What magic? Where magic? – 4. Whose magic?

[6th July – post shortened, by request]

I’d like to begin by quoting more of what Stuart Buck said to the assembled NSA tem in 1976.  I’ve added one word, because while keeping clear of most typical ‘blind spots’, he did overlook the possibility that what is contained in the written and/or pictorial text did not necessarily originate in the time and place the artefact was made.

from Buck to NSA at seminar 1976

As my readers will be aware, I have reason to be grateful for Nick Pelling’s willingness to treat decently with the newcomer, and to answer the usual research questions about sources, earlier studies on a topic, precedents and so forth. For many, there has been no other reliable source of information about such things, because regardless of his own preferred theories, Nick has always been willing to treat honestly with others’ work.

That said,  I should also make clear that each of us approaches the manuscript from a very different angle.

Pelling sees it as an historical problem and has said, in the past, that the aim of  historical research is to form a theory. 

My approach is  material and pragmatic –  identifying and then working to resolve any set or series of questions arising during close examination of an artefact.   In other words – my approach is question-driven and my research is thematic in style.  In response to an idea of the manuscript as about magic, my response is to consider the range and type of magical imagery and where – if anywhere – that item from the manuscript might belong.

In theory, there’s nothing wrong with forming an historical theory. In practice a theory, once espoused, can so easily lead researchers by the nose, to the point where they find confirmation everywhere and become so certain of the theory’s rightness that some have even invented fictions to serve as ‘patch’ over some hole in that theory.  I suppose what I mean is that theories can be seductive – as Newbold and too many others have found to their cost. The internal logic of an unfounded narrative can persuade even those who construct them.   It was the flaw in Wilfrid’s story, in O’Neill’s ‘note’ of 1944, in Brumbaugh’s papers and in much more Voynich writing produced since then. 

But at base, it is a problem of methodology, tools and validation.  

Methods – Thematic research vs theory-focused. 

A theory-driven study, on considering some possibility such as ‘magical’ content tends to define the parameters of study by their theory, not by the ostensible subject – the manuscript – and the topic for research – magical images.

Thematic research sets the limits of investigation without regard for any theory. Theories can wait, and wait indefinitely, and can – often must – be modified after the detail is, or isn’t, found to accord with both the historical evidence and the rest of  manuscript’s internal evidence.

In relation to the ‘occult Voynich’ narrative we see it informing method is  theory-driven because it begins where it ends: with the Germanic/central European theory or some other theory-defined locus, and without any effort to honestly consider the evidence for any other possibility.  No-one has, compared and contrasted – for example – alchemical images from Spain, England, Italy, France, Sicily and Byzantium or the Aegean islands and concluded that the closest in style to the Voynich images are characteristically ‘Germanic’.  But to be of assistance to those working to understand the written part of the text, such a process of investigation and elimination is important.  

I think I should give a specific example here of the contrast in method between a theory driven driven approach and a thematic, question-driven approach works in practice, but since much of my work in recent times has been doing ‘background checks’ on ideas currently circulating, it’s difficult to find an example that won’t upset someone.  I’m reminded of what Curt Zimansky once said, as he began a talk:

I HAVE CHOSEN TO TALK about a work that has been praised and damned for two centuries, about which one cannot venture an opinion without offending two thirds of one’s colleagues, and about which there is still no critical agreement. … Where there is so much disagreement about a work of acknowledged importance all parties will plead from the historical approach. I hope to show that … history can be used to re- inforce a prejudice, or can combine with critical techniques to obscure what is clear; that on the other hand a rigorous use of the history of ideas can rectify error and permit us to extend critical dimensions. (p. 45)

  • Curt A. Zimansky, ‘Gulliver, Yahoos, and Critics’, College English, Oct., 1965, Vol. 27, No. 1 (Oct., 1965), pp. 45-49.

Perhaps the best illustration is one which has no clear-cut conclusion, and so I’ll use the example of a recent enquiry into whether or not, as Pelling recently suggested, Quire 20 should be imagined two quires rather than one,

Nick Pelling remarked – almost as if were self-evident – that:  

… Q[uire] 20 … contains far too many bifolia to be a single quire,… I think may originally have been constructed as two separate gatherings Q20A and Q20B.

That’s the theory.

Readers might find it plausible; they may find it suits a preferred theory of their own. It may be a valid notion. It may not.

