been first to come to the manuscript with a ‘gut-feeling’ which is termed a ‘theory’, then to hunt within its parameters for snippets of information which are then used to adorn a ‘logical chain’ of argument, rather than having a broad understanding of any aspect of manuscript studies – including history – precede the formation of any theory whatever.
As it happens, however, this seminal flaw was never recognised among those for whom the manuscript was only of interest as the source for a presumed cipher-text. In decryption, the method so antithetical to solid academic studies in the critical sciences, is perfectly appropriate to tackling a written piece which is known, or reasonably believed, to be enciphered. The ‘theory-first’ approach is considered normative, practical and scientific. So once again, the template was never discarded, nor its inappropriateness recognised. And despite the protests of some few – chiefly Nick Pelling – no interest was shown even as late as the 1990s (or if felt, not pursued) in close assessment of the basic indicators: membrane finish, dimensions, inks, pigments and so on.
Attitudes to the imagery, and ideas of what constitutes valid comparison between images, largely remain an embarrassment in terms of the formal study of manuscripts. Most persons involved have taken their ‘Voynich theory’ and tried to produce interpretations of history or of imagery which in their view will accord with it. The acid test is to ask any ‘Voynich researcher’ about the body of data and observation from which their ‘theory’ was adduced. As often as not, there is none at all, or second- or third-hand repetition of unproven assertions gained uncritically from older publications, or from seemingly authoritative overview-sites published online.
System error creates error exponentially, and so it is hardly surprising that entire ‘parallel universes’ have then been established with regard to only one section of the manuscript, and that only by reference to theories about the subject of certain imagery, gained from imaginary first-premises. The corpus of the fantastic ‘Voynich European herbal’ is just one case in point.
From the outset, the obvious questions were never asked. Is the manuscript and/or its language more or less likely to have originated in the Latin-speaking world? Is it the work the product of one ‘author’? Do the materials and the construction accord with presumptions about the imagery and about the written part of the text? In all honesty I can say that none of questions were asked, or their asking welcomed, by those working on the manuscript when I came to consider it in 2008.
In the normal way, manuscript studies and authentications focus first on such things, so that omission of any serious evaluation of the object itself to that time is striking to any informed reader. Perhaps more understandable in the 1930s, its continuing omission is more difficult to understand. And while one does accept the opinions since provided by McCrone on the inks and pigments, and by the University of Arizona on the vellum’s date, the limits under which both sets of laboratory tests were undertaken, especially limits on the scientists’ discretion, is much to be regretted.
After the 1930s, the ‘Bacon cipher’ issue seemed simply to die away. It would be briefly revived in the 1950s, and then more determinedly by Brambaugh after the manuscript’s deposition at Yale.
What happened outside the public arena in the interval will be the topic of the next post. But the state of ‘Voynich studies’ at the time of Wilfrid’s death may be encapsulated by Thorndike’s comment in reviewing Grubb’s post-mortem publication of Newbold’s writings and notes.
Our chief criticism of [Burke’s translation of Bacon’s Opus Maior] is that it should have been more fully developed before publication. The other [book] should never have been published at all…. I should like to be able to force everyone who asks my opinion of the Voynich manuscript to read [Grubb’s volume] from cover to cover. I think it will either kill or cure.
American Historical Review, Vol. 34, No.2, (Jan 1929) pp.317-319.
Fortunately, Wilfrid had become better and more widely known as a reliable expert in medieval and renaissance manuscripts and printing. For a while, at least, his name was still intact.
Biography: Readers interested in Wilfrid’s character and the impression it made, might enjoy a book recommended first to the old mailing list by Dana Scott. Details here.