Note on palaeography and codicology – and my position on the written text.

Postscript – stop press.  For yesterday’s ciphermysteries post (read by the present writer today, March 1st., 2018), Nick Pelling selected three more YouTube videos claiming decryption of the manuscript’s written text.  There are presently half-a-dozen or so of such vid’s on YouTube, but the one Nick has chosen to post first comes from a Turkish family who – not unexpectedly – say the language is a ‘kind of Turkish’. 

Nicely serendipitous, given that my own post (below) went up just a few days ago… and Cuman is a language of the Turkic group… but now we will have to wait for any detailed evaluation until the  philologists at John Hopkins have their say.  That family  can at least be pleased that they  were re-posted at ciphermysteries, though I thought Nick’s commentary a little sparse, and was left unsure why Pelling would describe a claimed translation as a   ‘non-theory’. I guess the most intelligent explanation is that the claimed translation doesn’t present itself as merely theoretical.  Anyway, lets hope any comments it attracts are factual, focused on the decryption, and helpful for us all, whether the summations are on the positive side or not.  


I’ll add the video again at the end of this post. (A couple of annoying typos corrected 20/03/2018)


I  indicated recently –  in another place – that in my opinion the written text in Beinecke MS 408 may prove to be Cuman, a lingua franca of the  eastern traders and travellers over the higher ‘silk’ roads during the Mongol century.  I suggested further that the script used in this manuscript may be derived from Uyghur and/or Sephardi script.  I am well enough acquainted with the Nestorians’ liturgical language of Syriac to feel fairly sure that it is not a language or script informing the Voynich glyphs.


That there were other lingua francas used across the southern ‘spice’ roads has been known well enough to Voynicheros since 2011 when I posted about them.  At the time, it was still impossible to so much as suggest (without being ridiculed) that the Voynich manuscript was anything but the creative effort of some high-society  Latin Christian ‘author’ imagined resident in mainland Europe and with cultural horizons no wider.


By 2011, I had been explaining for three years that neither the manuscript, nor the facts of history, nor the imagery nor the manuscript sciences supported such traditional ideas about this manuscript.  The absence of any review, on any blog-site for the entire decade I’ve been publishing shows just how unusual the material was, and is.


Things began to shift from the ‘wholly Latin (western Christian) creation’ theme during the year of my sabbatical and I returned to find a slew of would be imitators – attempting to imitate the analysis of the Voynich map; adopting my comments on Armenian influence in southern India and south-east Asia – even an effort to base an entire theory around information in one post about perfumery and ‘small a’ alchemy.  The mice had had a year’s free run, and the habits of mouse-dom did not end because the cat had come home again.


Please don’t describe this as my ‘Cuman theory’ because the word  ‘theory’ is so used in Voynich conversations that for the majority it means no more than the words ‘idea’ or ‘notion’ or ‘novel-plot’.  People tend to experience the onset of that sort of ‘theory’ as a blinding flash of what they imagine is instinctive insight, and then to become as convinced of their new vision of the manuscript as any Paul after Emmaus.  Like many Voynicheros, Paul then spent the rest of his days trying to have others adopt the same ‘theory’ as an item of faith beyond all debate.


My background is in areas where imaginative ‘theories’ are not entirely welcome;  what clients want is information that they can, quite literally, bank upon and standing around hand-waving and offering nothing better than free-associations, best-guesses, and pure imagination with lots of ‘probably’ and ‘possibly’ and ‘my theory is’ just doesn’t cut it.


So – my opinion is that the language is very likely Cuman and the script quite likely derived from Uyghur script, influenced by others which I think include Sephardi script.


I very much doubt the text will be nicely homogeneous, orthographically or grammatically consistent or perfect and I should be surprised if it behaved nicely when computer-tested against such modern texts such as the American Declaration of Independence.  I  expect it might be so filled with loan-words, variant spellings, and efforts (correct or otherwise) to render phonetically sounds not native to the speaker’s first language that none but a human mind will be able to translate it, even if the basic language should be Cuman.  But I could be mistaken.  For an idea of what I mean, see:

  • ‘Of Portolans and Trabizond’, voynichimagery, 15th August 2012.