My first inclination, then, is to consider first that the Beinecke library’s description has the quire as a single gathering, one which was a septenion (seven bifolios) but from which the centre bifolio has been lost. 

Pelling says it has  “too many” bifolios and my immediate question is  “too many compared to what?” –  and the first answer which occurs (and which may be inaccurate) is that the comparison is to  ‘normal’ Latin custom. 

So where is the justification for altering the form of a manuscript to make it less obviously unlike a Latin ‘norm’?  Especially this manuscript whose quires contain so much else that has no parallel in any known Latin manuscript.

Its ‘fold-outs’ might be better termed ‘fold-ins’ since they most resemble scroll-lengths folded in to the size of the Voynich quires and stitched in. Then there is the assertion, always somewhat problematic, that ‘Voynichese’ is written in a humanist hand.  The Voynich pages’ lack of ruling out and lack of simplicity in their arrangement has always seemed to me to sit uncomfortably with that ‘humanist’ idea.  But it may be right.

In terms of codicology, too, there are other other a-typical quires, including what had been two quinions.

Quire 8 was- and quire 13 is a quinion.


Evidence for Quire 20 as  originally a septenion seems clear enough.  The quire was sewn as a single quire.  

But perhaps  it merits some digging to see how unusual were septenions and quinions in the fifteenth century.

A first search through JSTOR produced an article which looked promising.

  • E.K. Rand, ‘A Harvard Manuscript of Ovid, Palladius and Tacitus’, The American Journal of Philology, Vol. 26, No. 3 (1905), pp. 291-329 (39 pages).

Rand’s article is of interest not least because the the ‘humanist hand’ is usually said to have been first used by Poggio Bracciolini and one of the two manuscripts bound in the Harvard volume is written in that humanist hand, includes an introductory letter to Bracciolini, and consists of a single septenion. 

The other of the two manuscripts bound to form that volume consists of ten quinions.

I’ll quote some of Rand’s commentary.

The most important fact omitted in Quaritch [the seller’s catalogue] is that the volume includes two separate manuscripts; they are noted here as MS I and and MS. II. The contents of the volume are as follows: fol. I-4. Two uniones, added when MS. I and MS. II were combined. .. Both MSS., naturally, were written before the date of binding. ..

Manuscript 1: “This manuscript consists of a single septenion. It has 22 lines to the page.  The text occupies 12.2 x6.4 cm.
fol. 5. Rynucius Poggio suo Oratori Eximio I felicitatem (in red) Ille Rem optimam et sibi salutarem ….. (fol. I8). At inuita nemini datur effugere fatum. (One line blank) FINIS.
fol. x8.’ Blank.
An unpublished letter of Rinucci da Castiglione to Poggio, with translations of the Athenian decrees contained in the De Corona of Demosthenes. The letter must have been written before I459, when Poggio died; probably before I453, when he left Rome; and possibly much earlier still, as he was studying Greek with Rinucci as early as 1425. See Voigt, Wiederbelebung des klass. Alterthums, 1893, II, pp. 45, 84. The present copy might well have been made about the middle of the century.

Manuscript 2:

MS. II. This manuscript consists of ten quinions. It has 23 lines to the the text occupies 12.7x 6.4.

N.B. This manuscript contains a work by Ovid, and while it is Ovid’s Heroides, not his Metamorphoses, it’s only fair to mention that Koen Gheuens has (or had) a theory that the Voynich text represents matter from the Metamorphoses.

While I am unable to agree with Koen’s reading of the Voynich drawings, I have long been of the opinion that the oldest chronological layer informing the drawings – if not all the drawings in every section of the Voynich  – is Hellenistic, first enunciation having been, in my opinion, contemporary with the Seleucids. A later layer I date to the 1st-3rdC AD and  latest of all (barring some pigments and marginalia) to between 1290-1330.  Manufacture of our present manuscript-as-object I date, naturally, to the early fifteenth century..
  • Harvard has made the volume available online. Phillipps MS 6748 describing it as “an anthology of humanist texts” and dating the whole, as bound, to between 1425 and 1500. 

What was being copied by these humanists and scribes were copies of ancient and classical texts in versions previously unknown to Latin Europe, being brought or fetched into Italy, by Byzantine immigrants.

