This post treats issues of method and the fraught question of when theory-formation is a benefit to the study and when a hindrance. I hope revisionists will find it food for thought, but anyone with an investment in some Voynich theory, and especially a theory focused on Latin European personalities, might like to stop reading now. Besides, it is a long essay, not much enlivened with pictures.
I would actually prefer not to to treat this topic at all. Theorists’ responses are easily predicted. But it must be done – Fiat jūstitia …. as the Roman said about Libra.
“Divided minds” – logic and illogic in Voynich research.
A belief that ‘images are easy’ is arguably the first and longest enduring systemic error in Voynich studies, but may explain why so little effort has been made to study techniques of analytical method (though Voynich sites may entitle their non-analytical matter ‘analysis of the imagery’). From 1912 to the present, the study’s history shows little sign of efforts made to understand how images are assigned to their place, time and social community. Theory-driven ‘nearest fit’ has been constantly imagined sufficient whether the Voynich theory being posited had the manuscript a product of medieval or of modern times, and attributed it to some part of western Europe, to the Americas or elsewhere..
Such magnificent indifference to objective criteria is not so prevalent in other facets of the study.
A competent cryptologist, when he or she crafts a theory about Voynichese, remains conscious of the theoretical model’s being no more than an analogy, and takes notice of both what does and what doesn’t accord with the primary evidence. If the theoretical model proves a poor fit, it is discarded, although the aim is to devise one so close to the original that it will help explain what has been so far unexplained.
In the same way, a botanist might posit a theory about a plant’s identification, and test it by balancing points of similarity against points of difference between a textual description and the living plant. A linguist will also balance points for, and against, a theoretical model.
There is no confusion in their minds between the actual object which constitutes the standard, and the theoretical model which may achieve or fail to agree with that standard.
Quite the opposite habit pervades assertions made about this manuscript’s images in almost everything written since 1921. In that case, whenever the hypothetical model fails, the usual practice has been to ignore the differences, deem the primary evidence flawed in failing to agree with the theoretical model and thus, in effect, the analogy is taken as the standard and the primary evidence discarded. Whatever details are deemed a near-fit (right or not) are given an importance out of all proportion, while the details which don’t are treated as of no consequence. At best they are merely ignored; at worst attributed to an equally hypothetical creation described as ‘author’ or ‘artist’ and blamed for the work’s not conforming to the analogy.
What is most curious is that, in Voynich studies, a single researcher may switch from the rational and analytical mode to the emotional and inverted mode, depending on whether they are thinking about the manuscript’s written or pictorial evidence.
The switch is signalled by the degree of interest expressed in the reasons and evidence informing dissent from the hypothetical model: in other words, in why the dissenter dissents. As a linguist or cryptographer, that researcher is likely to be interested in comments about flaws in a given model. In emotional mode – and historical scenarios seem to have a strong emotional factor – the response to similar comment is more likely to be little other than expressions of personal hostility. The level of that researcher’s interest in the primary evidence’s divergence from that theory is also reduced to the minimum.
And, in fact, theoretical-historical models appear to render differences invisible to the theorist. You can explain those differences, describe them, illustrate them and provide a yard of documentary evidence – but the theorist may well see nothing and read nothing and absorb nothing which he or she interprets as a threat to their theory. Over-identification with a theory is the point at which Beinecke MS 408 becomes the ostensible but not the real subject of a person’s interest.
It is theory-induced blindness which produces illustrations of a myriad zodiacs and the assertion that the ‘nearest fit’ is the one which suits the theory. Or which reduces the question to a single emblem only (as the German theory has done with the ‘archer’ figure). The person is no longer thinking, ‘What were these images meant to mean?’ but something more like, ‘Since I know my theory is right, what about this part of the manuscript can be fitted to it?’ What is absent from such exercises, in connection with the month folios specifically, is any effort to explain those drawings as whole images, or to understand why its series of central emblems doesn’t, in fact, form a zodiac sequence at all – not even a truncated zodiac sequence.