The Codex Cumanicus is an obvious first reference.  Here’s the lovely parrot from its folio 58, the same picture I posted a year ago when I started thinking Cuman a serious possibility.

detail from the Codex Cumanicus f.58 – as reproduced in ‘Parrot and Fox’, voynichimagery, January 26th., 2017.Quite apart from the indications of non-European origin and character that are provided by the Voynich manuscript’s imagery, the codicological evidence itself strongly suggests the same, and one can only regret that the televisual needs of the Austrian broadcasting company prevented either of the technical reports from following the usual course.


Had McCrone or the University of Arizona been handed their brief and allowed to get on with it, we should surely have learned more than we did. McCrone might have provided a full list of the pigments and identified the pigment binder.  These things together would have immediately narrowed both the temporal and geographic range.  As it was, the television company dictated the nature and type of substances, and even the specific details, from which the pigments were to be taken.  We ended up with a tedious description of the same ink, and various samples of exactly the same pigment in more than one case.  One of McCrone’s employees signed the letter in his own name and the date on the header was ‘April 1st’.


Again, when the University of Arizona assisted with radiocarbon dating, the television anchorman/representative informed the specialist that the permitted four samples were to be taken from here, here, here and here – thus immediately denying the sampling procedure any scientific rigor.  That is not to suggest that the date range obtained for those higher quires was not accurate; but it leaves us quite ignorant of whether or not the lower half on which those quires were stacked might not be older.  In fact the lowest sample (f.68) did return the earliest raw date – 1400 AD.


I have heard people try to excuse the issue with that radiocarbon dating by suggesting the television company and/or its anchorman picked ‘representative samples’ but since the aim of radiocarbon dating is to date the substrate, and non-specialists could only define ‘representative samples’ by their impression of pictures and writing, any definition of  “representative” was not the appropriate one.


Nor, in fact, were they ‘representative’ in terms of the manuscript’s ‘Currier languages’ or the codicology, because three of the four samples were taken from the top third of the manuscript and all three are written in Currier ‘A’.  All three are quarternions: that is quires formed of four sheets  folded in half (bifolio) to give eight sides or ‘pages’, these being usually described as the front (recto) and reverse (verso)  of each folio.


The problem is that the rest of the manuscript is not formed of quaternions, though quarternions are  the norm for manuscript of Latin or of Byzantine provenance.  In fact the rest of the manuscript is made of quinions, of fold-outs unlike any in another Latin manuscript, and the last, known as ‘Quire 20’ though it is the nineteenth in the present manuscript, contains six bifolia: a senion.  So rare is this in the western Latin tradition, overall, that most codicological texts don’t even include the word.


So those samples offer a skewed image of the manuscript as a whole, and just as with the instructions to McCrone, inhibited the scientific assessment of the manuscript rather than elucidating it to the degree those specialists might have done if given their brief and allowed to go about the work in which each was a specialist.  One does, of course, thank that television company for being willing to pay for the tests, and for the publicity which resulted in the wider world.  One may still regret that we learned so little overall about the reason for the manuscript’s plainly non-Latin elements and characteristics.


Quinion: Quire 13

About ‘Quire 13’ some interesting observations were offered some years ago by Glen Claston, and a couple of years ago, Nick Pelling revisited that conversation, kindly responding to a reader’s request. I do not agree with  Claston’s use of  ‘medical’, ‘balneological’ etc. even if Claston was aware (as many newcomers are not) that such terms were never meant as formal descriptions but were invented adopted as ‘tags’ for the sake of online conversation.


Nor do I think the content in Quire 13 was ever other than it is now, though I could agree it had been copied from more than one exemplar.  There is no reason to suppose, however, that this is why the quire is a quinion.

To sketch the possible implications of these quires in the untested second half, here are a few among the technical studies online:


Caspar Rene Gregory’s seminal study of Byzantine manuscripts, written in 1886, had already a basis broad enough to permit beginning with a simple statement:

‘The unit of construction for a Greek manuscript is the quaternion or quire of four double leaves or of eight pages’

  • Caspar Rene Gregory, The Quires in Greek Manuscripts, The American Journal of Philology, Vol. 7, No. 1 (1886), pp. 27-32

And a more recent writer could say as confidently of the Irish tradition that,

The standard for Irish medieval manuscripts was the quinion, or five sheets of parchment folded to make a gathering of ten leaves or twenty pages.

  • T M Charles-Edwards, Early Christian Ireland (2000).

and similarly, Gacek,

‘ the overwhelming majority of parchment quires in Arabic manuscripts consist of quinions (i.e. five bifolia, ten leaves).