So, we  have one useful example. The Harvard volume’s manuscripts were certainly made and written in Italy. Both are dated to the fifteenth century, with the septenion perhaps inscribed within a short time of the Voynich manuscript, and since the Harvard volume shows the humanist hand already in use (before 1440?) we don’t have to abide by the usual dating for that hand’s appearance in Italy. The layout issues remain, but there seems no obvious reason for supposing septenions weren’t being used in fifteenth-century Italy.

A cross-check with Beit Arie, reveals another example. Beit Arié says that septenions are very rare in Jewish manuscripts before listing 31 examples. He says they are “found only in seventeen paper manuscripts, in thirteen manuscripts with mixed quires from Spain, Italy, and Byzantium, and in one Italian parchment manuscript.” Beit Arié does not distinguish vellum from parchment.

  • Malachi Beit-Arié, Hebrew Codicology: historical and comparative typology of hebrew medieval codices based on the documentation of the extant dated manuscripts using a quantitative approach‘, unrevised (2018) preprint  of

So there’s an interesting possibility – that text in  Quire 20 might have been copied onto vellum from a septenion quire in a paper manuscript .

Note to self – what is known of the parchminers’- and stationers’ network for the early fifteenth century? Quires ready-made?  

Like the Voynich manuscript, the Harvard volume also contains quinions.

Less rare than the septenion in Latin European manuscripts, they were actually standard for quires in medieval Arabic manuscripts. 


Altogether, so far, this information raises a possibility that quires of this type, being atypical for Latin European works, may have been – for some reason as yet unexplored –  deliberately selected and/or what was on offer from the stationers or parchminers, who not only sold sheets of membrane but ready-made quires. 

Another possibility is that use of the quinion, and/or the septenion, was some quirk of the small circle of 15thC Italian Graecophiles and/or a usage familiar to the Byzantine Greeks who had emigrated into Italy. 

A third possibility is that the ancient and classical texts which are copied into that Harvard volume had been of that form when brought into the west, and so the form as well as the content was being imitated – perhaps as sign of authenticity, or to avoid risking quires of a rare text being separated.

The ‘Graecophile/Byzantine’ possibility took prioriy –  first, because if it were found to so, it might tell us what sort of text may be written in the quinions and septenion of the Voynich manuscript. 

Naming a paradigm.

Which of the various Byzantines sent to fetch manuscripts from libraries of the Greek domains shall I mention here?  Perhaps one from the second half of the fifteenth century.  Janus Lascaris will do well.

Janus Lascaris was known to the Latins as John Rhyndacenus ( i.e. from Rhyndakos in Asia Minor. The Rhyndakos village and river are now known as Mustafakemalpaşa,

He was a scholar from an eminent Byzantine family, and being native to Asia minor was surely as familiar with the other Byzantine capital of Trebizond/Trabzon on the south-eastern shore of the Black Sea.

Having come to Italy under the patronage of Cardinal Bessarion, Janus Lascaris was made welcome in Florence by Lorenzo de Medici after Bessarion’s death.

He was then sent twice  by Lorenzo to fetch copies of ancient and classical texts from ‘eastern parts’ and we are fortunate that some of Lascaris’ personal notebooks survive, listing titles wanted, titles sighted and titles bought.  One of these notebooks records his itinerary, which ended with his returning to Italy with 200 manuscripts from Athos:

From Florence Lascaris’ itinerary took him to Ferrara, Venice, Padua, Corfu, Arta, Salonica, Galata, Sozopolis, Athos, Salonica, Crete, Apulia, Corigliano, and Monte Sardo. He lists manuscripts acquired or at least seen at each of those points. We shall restrict ourselves to Athos where, apart from one book at Chilandari and another at “Simenou» .. he confined his attention to the collections of Vatopedi and Megiste Lavra.

I’ve had reason to mention Vatopedi before. To save readers the effort of finding the earlier reference, here’s the critical paragraph again.

That the texts of Strabo, and of Dionysius of Byzantion were still known and copied in Constantinople during the early fourteenth century is proven by the deservedly famous Vatopedi manuscript, a compilation of texts from major and minor classical authors describing the sea-routes of the Black Sea, Red Sea and to as far as England. It is difficult to think other than the compilation was made for contemporary needs, and these may have included the needs of foreigners resident in the enclaves of Pera and within Constantinople, wanting to know those routes.

from: D.N.O’Donovan, ‘The skies above Pt.5: bodies in baskets’, voynichrevisionist.com (12th September, 2019).

