Anomalies – consider this…

This post – call it an essay – is a little over 3,000 words long.

In the later years of my studying Beinecke MS 408, I found that most of the digging was about reconciling and explaining apparent contradictions and anomalies in the primary evidence or secondary Voynich narratives. Sometimes the problem proved to be a consequence of old ‘Voynich doctrines’, theories accepted without scrutiny and so on. At other times the problem evaporated as I learned more. The ‘Mark Twain’ situation.

At the moment, I’m interested in another such knot -part palaeographic, part historical, and part contextual. I could stop blogging while I try to sort it out, but most of it isn’t in my field and I’m not sure I will be able to sort this one. But perhaps you may.

So welcome (I suppose) to my current ‘anomaly’.

The problem.

The ‘4’ shape, whether as numeral or as cipher-form is not attested, in the western side of the Mediterranean or in western Christian Europe (i.e. ‘Latin’ Europe) earlier than about the mid-fourteenth century, with just one exception known to me so far.

However, that closed, upright ‘4’ shape – whether or not as numeral – is present in the Voynich manuscript and (this is the issue) with glyphs having what I described in a voynichimagery post of 2015 as ‘elongated ascenders’ -observing the trend towards plain English.

(detail) Yale University, Beinecke Library MS 408 f.5r

Here’s what I mean by glyphs with elongated ascenders. As you see, they occur in the Voynich script in positions apparently other than the initial, and they occur in close proximity to glyphs of that upright ‘4’ shape – and others very similar but having the left side of the ‘eye’ a little more curved. (I don’t mean the ‘9’ shape).

The term ‘gallows’ by which the ascender glyphs are known today in Voynich studies is a complete misnomer. The true “gallows sign” describes a motif used in letters from Tudor England, urging the recipient to reply on the instant. It doesn’t look much like these.

In passing, I’d like to draw attention to the mark (diacritic or vocalisation?) over the glyph you see as the second of the second paragraph in the clip above from folio 5r, and then as the first glyph of the following line. A mark similar to that in the second instance appears among the central inscriptions in the month-folios.

The easiest way for me to explain why I regard a ‘4’ shape in proximity to elongated ascenders as a problem, is to first indicate the date-range for ‘4’ shaped numerals, then refer to Hill on the history of Arabic numerals in parts of Europe, and then return to the issue of elongated ascenders – as well as I can from my own research into their history within Europe.

I’m fairly sure that anyone competent in comparative studies in palaeography might dispose of the whole problem in a sentence or two, but none were publicly involved in the study in 2015, and I have no introduction to any now.

The earlier post that I’ll quote from, further below, was entitled ‘Who wrote the gallows?’, and was published in October 2015 through the voynichimagery blog.

I’ve had little difficulty in finding examples of a ‘4’ shape compatible with even the earliest date returned by the Vms’ vellum samples, one of which returned a raw date of 1400 AD. I’ve shown two of these early instances already, but to save scrolling back, here they are again:

detail from: Abraham Cresques’ pictorial compendium, made in Minorca for the court of France. completed 1375. Known as the ‘Catalan Atlas’.

I also consulted Smith’s History of Mathematics (2 vols), but now must express my intense gratitude to the friend and master printer who has sent me a copy of a rare monograph:

  • G.F. Hill, The Development of Arabic numerals in Europe (Oxford: Clarendon, 1915).

Hill’s work, like most products of the nineteenth-century Anglo-German school of thought, has definite benefits for this study, but certain deficits too of which readers should be aware.

On the positive side: In 1915 an English scholar might still take his time to amass information for a projected monograph or book, taking ten, twenty or even forty years to finally produce as complete a study as possible, one to be of value for decades to come. As a result, one may still find that a book written last century contains information of greater detail, depth and range than one published last week when it comes to technical and historical studies.

Hill also had the advantage of living in Europe before European countries adopted the Mongols’ ‘total war’ approach, so he was able to consult scholars, manuscripts, libraries and collections which, by the end of the second world war, would exist no longer.

On the negative side, Hill’s study suffers from the same presumptions as those affecting his contemporaries Wilfrid Voynich, William Newbold and William Friedman, all of whom were born in the nineteenth century and reached maturity before the beginning of the twentieth.