  • Adam Gacek, Arabic Manuscripts: A Vademecum for Readers (2009) p.212.


So then where do we find quires resembling the Voynich manuscript’s Quire 20?

Senion: Quire 20

Germany.  The example is too late, being dated to 1450, but very few German examples adduced as comparison for Beinecke MS 408 have not been too late.

Codex Germanicus 6 is described as “a personal notebook, in Middle High German, dated to 1450. Rabin  note that “the 25th quire is very likely to have initially been a senion, but appears in the bound manuscript as a 7-bifolium quire” (p.128)

  • Ira Rabin, Oliver Hahn, and Mirjam Geissbühler, ‘Combining Codicology and X-Ray Spectrometry to Unveil the History of Production of Codex germanicus 6’ [pdf]


In Armenia,  between the 13th-16thC.

  • Dickran Kouymjian, ‘Notes on Armenian Codicology.  Part 1: Statistics Based on Surveys of Armenian Manuscripts’, Comparative Oriental Manuscript Studies Newsletter (4th. July, 2012) pp.18-23.  [pdf]


Jewish manuscripts,


Professor Beit Arié  is an eminent specialist in comparative codicology, which is the sort of specialist you need when an object has no certain provenance as yet – and as yet the material object that is Beinecke MS 408 still has none.


The quinion composition was the standard composition in the Middle East (in all Semitic manuscripts) and in Italy in all the dated manuscripts, which comprise around a third of all the dated Hebrew manuscripts. Nor are the several sizes of senion composition, which was the secondary composition used in Hebrew manuscripts in zones of Sefardic book culture (and was common among Latin parchment manuscripts in the 13th and 14th centuries, see below n. 48) consistent with the assumption that quires were created by folding the sheet. Collette Sirat proposes that the odd (quinion – five folios) composition of the early non-Hebrew codex quires written in the Orient were suitable for papyrus, which was rolled into a scroll after its manufacture, and then cut into any number of sheets,.


(This allusion to papyrus is compatible with remarks made by Georg Baresch in a letter written to Athanasius Kircher – a letter that has been oddly ignored by the same persons who treat as if it were fact a trivial bit of alleged-to-have-alleged gossip about an emperor).


All the above matter has been considered here in earlier posts, and readers referred to Philip Neal’s transcriptions and translations for the original documents.


I understand that some Voyncheros have no time to follow links and references so here is the content in that ‘note 48’ mentioned above:


48 For examples of manuscripts localised in the Orient and Italy and inscribed by immigrants who continued to employ the script of their country of origin for many years but whose book craft reflects – entirely or at least partially – the practices of the region in which they were active, see Beit-Arié, Hebrew Codicology, pp. 104-109.  MS Oxford, Corpus Christi College 133 is to my mind the most instructive illustration of the problematic and complicated nature of Hebrew codicology and palaeography, due to the unique mobility of the Jewish people. This manuscript, inscribed in the early, square and semi-cursive Ashkenazic script, is a copy of a prayer book whose unique text version is presumably descended from the custom of French Jewry, used by the Jews of England before their expulsion in the twelfth century. On two pages which were left blank (fols. 349v and 350r) numerous records were added attesting to payments received from prominent Christians from all over England (be it Bath, Norwich, Exeter or Winchester), half of which had been identified as having been active at the end of the twelfth century. Most unexpectedly, the records — inscribed no doubt by a moneylender who was manuscript’s owners at around 1200 C.E. — were in Judaeo-Arabic in a Sefardic (Andalusian) cursive script! The Sefardic owner of the prayer book noted that his records included מא כאין לי מן די אנא הונא פי אנגלטירה (‘all that I own since being here in England’. See Beit-Arié, Makings, p. 138 (translation and transliteration of the document by Ephraim Wust), pp. 147-148 (plates of the records); Beit-Arié, England, appendix (identification of the local English notables by Zefira Anton-Rokéah).


from: HEBREW CODICOLOGY: Historical and Comparative Typology of Hebrew Medieval Codices based on the Documentation of the Extant Dated Manuscripts Using a Quantitative Approach (Preprint internet English version 0.1)  [pdf]


It will be obvious enough that I do not discount the possibility that part, if not all the matter copied to form Beinecke MS 408, survived to the early fifteenth century because preserved first by the Jews of the south-western Mediterranean and perhaps even by the great Franciscan of Oxford. Bacon certainly Hebrew, recommended its study, and also knew many of the eastern plants that I have identified in the botanical section.


