c2000 words

The author’s rights are asserted.

It is now more than a decade since I pointed out, for persons then involved in Voynich studies, that while the calendar diagrams’ central emblems use a visual language near-enough to Latin conventions, the diagrams themselves do not.

Given the enormous optimism, self-confidence and positivism one finds in Voynich writers working outside their areas of special competence – and which is surely needed to address so problematic a manuscript in the absence of prior studies – I expect my opinion will be unwelcome that any correct reading of these diagrams (if not of their written labels) will need specialist knowledge at a level we associate with such names as David A. King, Elly Dekker, and the late David Pingree and Paul Kunitzsch – Kunitzsch’s death in 2020 ending one of my own long-held hopes for this study.

The diagrams raise a number of highly technical issues which only a deep grounding in the history of medieval astronomical texts, tables and charts can clarify. Many of those issues will be invisible to a general reader and amateur theorist, especially any misled into thinking that all one needs are “two eyes and commonsense” and some computing skills.

I had hoped to avoid pouring such cold water on enthusiasts who enjoy guessing or who have confused traditionalists’ repetition of old theories with statements of fact.

I include this post so that my silence may not mislead readers of this blog into thinking that I believe the Calendar section expresses nothing but the habits of western Christian (‘Latin’) Europe.

Whether we consider the ninth century, the twelfth century, the mid-fourteenth century, or the early fifteenth century, astronomical knowledge involved wider and more complex interactions than the usual historical summaries suggest.

It is more than a decade since I realised that there is an inherent conflict between the iconographic information provided by the Calendar’s central emblems as against the diagrams as such.

Take, for example, the long-enduring assumption that each of the calendar’s anthropoform figures represents a day (or night), or that each star in each diagram does so. The stars, and the nymphs have been counted by various writers over the years – in publications, as in conversations to mailing lists and forums. Most recently, I understand from one amiable forum member, Anton Alipov has counted them again and shared his results at voynich.ninja.

The rhyme everyone knows today was known in medieval Europe by the ninth century. In modern English it runs,

Thirty days hath September, April, June and November.

All the rest have thirty-one, excepting Februrary alone

which has 28 days clear, and twenty nine in each leap year

Even if we were to treat the doubled months as split months and count their stars together, still the tally must read (according to the copy I’ve been sent)

• March: 29 or 31(?)
• April: 30 [As 15+15]
• May: 30 [as 15+15]
• June: 30
• July: 30
• August: 30
• September: 30 [and one extraneous star]
• October: 30
• November: 30 [and one extraneous star]
• December: 30

The logical question to ask (one would think) is where and when we find calendars of comparable design, ones lacking any evidence of intercalation?

That has never been the response made in the past.

Those unable to contemplate the possibility of non-Latin character for the manuscript’s contents (or who can imagine it, but find the idea preposterous) have veered off and created alternatives – often by inventing imaginative-hypothetical theory-patches mis-represented as the fruit of historical logic. The basic traditionalist position is that if the manuscript’s content doesn’t look Latin, or act Latin, then it jolly well ought to, and really does “underneath it all” and/or that the author/draughtsman got it wrong, poor thing. 🙂

It must be understood that the “all-Latin-Christian-European” theory-narrative IS the traditionalist theory because the study’s founding fathers – Wilfrid Voynich, William Newbold and William Friedman – began by assuming it an autograph composed all at once by a thirteenth century Englishman, or by some other European male important enough to figure in Europe’s story of its own intellectual advance to the mid-twentieth century.

Especially for the Friedmans (and thus for Mary d’Imperio) even to suggest the content included “foreign” matter was offensive, because to them the foreign implied the inferior and unimportant.

Added to this was the theory that the written text should prove to be a consistently-spelled and neatly grammatical plain-text because without such standardisation (as they thought) encryption and decryption became impossible. That it was an encrypted text of ordinary prose or poetry was the cornerstone – the non-negotiable element – in the theories they created.

For the time of Roger Bacon, Scot et.al., that meant in practice assuming the text written in one of the liturgical languages and given their bias – it meant Latin, English or German, none of which is indicated by the usual statistical analyses. The same assumptions and prejudices so common in the nineteenth and earlier twentieth centuries are why Panofsky’s recognising non-Latin elements – presumably in the manuscript’s layout and drawings – was not taken seriously by the Friedmans and by Mary d’Imperio was imagined mediated by some Latin figure. Hence the references to Ramon Llull and anachronistic allusions to a consciously Christianised Cabala.

As so often, Voynich theorists have attempted to assert a section’s meaning, or a drawing’s meaning, though paying scant attention to the form given an image or section – as we’ve noted recently in discussing the series of emblems used as centres for the Calendar diagrams.

Inherited bias, within the traditionalist theories, seem to me to explain why a hundred years and more have passed without any Voynich writer asking, and seeking to understand even the simplest of questions about this section: such as “Why do the central emblems not form a zodiac sequence, even of just these 10 months?” Or “What kind of calendar might have 30-day months for every month from April to December, inclusive?”

The larger questions about calendars and the history of astronomical works are not within the brief of an iconographic analyst; what we can address is the curious choice of emblems to fill these diagrams and why they present such an odd mixture of zodiac-like and non-zodiac like forms.

I would add another question – why do they include forms which appear in some cases compatible with images found in England and in France over the period from the eighth to the twelfth centuries, but with other emblems unattested in Latin works until the mid-fourteenth century?