Like them, Hill defines ‘Europe’ as England, Germany and Italy, with a just glance towards France and no reference made to the Iberian Peninsula, the Balkans, or elsewhere unless an Englishman, Italian or German was involved. Italy was included by virtue of the Renaissance and ancient Rome, and Ramon Llull was acceptable to the older Voynich theorists only because the Voynich text was presumed encoded or enciphered and his was the only name that came to mind in relation to Panofsky’s saying (in 1932) that he attributed the manuscript to ‘Spain or somewhere southern’. Llull’s name has been floated, intermittently, since the 1920s and is risen to the surface again recently – who knows why – but I doubt if one in a couple of hundred Voynich writers has bothered to read anything of Lull’s writings. There is absolutely nothing of his worldview in the Voynich images, and his system for memorising texts has about as much connection to the Voynich plant-pictures as a greeting card to the Boboli gardens.

Spaniards and even the French were simply not considered in weaving a Voynich narrative and that remained so from 1912 almost to the present day. As for writings of non-Europeans, or even of European Jews, they are not included in Hill’s work, any more than in the Friedman’s idea of medieval Europe, or the names considered as the Voynich manuscript’s possible ‘authors’. d’Imperio’s including medieval Jews assumes their work relevant only to religious/magical/superstitious matter and even then envisages it having been filtered and ‘translated’ by some Christian intermediary. What is embarrassing to us, these days, is her being evidently oblivious of the Jews’ scientific literature and their role as scholars and not merely as hired translators.

However, one simply has to strive to correct such old and sadly persistent blind spots in relation to Voynich studies, while observing a solemn silence over the still more embarrassing and ill-informed, if mercifully few, pseudo-Jewish ‘Voynich’ theories.

It was precisely because I’d come to know where the older writers’ blind spots were that I went to those areas first – the Iberian peninsula, the Balearics, the artisans and masters of crafts, the Jews and the mariners, the traders and cartographers, .. and found those two early examples immediately, though admittedly both are sources I know well and to which I’ve had reason to refer before in explaining drawings from the Voynich manuscript. There are surely more fourteenth-century example to be found, but those two suffice for now.

Within Hill’s predictable limits, his study remains very valuable indeed. Here’s a clip from his annotated Tables – which I’ve also sent to Nick Pelling in case the cryptographers would like more information from it.


My own research into the manuscript concluded with a date for the final recension of the imagery, save a few peripheral additions and marginalia, around 1350 (it had been 1330 but the form of the ‘tower’ in the north roundel, which is not the North emblem, adds 2 decades to that date. The tower, which I identify as the Galata, gained its form in three stories in 1348/9 and ceased to have that quite that form after 1445. As it happens the latter date is useful marker too, as we’ll see. I accept, of course, the radiocarbon-14 range for our present manuscript.

Evidence from Hill.

It has become my habit, when quoting matter likely to upset adherents of a theoretical ‘Voynich’ narrative, to reproduce the passage rather than transcribing it.

So here’s Hill on the ‘4’ shape as numeral in England, Italy and Germany.

Although Hill tried, as he says elsewhere in that introduction, to ‘sweep into his net all examples to 1500’ he, like Friedman, was oriented to think of history as determined by official and ‘high’ works, rather than the works of artisans or commerce. Breviaries rather than invoices. So earlier examples may still turn up if anyone cares to investigate German documents.

Here’s Hill’s entry for that early 14thC manuscript from Florence.

*NOTE – Hill has a number of thirteenth-century examples which verge on becoming a similar form but which represent ‘five’)

Without pausing to explain why the following might excite old die-hards like me, here’s another set of numerals together with Hill’s note.

The more traditional form for the numeral ‘four’ resembles another Voynich glyph.

In the version above (from a digitised ‘Voynich alphabet’) the thickened stroke should be ignored. It doesn’t appear for the old ‘four’ as a rule, nor for the Vms glyph.

The range over which Hill collected his examples included copies of textbooks, religious handbooks such as breviaries, texts on mathematics including compotus, and coins, monumental work and church-bells. But no informal notebooks, no financial documents, nothing from the trades or commerce. Of artisans, only painters as ‘artists’ were included, and these chiefly from Italy. The nineteenth-century Anglo-German idea of ‘Europe’s intellectual history’ – yet again.

While it might well be said that cryptographers need skill in mathematics, I do wonder now whether the Urbino ledger of 1440 was rendering ‘4’ as a cipher, or simply translating a form for ‘quo’ unfamiliar at that time to Italians and in a time when the ‘qo’ might yet be mistaken for ’98’ or ’96’ or ’48’ or ’46” or even ’58’ or ’56’. This clip illustrates forms in a fifteenth-century English manuscript. Regular readers will have seen the illustration before, but it will do.

Untidy-ing the theory.