In the meantime – maybe Cuman.




for Oxford, Corpus Christi College MS 133 (not to be confused with Cambridge University’s Corpus Christi College MS 133), see

  • Peter E. Pormann (ed), A Descriptive Catalogue of the Hebrew Manuscripts of Corpus Christi College, Oxford,  (Cambridge: D.S. Brewer, 2015) where it is noted that seven of the thirteen Hebrew MSS which Corpus now holds were left to the College by John Claymond (d. 1536), its first President.
  • Eli Brackman, ‘The Oxford Haggadah: A Hebrew manuscript of Corpus Christi College, Oxford‘, web-page article, undated, for the Chabad Society, University of Oxford.

for Papyri

  • Thomas Christiansen, Marine Cotte, ‘The nature of ancient Egyptian copper-containing carbon inks is revealed by synchrotron radiation based X-ray microscopy‘,
    Nature, Scientific Reports Vol. 7, Article number: 15346 (2017).
  • Graham Shipley, The Greek World After Alexander 323–30 BC (2014)  pp.196-201.  For readers with no solid background in archaeology and ancient history, or who find those subjects of little interest, Shipley provides a very clear and accessible explanation of why no definitive geographic or temporal range can be offed for the use of papyrus as a writing material whether before or after the 3rdC AD.   In the Mediterranean, it was certainly being used for papal edicts and for documents in the Sicilian court as late as the eleventh or early twelfth centuries.

Cuman and related matter, including Franciscans.

  • Routes and Scripts (page), voynichimagery, September 11th., 2012  and  ‘Trade Routes and Scripts’ (blogpost), voynichimagery,  September 20th, 2012.
  • ‘Ottonian Manuscript’ (blogpost), voynichimagery, 31st October 2012.
  • ‘Across the North – Intro: fol.86v and prototypes for the Month-emblems’, November 21st., 2012.
  • .. and many others.  Search the blog for ‘Cuman’ but don’t miss…
  • ‘Parrot and Fox’ voynichimagery 26th. January 2017.

Franciscans and the roads east.

So much material from these posts was co-opted without proper acknowledgement that most of the posts are private now.  To give a general idea of the range of research which underlies my  ‘Cuman opinion’ I’ll quote a little from one of them,  ‘A thought about f.1r’, voynichimagery, February 5th., 2013.

…. As keen readers of these pages will know, I’ve suggested here that the red ‘bird-over-fire’ sign in the left margin of f.1r may be afol 1r blogreference to the aquilonarium Tartarorum – as the northern route was described in a letter written at the end of the thirteenth- or beginning of the fourteenth century –  allegedly sent from China to France or Italy by the Franciscan friar (and former military advisor to Frederick II of Sicily), John de Montecorvino, made bishop in China.  It would also serve well as his own ‘seal name’ in Chinese fashion: the mountain of Etna having been his home and he ‘corvino’.

Whether or not John of Montecorvino wrote it, the letter’s date does confirm that by that time, the custom already existed of describing the route in that way,  which is the important point in relation to dating….


Other keywords such as ‘Caffa, ‘Black Sea’, ‘Mongols’ ‘Sephardi’, ‘Uyghur’ and so forth should turn up quite enough to make you feel tired at the thought of reading it all – but I do like to be clear about the facts and artefacts before offering an opinon.  When we’re talking about opinions outside my area of specialisation, ten years seems fair enough to me.  It takes more than common sense and two ears to have opinions about comparative linguistics; if only people would realise it takes a lot more than common sense and two eyes to have valid opinions about imagery, too.



  1. and of course, given the Manichaean 5-element system I’ve described in explaining a detail in f.77 I should have added details as given in

    Medieval Manichaean Book Art: A Codicological Study of. Iranian and Turkic Illuminated Book Fragments from 8th–11th Century East. Central Asia, (Nag Hammadi & Manichaean Studies), Stephen Emmel, Johannes van Oort (Eds.), volume 57, Leiden-Boston: “Brill Academic Publishers” (2005).

    Brill itself names as author Zsuzsanna Gulácsi (2005)

    but I note online a Researchgate reference which credits instead Rachel Goldenweiser.

    Anyway, the table you want in that book is on p.64.


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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s