After a century in which Newbold’s impression has been echoed and reasserted without variation it is difficult for modern reader to perceive the series as other than a Roman zodiac, or to realise how much virtual violence must be done to maintain that theory. Images which are there have to be imagined not there. Pages which are not there have to be imagined as being there, with non-existent pages imagined present, and their surfaces covered with hypothetical/imagined content.
One has also to conjure up a single ‘artist’ – when the evidence of several is plain enough – and then accuse that imagined figure or even all of them of a staggering incompetence and ignorance while at the same time (to maintain such theories as the ‘German’ theory) of such superb competence that they could draw a crossbow to scale within the space of one centimetre square. It has to be supposed that not only the imagined ‘artist’had managed to remain unaware that a Roman zodiac has 12 figures with none repeated and all in set order, but everyone else connected with the manuscript’s production had also managed to remain ignorant of a series which was to be found, in the Latin environment, carved on the exterior of churches, made in mosaic in public places, and used to illustrate manuscripts and calendars both liturgical and secular. Not only the artist(s) as I say, but the scribes and the overseer of work. And then one must imagine, further, that this ignorance survived in all of them while one is asked, simultaneously, to suppose that the person(s) for whom the month-folios were being made was an astrologer of some sort.
It defies reason and the historical evidence. But apparently did not quite beggar belief.
I’ll turn again in the next post to the matter of variant depictions for the zodiac in works produced in Latin Europe the immediate point being that, once again, the focus of attention slid from the primary evidence to a theoretical model for which one’s credence is demanded (with penalties ad hominem for refusal) and this despite any formal argument’s being presented, or any effort made to explain what is actually on the manuscript’s page.
The primary evidence’s failing to concur with the hypothetical model is treated in the same way. One is encouraged to ‘just ignore the nonsense’ or to blame the source itself. It *ought* to conform to the theory.
Is it any wonder that a century’s elaboration of that ‘Latin origin’ theory has not elucidated a single phrase of the written text?
The non-zodiac shall be deemed “a zodiac”; the purpose for which the month-folios were made shall be deemed astrological. The ‘logic’ invoked to persuade one to accept that what is not so shall be so is not (as often asserted) any historical logic but the sort of internal logic we find in the best historical fiction.
To suppose that scribes so obviously competent as those who made the written part of the text (at least) wouldn’t know the series of 12 constellations in order, and that the labours of the months linked just one to each of those 12 months is to defy the historical evidence. The logical conclusion is, surely, that the series in the month folios diverges from the zodiac’s standard sequence and order of ’12’ (or, if you like, of 10) for a reason.
Discovering that reason must be part of researching the manuscript if the aim is to understand the primary document and that certainly can’t be done by pretending the primary source is other than it is.
Which is why, in my opinion, creation of theory-driven historical scenarios which presume what is not known is known is an inappropriate method, no matter how traditional in this study.
It leads to that unreasonable confidence which has one theory claim some creature, or plant- picture shows a New world species while another says the month-folios must speak of Christianised astrology and magic, or which – finding itself stymied by the plant-pictures – resorts to airy declarations that whatever it has not provided with a theoretical ‘nearest fit’ is to be dismissed as ‘the artists’ fantasy or personal whim. If such guesswork was presented by one person claiming responsibility for it, the matter might be debated rationally, but such things are often decided as if by some anonymous bureaucracy or by public acclaim, disseminated by a general weed-seeding, as produced out of analogy by god-knows-who, and then as something ‘everyone agrees to’ defended by the masses to the hilt – belief defining dissent as heresy.
So the normal relationship between primary evidence and that posited analogy is inverted, the story elaborated and more impositions laid on the primary source, and while individuals are eager to accept credit for a ‘genius idea’ it is rarely that the same individuals produce any formal argument for which they accept all responsibility. By ‘formal argument’ I mean one which balances arguments for and against a proposition, adduces verifiable evidence and accurately documents both sources and any precedent.
Traditional method, in Voynich theory-making, is fundamentally just poor method. The way the images are treated in service to such theories is not remotely like the way pictures are normally approached, described and assigned their time and place of first enunciation or of subsequent copying.