If we suppose any relationship exists between the ‘4’-shaped Voynich glyph and the similar form for the numeral ‘4’, whether direct, or employed as cipher, or anything else, then the seemingly obvious conclusion is that the text ought to have gained its present form in the fourteenth or earlier fifteenth century, up until 1438, in which case it is most unlikely to be of German Christian origin, only a little less unlikely to be of English origin and even if Italian not yet likely to be in any cipher-ledger known so far in Voynich studies. One must have some doubt about the relevance of the Urbino ledger, and it is too early for the Milanese ciphers noted by Pelling in 2006. Of course here too, it may simply be that no-one has yet investigated, or if investigated, not yet published other records.

We may say, at least, that the form is most likely to belong within the same general area in which we’ve noted instances of western MSS containing, as does the VMS, quires both septenion and quinion.*

*for details, see earlier post, ‘What magic? Where Magic? 4: Whose magic?’ (July 5th., 2021) 

We might describe that region in the most general terms as south-western Mediterranean, but with an accent on places linked directly to the maritime trading routes initially dominated by Genoa and Venice and their entanglements both by sea and within Italy.

In order of date, we have the ‘4’ shape in a Florentine MS of the early fourteenth century, the Venetian mercantile handbook MS Beinecke 327 in the mid-fourteenth century, and the Jewish cartographer of Minorca in 1375, then a similar form, certainly, in a cipher-ledger of Urbino in 1440 and somewhat later, in ciphers used in Milan, though the last post-date manufacture of our manuscript.

It is almost neat. Except.

Except for those glyphs with elongated ascenders.

In 2015 my aim was to only to track the history and context in which such forms occur in the west.

The first historical example came to notice in Voynich studies thanks to Jim Reeds who provided it and, at the same time, re-introduced* Cappelli’s Dictionary to Voynich studies, Yet someone had already described these as ‘gallows’ letters or glyphs, for it was as ‘gallows’ he described them.

*Aficionados may recall Erwin Panofsky’s acidic response to Question 13 of Friedman’s preposterous ‘Questionnaire”.

The following, in blue, from that post of 2015:

That first example (above) comes from an official charter, now in Parma but from the monastery of San Savino, in Piacenza. When this example was introduced to the Voynich world by Jim Reeds, via the old mailing list, he added this:

But I can refer to one of the few photographic facsimilies in … Cappelli’s Dizionario (the 1967 reprint of what appears to be the 1929 edition), namely “Tavola IV”, which shows a letter [of] 1172 [AD], 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.”

So it’s not just a letter, but a charter.  And the script is described as Caroline, despite being late 12thC.

Pelling, whom experience has shown invariably accurate in crediting original sources of information, even if they oppose his theories and opinions, mentioned that Barbara Barrett argued for the Voynich script’s being derived from a Caroline hand and not – as Pelling has long held – that the Voynich script was written by a scribe influenced by the humanist style.

I’ve not yet seen Barrett’s articles (I’m now trying again to see if the Fortean Times can supply a copy), but I note that neither Barrett, nor Pelling, considered any other form of script save a Latin script as informing the manuscript’s written text as we now have it.

I should like to see the opinion of someone expert in comparative palaeography, and one day perhaps we shall see a detailed palaeographic study of the Voynich script in print.

Such elongated ascenders don’t occur in the body of a humanist text. The time-frame is wrong. By the early fifteenth-century, only the faintest trace remains of the style informing that charter from Piacenza and even in that twelfth-century charter, such forms were used only for the headers and not within the body of a text.

The Caroline hand is certainly earlier – the standard limits for it that you’ll find in the text-books, wiki articles and so on is “8thC-11thC inclusive“.

At this point a theorist might be inclined to invent a theory-patch by saying something like, ‘OK, so a fifteenth-century Italian scribe copied/enciphered a tenth-century text’ – but such anodynes won’t do.

Our maxim as revisionists remains, “No evidence, no exposition – no case”

On the other hand that ‘patch’ might be expressed as a question, or rather a group of questions, and so begin a potentially useful line of enquiry.

As I hunted other examples of similarly formed letters, ignoring the distraction of an over-elaborate variation adopted by the chancery of the Holy Roman emperors, I found that they were all (that is, all that I could find, not all there may be), issued from the papal chancery or scribes of papal delegates and all but one – which I’ll mention later – were of similar type, not letters but charters.

To find another text from the Latin domains in which elongated ascenders of any kind occcur within the body of a text, I had to go to an ever-earlier period and it was thanks to Jonathan Barrett and his blog, ‘A Corner of Tenth-Century Europe: Early medievalist’s thoughts and ponderings” that I had the following – again a charter, and again from a Papal delegate. It’s the authority to establish a Spanish convent, in Ripol (mod. Ripoli) in the province of Girona (medieval Gerona). It is dated to the tenth century.