I’d point here not only to November’s crocodile but to the history of the Arcitenens type. The Parthian type with its goat-legs appears early, in the work of one Anglo-Saxon monk who also worked in France, and as a fully human figure in the 9thC, but it was not the form preferred thereafter in Latin manuscripts’ representation of the 12 zodiac figures and seems to disappear soon after from the Latin sources.

Nonetheless in its old Pan-like form it reappears in one Jewish manuscript* that the holding library dates to the 15th-16thC, and whose chief text is the Six Wings of Immanuel Bonfils of Tarascon (1300 – 1377). And the same manuscript has a prawn-nosed lobster for Cancer. I cite the example only to show that history – including the history of images and forms – is no simple “forward-march”.

*On this see first: Gerrit Bos, Charles Burnett and Tzvi Langermann, ‘Hebrew Medical Astrology: David Ben Yom Tov, Kelal Qaṭan: Original Hebrew Text, Medieval Latin Translation, Modern English Translation’, Transactions of the American Philosophical Society , New Series, Vol. 95, No. 5 (2005), pp. i, iii, v-vii, ix, 1-61, 63-121.

Alfonsine Tables.

The part played by Jews, including Jews from French-speaking regions, in the translations made for Alfonso X of Castile is another subject unsuited to amateurs and speculators, for it is still debated by scholars who may fairly be described as eminent specialists in that field. When such scholars as Pingree and Mercier are unable to agree about transmission of the Persian Syntaxis or Byzantine reception of the updated version of Ptolemy’s Handy Tables, the issues can hardly be resolved by less well-informed writers – yet such matters must impact on how one explains the extraordinary number of stars which seem to be referenced by the Voynich calendar.

Alipov’s recent count gave a total of 299 figures of the star-holding type (he calls them ‘nymphs’) and a total of 297 stars – presumably including some he describes as extraneous (ornamental?).

I call the number extraordinary because any survey of astrolabes and other flat representations of the heavens, produced before 1440 (and more realistically to 1350 AD), generally include between 17 and 30 stars, with 50 being an unusually large number. Similarly, one does not find in the Latins’ calendars, breviaries or books of hours from which so many Voynich writers have taken their zodiac images such things as star-tables or lists, nor do their months consist of mostly of 30-days.

From time to time, since Jim Reeds’ mailing list was opened, individual researchers have tried to raise the matter of the lunar calendar and the lunar asterisms known as lunar mansions or as towers – only to have the topic submerged, ignored or bulldozed under some determinedly Eurocentric theorising – typically focussing on the Picatrix in pretty much the same way that “southern and Jewish” has been transmogrified by theoretical narratives about Ramon Llull and Christianised Cabala.

Illustrations in copies of the Aratea may add red dots to mark stars, and Elly Dekker, in 2010, published a paper on the Leiden Aratea* which shows it referencing more than 600 stars by the red dots with which its pictures of the constellations are adorned. How much work was required to identify those stars, her paper shows plainly enough an although I include here one table [Table 3] from that paper, I cannot emphasise strongly enough that no use should be made of it to invent or patch a theory – at the very least the trouble should be taken to read the paper in full and realise just how much expertise is required even to identify stars embedded in an illustration of a constellation.

Leiden Aratea (Leiden, Universiteitsbibliotheek MS Voss. lat. 4° 79) – produced in the first half of the ninth century for the court of Louis the Pious (814-40). It is not a typical work of that time, but an exceptional one – in its size, artistic quality and content. It contains images of forty-two constellations as we count them now, and the Pleiades.

*Elly Dekker, ‘The Provenance of the Stars in the Leiden “Aratea” Picture Book’, Journal of the Warburg and Courtauld Institutes , Vol. 73 (2010), pp. 1-37. Accessible through JSTOR.

While respecting the level of scholarship needed to attempt an accurate reading of the Calendar diagrams, we may continue to investigate the central emblems which – I’ll say again – do not appear to me to agree well with the character and content of the diagrams proper .

Two more passages worth thinking over before we turn to those manuscripts I’ve been promising (one very early semi-Christian calendar, and Oxford Bodleian MS Douce 313).

Both these passages (below) come from papers by Raymond Mercier, a former editor of the Journal for the History of Astronomy.

The first relates to the ninth-twelfth centuries; the second to the early fifteenth century.

In the first Mercier notes a curious instance of alteration/adjustment in a twelfth-century Latin text, apparently attempting to harmonise the Christian with the Jewish calendrical system, while at the same time back-dating it to the ninth century. The author of the tables he mentions – Luhot ha-Nasi – was Abraham ben Hiyya, known as Sarasvorda , who was born in Barcelona c.1070, and who died in Narbonne or in Provence in 1136 or 1145AD.

In this second passage from one of Mercier’s papers, he is speaking of events which occurred close to when the Voynich quires were made (1405-1438).

I would add, as a simple matter of fact, that the Persian New Year began in March, and we learn from Ibn Majid, a fifteenth century navigator who knew his stars, that the eastern mariners counted their sailing year from the date of the Persian New Year. It was important to count one’s days on those eastern maritime routes, because if wrongly calculated, the monsoon winds on which their navigation relied might be misjudged with disastrous consequences, physical and economic.

additional note (13th October 2022) on the moveable date of that Persian New Year relative to the Julian calendar, and the Arab navigators’ practice of counting their days pp. 361-2 in G.R. Tibbett’s English translation of the ‘Kitāb al-fawāʼid fī uṣūl ʻilm al-baḥr wa-al-qawāʻid’ of Ahmad b. Majid al-Najdi, the translation published as Arab Navigation Before the Coming of the Portuguese…etc. Any reader who is particularly keen to have the information but not quite so keen on the book’s price is welcome to email me and I’ll type those two pages.