And I suppose that to show the foregoing remarks are not themselves just theory, I must now add an example, but since examples are often confused with personal attacks on whoever’s work the example comes from – even if the person is dead – the safest example is one I’ve already spoken about at voynichimagery.
Below is an illustration which Ellie Velinska produced for a post to her blog in 2014. It sets the diagram from folio 68v next to a detail from one copy of Oresme’s Treatise on the Sphere and was very warmly received, as you’ll see from the comments made to that blogpost. ( here). I’ll leave my own comments to the end of this post, but as you’ll see if you follow that link, none of Velinska’s commentary addresses points of difference between the ‘clips’. She offers no analytical discussion of the Voynich drawing, nor tries to explain its intended purpose or its particular form.
Since I can’t treat every historical-theoretical narrative proposed since 1912, I’ll keep to the oldest – that which interprets the manuscript, and specifically its pictures, by analogy with western Christian (‘Latin’) culture during the thirteenth- or fourteenth- or early fifteenth-century.
Here, in brief, is the negative case:-
Hallmarks of medieval Latin (i.e. western Christian) imagery.
As did every other cultural community, that of medieval Latin Europe expressed its own world-view using a distinctive repertoire of graphic and painterly techniques – the ‘language’ of art.
The conversations between maker and intended audience speak of that world-view they shared, doing so by both the style and the content of an image. It is by recognising both form and informing thought that an image may be described as an expression of medieval Latin culture and assigned its origin in some particular region.
That world view characteristic of medieval Latins was informed by an idea of universal hierarchy, this vision including everything from heaven, through earth to hell – all of which were equally ‘real’ for them. Their fixation on relative position in that universal hierarchy meant that every visual conversation emphasises the ranking accorded each element in a picture, whether animal or person, cloud or fish, angel, devil, noble or peasant. Except when used symbolically – as the lily might be used as symbol for Mary, Christ’s mother – all natural things of earth were assumed subservient to mankind, and within mankind the western Christian was assumed ‘properly’ superior to all others. (This heritage and pre-disposition is why European society was initially outraged by Darwin’s theory of natural selection, and also why after a period of adjustment, Europe so easily translated it into what is a ‘theory’ in the sense of a fiction: social Dawrinism).
But the medieval Latins’ attitude is expressed in imagery by showing man surveying as possessions the land and all it contains. It also elevates the deeds of the male over those of the female; it sets nobles on thrones, on chairs, in high-towered castles and on horseback – thus literally as well as conceptually over those deemed ‘lower orders’. War is another constant: the struggle between the upper righteousness and lower sinfulness; between Christianity and all other belief systems; between ruler by inheritance and the elected head of the church, between angels and devils. Literally depicted forms, and allusion to beauty as expression of ‘higher’ rank, and goodness were among the techniques by which position in the universal hierarchy was envisaged and, so, communicated.
Images expressing this world-view occur even in Latin herbals, in their introductory images (as in the Manfredus herbal and the Anicia Juliana), or in images scattered through it, and in such things as dedicatory inscriptions and colophon.
From the Voynich manuscript, those factors and themes and above all that perception of the world itself are overwhelmingly absent. It is the most resounding silence and failure to appreciate its importance has been the single greatest failure of the many historical-theoretical models devised to explain the manuscript.
There is not one depiction of a king, of a throne, of a man on horseback, of a figure recognisably from the calendar of saints. There are no halos, none but late-added crowns; no bishops (though one preacher in a Mongol robe appears in one detail, an addition to the older material which I date to the late thirteenth or early fourteenth century, and one reason I date the material’s introduction to the Latins to that period.). no massed armies, no servants in peasant costume, no image of a seated authority figure before a kneeling inferior. There is no allusion to Christian belief in the images, none of those emblems by which a saint was identified. The only sign of war is a single small detail among those on folio 80v. It shows a male figure dressed in what appears to be Roman military dress, and in the act of enslaving a female). There are no lovely noblewomen, no devils, no winged figures at all. There is no reference to class distinction by the usual depiction of silk and ermine robes. The western preoccupation with fabrics is found only in the late-added heavy pigment laid over some figures in the month-folios – another indication of late translation into Latin domains. The arms and fingers of the great many female figures are unadorned with jewellery. There are no interior scenes, no nuns or clerics, no travellers with staff and scrip, no vessels with handles and bellies; no emphasis on objects as tokens of status; no images of the hunt, no ‘horse, hound and falcon’ images and only one figure in the entire manuscript who is shod.