I’m fairly sure that here again a professional palaeographer might have done better, and quicker, but from the 1990s until very recently, we had none who were publicly engaging with Voynich research, so one did what one could.

Papal establishment charters continued evoking the style of elongated headers to as late as August 1412, but this late example shows only a faint echo of the twelfth-century one noticed by Jim Reeds. It is among a set of documents sent to Scotland after a local bishop had, somewhat presumptuously for the time, already granted a charter for the establishment of the University of St.Andrews. The papal documents therefore just confirmed that charter and added to it, including the official ‘Blessing’ in which the vestigal form appears. A somewhat blurry photo of it be seen here.

Against this, a Papal letter ( bull) of 1216, though addressing the rights to which the Order of St.John Hospitaller were entitled, contains no ascenders save the ordinary ‘f’ and old form of ‘s’.

So that’s the problem. What appears to be a ‘4’ form which we should date to not earlier than the beginning of the fourteenth century, combined with glyphs having elongated ascenders that appear within the body of the text, like something written in the tenth century.

An anomaly if we suppose the written text first composed by a Latin in western Europe. And what Latin, in tenth-century Europe, or even early fourteenth century Europe, could have produced the drawing on folio 13r?

Other restrictions.

I’m hoping now to get access to material in the Fondo Datini to see how commercial documents were being written in fourteenth century France and in his hometown of Prato, as well as hoping there might be some way to see examples of commercial and financial ledgers from the papal court in Avignon, whether under the accepted Popes or their successors in the region after 1376, the so-called ‘antiPopes’. I’d also be interested in seeing any form of ’emissary letter’ which might exist from before 1450.

While I’m strongly inclined to agree with Panosky’s initial assessment of the manuscript, when he classed it as a product of ‘Spain or somewhere southern’,* it is with the caveat that I do not think the whole content created by any single person, nor originating from that region. I consider ‘Spain or somewhere southern’ an important. late, halt in the route of transmission, though distance and through time.

*by which, as I read his comments, he meant the appearance of the contents.  Whether or not he was considering the vellum’s relatively rough finish (as another early commentator remarked) and its binding, I think at present the weight of evidence for the current work’s manufacture lies with Italy, or a region with strong influence from one or more of the Italian city states.  Again, the judgement of professionals in comparative codicology, and those with years of practical experience analysing the structure and materials of medieval manuscripts, are best able to judge that.  – Note added 29th Nov. 2021.

It may be written text in Beinecke MS 408 was composed between 1350 and when our current manuscript was put together, but I cannot think so while this apparent anomaly remains unexplained.

Feel free to leave a comment here, or a link to your own work if the problem interests you.

(detail) Yale, Beinecke MS 408 f.4v

3 thoughts on “Anomalies – consider this…

  1. Among more recent studies of the Arabic numerals are two papers by Charles Burnett,

    “The Coherence of the Arabic-Latin Translation Program in Toledo in the Twelfth Century,” Science in Context 14 (2001): 249–288;
    “Indian Numerals in the Mediterranean Basin in the Twelfth Century, with Special Reference to the ‘Eastern Forms,’” in From China to Paris: 2000 Years Transmission of Mathematical Ideas, ed
    Yvonne Dold-Samplonius et al. (Stuttgart, 2002), pp. 237–288.


  2. I see today that a wiki article describes that Florentine copy of the Liber Abacci as ” Ms. Biblioteca Nazionale di Firenze, Codice magliabechiano Conv. Soppr. C 1, 2616″. I believe the wiki author is mistaken in dating that mss to 1228, but quite correct in noting that Leonard’s travelling around the south-western Mediterranean had first made him aware of the Hindu-Arabic numerals, of which there were two main types. This has long been known to western scholarship with (for example) Solomon Gandz writing, in 1931,

    “The Arabs … distinguished two different types of numerals and characterized them by two names, the Hindu and the ghubār numerals. The Hindu numerals were common among the Eastern Arabs and are, at present, still usual in the Arabic world. The ghubār numerals were found in Spain among the Western Arabs..”

    Solomon Gandz, The Origin of the Ghubār Numerals, or the Arabian Abacus and the Articuli’, Isis , Nov., 1931, Vol. 16, No. 2 (Nov., 1931), pp. 393-424.

    Among current studies of numerals, their form(s) and history, Charles Burnett’s are to be recommended.
    One reason for my regularly beginning at the beginning, as it were, is to be clear about what might reasonably have been known to someone living in the days of Wilfrid Voynich, Erwin Panofsky, or the Friedmans. So, for example, both Wilfrid and Professor Newbold had died before Gandz’ article appeared in Isis.


Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.