There are no vine borders, no interlace and knotwork designs, no drolleries formed by the fusion of human and beast into one creature – in fact no evidence in any of it of an inclination to indulge in fantasy. This renders unlikely too, the sort of excuses being invented at present to cover the fact that the old theory of the plant-pictures as a form of Latin herbal is bankrupt – something which Tiltman understood by the 1960s. The latest rationalisation asserts whatever plant-pictures frustrate even the ‘near-fit’ approach are merely the product of fantasy or whimsy. For this new theoretical elaboration I find no evidence within the manuscript at all. Symbolic, allusive and mnemonic devices certainly, but none without relevance and none personal whimsy. They are not beyond understanding. From what little is said in public, the ‘whimy’ idea seems to be another effort to find post-hoc justification for something in d’Imperio’s book, and to rely on arbitrarily transfering ideas which Marco Ponzi offered about Cambridge Bodleian Trinity MS O.2.48 to Yale, Beinecke MS 408. One would like to see that provided a formal argument.
Nor is there anything of officialdom in the Voynich images; no official figures’ granting gifts or meting out judgement. There is nothing of rule and government whether of religious or secular organisation. Not so much as a male holding a walking stick, a staff of office, a crozier, or a sceptre. (I’ve listed those in Latin order 🙂 )
…in sum: the entire fabric of the Latin Christian world, its culture, informing ideologies and world-view – the very means by which an image is assigned Latin origins – are just not there.
THAT is why specialists in so many areas of medieval western culture have refused to endorse theoretical arguments, and denied overtly or tacitly that the manuscript is “one of theirs”.
It is remarkable, even astounding, that the logical inference has so rarely been taken: that the reason imagery in the Voynich manuscript doesn’t look like an expression of Latin culture might be… because it isn’t.
The possibility receives further support from what I describe as ‘avoidances’ in the imagery.
Apart from later accretions as e.g. the month-names; additional images set onto the back of the Voynich map and a few other specifics, these avoidances are so pervasive across the various sections that I take them as indicating a cultural norm, and one which was certainly not Latin and which represents an important phase in the content’s evolution and transmission: a chronological stratum. The following details reproduce notes from my research log for July 13th., 2010, with some of the marginal notes subsequently added as I began looking into the questions raised. The notes were for personal use, as brief guide for research during the months to follow, and I daresay some will read a bit cryptic. But anyway here they are, verbatim. The first notes are in italics. Marginal notes in plain. Today’s comments in blue.
- No use of instruments – neither ruler nor compasses. [exception: folio 57v.] Pages not ruled out. No evident mark of wire, nor of pricking for this purpose- has textblock been trimmed for a later re-binding [later marginal note] – Rene Zandbergen says it hasn’t been and that he hopes I too may one day hold the manuscript in my hands, as he has done). Folio 57v as late addition – possibly very late Cf drawing-style in illustration for Kircher’s “Machauter” and note his source – but perhaps the ink is against a date in the 17thC. Other, less obvious, exceptions – some plainly informed by Latins’ traditions – seen in drawings accommodated by using the map’s reverse. I’d date these to the fourteenth century. [Further marginal note] In a post to second mailing list, in 2007, Rich Santacoloma notes that 57v shows the prick of a compass or dividers, and later that it has three distinct centres]. I would finally write extracts from this part of the research – that is, about f.57v -at voynichimagery over February-March, 2013.]
2. Not only no ruled lines but no perfectly straight drawn lines. This also reason for no ‘ruling out'(??) Rapidity with which one scribe comes and goes from the ‘bathy-‘ section, after using some as ‘improvement’ on the original. Effort to copy the original material so exactly… was the 15thC copying informed by any knowledge of the tabus, or not? On this last point I think the balance of evidence is against the copyists understanding the earlier avoidances – a better definition than ‘tabus’.]
3. Avoidance of crossed lines. No interlace, no ‘x’-form among the glyphs. Discounts the Insular, Coptic, Latin, mainstream Arabic, Armenian as well as the Byzantine traditions (except in some superficial ornament, which Pelling calls ‘cross-hatching’ and supposes invented in Renaissance Italy. Re architectural structures added to the map – c.13th(?) century – check comparisons.
4. No ‘boxes’ – including no triangles. Nothing with sharp right-angles except a few late errors in copying and an (original) emblem for ‘south’ on the map – though even that is surrounded by an apotropaic ring – “shield against the fires implied”. North-oriented worship? cf. Harran. (Tamara Green). Containers in the root-and-leaf section, even simple cylinders, are bent to avoid the angular ‘box’. Very unusual avoidance. Perhaps related to observation that nothing natural to the world is ‘ruled line straight’? Arcs of horizon and heavens are conceptual, not physical. Can’t identify the community. (must check ethnological studies – ugh!)
5. No literal depiction of any living creature. [perhaps one reason for the plant-pictures’ not showing specimens as we think characteristic of the Mediterranean world. But it may have been just convenient to group by location and use]. ‘Violas’ image an obvious ring-in – its maker clearly understood the principle but it wasn’t natural to him. He had no idea how to form a root-mnemonic or use the ‘shorthand’ motifs. And he defines a plant by its flower(!). His composite image is drawn as range of viola species occurring rom east to west (or vice versa? A Latin, perhaps – his mind works differently from the original makers’. The system (and key to text?) had been explained to him; but it isn’t his natural way, or training – doesn’t quite “get it”. Get opinion on petals. micrography? [ I later had the advice of an eminent specialist in the history of Jewish paleography about this posited micrography in f.9v but the reaction of the ‘Voynich community’, at that time, was strongly opposed to any suggestion of eastern or Jewish ‘authorship’ – so much so that I did not feel it right to name the specialist. Some years later, Zandbergen and Prinke wouldproduce a book in collaboration with a writer named Stephen Skinner, who described a ‘Jewish’ theory so appallingly ill-informed as to be offensive. Since I’d been explaining for some years by that time, details indicative of Jewish influence, the specialists who’d been advising me were disconcerted and to ensure no connection was imagined between that essay and the work I’d published between 2009-2017 it added another item to my reasons for deciding to close voynichimagery soon after.]
6. No repetition [i.e. replication, in the images of the living creatures]. This is a remarkable achievement if you think about it, and realise how very many female figures had to be provided with distinctly different form-and- face without resembling a living, breathing, human. Even in our present volume – which is reasonably supposed made by Latins who would have a very opposite inclination, we find only the occasional ‘slip’ by which a face looks ‘real’. [My favourite ‘shapely lady’ is one of those slips and it is a most valuable evidence that the draughtsman could have made them all just as ‘real’ if he’d been free to do so – just as he might have fallen into an easy repetition for the figures around the month-folios’ tiers. [Only some strong and probably religious principle would have prevented the earlier makers’ avoiding both literalism and replication. The fifteenth-century copyist who ‘slipped’ in making one very ‘shapely lady’ was certainly not working in monastic scriptorium. Nor, I should think in a fifteenth century European Jewish community. I never found but two references to a prohibition against such repetition [‘replication’] -one in connection with use of draw-loom fabrics in post-iconoclastic Byzantium, and another as a suggested reason for the fact that although knowledge of printing is attested within the Arab speaking world as early as the 10thC – it was almost immediately rejected and texts continued to be produced by hand for centuries more. I’ll have reason to mention this another time and will add the references there. It implies an aversion to magical practice, by the way.]
I do not pretend to have found answers to all my own research questions, but enough avenues opened to allow a reasonable explanation for these non-Latin characteristics.
Example – Ellie Velinska and Oresme.
To my knowledge, none save the present author registered any hesitation about embracing Ellie’s proposed match.
Ellie was not trying to say that the two images came from the same artist, or even the same atelier. Her argument – more implied than argued – was that similarities (only) between the two support an hypothetical Voynich history into which may be drawn the person of Nicolas Oresme, as well as the French royal court.
Ellie’s impression may one day be proven accurate in general, but to argue the ‘Oresme and Charles’ case, one would have to show that the manuscript’s vellum, its finish, dimensions and binding, its style of binding and much more are attested for that time and region or – if that is impossible – that the same detail occurs in earlier and in less formal, copies of Oresme’s Treatise.
Her case is weakened by the manuscript’s unobserved and unmentioned differences from the ‘match’ but even more by the fact that we do not find that form in other and earlier copies of that Treatise. Given the style and technique displayed by her chosen copy ( BNF fr.565) it would not be unreasonable to suggest that the detail in question may have less to do with the treatise itself than to the stock repertoire of some particular atelier. Below is the first image from a copy made during the first years of the fifteenth century ( BNF fr.1350)..
As is usual with implied rather than stated inferences associated with ‘matched clips’, their appropriateness or otherwise is often shown by re-contextualising the detail adduced as ‘nearest-fit’. In this case, it becomes apparent that the chosen ‘match’ is from an mage entirely characteristic of Latin thought. The chair tells us who is of the highest rank; the crown and sceptre denote royalty. The various types of hat refer to national and ethnic character. These are the ‘astrologers and diviners’ whom Oresme wishes the king to ‘put behind him’, echoing Christ’s words in the gospel. There’s no doubt it’s a product of Latin culture as well as of Latin making. Now consider the differences in attitudes, and forms for the human figures compared with what we see in the Voynich manuscript. The one aims at ‘realism’ as plainly as the other does not. Here there are no ‘broken’ arms and shoulders or deformed faces. Certainly no avoidance of replication… and just look at those robes and fabrics.
On the positive side…
There is certainly other evidence which permits more general argument for the manuscript’s content having been, at some time, in French-owned territories.*
*or rather, of French cultural influence. (note added 16th.March)
We have the orthography of the month-names, which agrees closely with forms found in Judeo-Catalan, Occitan and Norman French. According to Sixto – and I haven’t checked this – there were Catalan Jews in north-western France. The same orthography occurs on an astronomical instrument, made in Picardy, which was sent to England at some time. (I haven’t checked the object’s history yet). Koen Gheuens’ discussion of the ‘double lobster’ indicates dissemination of that form through Alsace and/or Flanders during the earlier decades of the fifteenth century and from an Anglo-Norman speaking source. My own research into the standing, fully human archer as token for the constellation Sagittarius also led me to conclude that its first remaining expression – in a mosaic near Lake Tiberius – occurs several centuries earlier than its first extant example in Europe and the means by which it came was probably workers in glass, who brought with them the red tesserae found in quantity in that region and whose technical secret, according to a contemporary Latin’s account, was by then known only to one or two families or clans of the eastern Mediterranean shore. The oldest western example is in a rich window originally in Braisne Abbey. Since that time, my illustration has been widely re-used by Voynich writers but minus reference to the associated evidence or argument as to what significance should be taken from the sudden appearance of this type, previously unattested in Latin art.
But altogether there is evidence enough to argue some link between the month-folios and France, but nothing like enough evidence to support a theory that the content originated there or that it has any connection to Oresme’s Treatise.
And that is without beginning to address the differences between her ‘matched’ clips or to explain the intention of the original image from folio 68v. Which last, surely, should have been attended to first.
On the history of Oresme and his works, and his detestation of astrology and divination see the short essay
- Mackley, J. S., ‘Nicole Oresme’s treatises on cosmography and divination: a discussion of the Treatise of the Sphere’. Paper presented to: Starcraft: Watching the Heavens in the Early Middle Ages, University College London, 30 June – 1 July 2012. (available online as a pdf).
Nick Pelling recently revisited the ‘Oresme’ theme
- Nicholas Pelling, ‘Nicole Oresme’s “Treatise on the Sphere” revisited’, ciphermysteries (Feb. 15th., 2020)