Manuscripts copied by Janus Lascaris again include quinions, Young mentioning specifically, 

Venice, San Marco, Codex XCII, 7 (gr. 522), ff. 181-198, given in 1468 by Cardinal Bessarion, describing it as “a handsome vellum codex, of 23 quinions, 268×193 mm.” and Am(brosianus) D 210 inf. (gr. 940), “a parchment quinion, 292×160 mm., contains Theognis vv. 1-618 only, 31 lines a page in a writing-space 180×80 mm., … It follows another quinion with an unfinished Timaeus Locrus De natura mundi, on parchment leaves, 274×157 mm.,, J. Lascaris writing 35 lines a page in single column.”  And again, Florence, Laur. XXXI, 20, ff. 35v-57r, leaves being lost between ff. 41/2, 42/3, 44/5, 45/6 … the second and fourth broadsheets of a quinion having gone astray.”

  • Douglas C. C. Young. ‘A codicological inventory of Theognis manuscripts With some remarks on Janus Lascaris’ contamination and the Aldine editio princeps’, Scriptorium, Tome 7 n°1, 1953. pp. 3-36. 

A much more detailed account of Lascaris’ travels, including mention of Lascaris’  notebooks:

  • Graham Speake, ‘Janus Lascaris’ visit to Mount Athos in 1491’, Greek, Roman and Byzantine Studies 34 (1993), 325-30.

Review of information so far.

  1. Evidence that the humanist hand was employed as early (perhaps) as 1425.
  2. Evidence that works of that date, in the humanist hand, using quinion and one example of a septenion can be identified as work accomplished in Italy.
  3. The interest of humanists was chiefly in copying classical and ancient texts brought from ‘eastern parts’.
  4. The page layout and scribal customs seen – so far – in the Humanist and Byzantine scholars’ works do not accord with those of the scribes who worked on the Voynich manuscript.

Interim conclusions

From the evidence sighted and cited so far,  I can draw only one conclusions –  that there is not enough in the historical record to support Pelling’s theory that Quire 20 has ‘too many bifolios’, and enough to dispute the idea. No justification exists for attempting to presume that what was the usual habit in central Europe should be imposed on Italy, Spain and southern regions.

The quire-stitching does not support the idea, and the historical evidence considered so far  shows that while septenions would appear to be rare, they are not unknown and we have already seen two attested from Italy, one being reasonably attributed to the first half of the fifteenth century and inscribed in a humanist hand, the other being sighted but not described in more detail by Beit Arié,


To end this post-  description of another fascinating manuscript held by Yale. Made about a century before the Voynich manuscript, it even includes mention of alchemy (and no I don’t have ay ‘medical Voynich’ theory. I avoid having theories; I find they interfere with work..



Yale University, Harvey Cushing/John Hay Whitney  Medical library Manuscript 28. (here). 

Medical compilation (“codex paneth“). Northern Italy, (Bologna ?), 1st quarter of the 14th century. Vellum; 685 folios; 2 35 X 337 mm· (Yale Medical Library, Ms. 28)

The curriculum of a fourteenth-century medical school was based on works of Hippocrates and Galen, rounded out and brought up to date with writings of Arabic origin and the best contemporary physicians of Salerno, Bologna, and northern Europe.

The “Codex Paneth” preserves precisely such a collection of medical tracts.

The forty-two separate tracts have been identified by K. Sudhoff. Included are several works of Hippocrates and Galen, others by such authors as Roger of Salerno, Rhazes, Albucasis, and other lesser known French and Italian writers.

Each tract is introduced by a historiated initial showing physicians in discourse with students or patients with various afflictions. The subjects covered in the compilation include anatomy, bloodletting, acute illness, diet, urine, the pulse, diseases of the eye, childhood diseases, herbal and lapidary remedies, alchemy, astrology, medical recipes, veterinary medicine, and an Arabic-Latin vocabulary (fols. 235^2 38V).

A large portion of the manuscript, some one hundred eighty folios, deals with surgery, not usually included in traditional medical texts. Perhaps the most interesting of the surgical treatises is that of Albucasis (fols. 200r~318v) in the translation of Gerard of Cremona.

It contains two hundred fifteen paintings of surgical instruments, inserted in the appropriate places within the text. The manuscript was written by two or possibly three scribes in northern Italy, probably at Bologna, the most likely place where such a vast compendium of current medical writings might be found at this time.

Two distinct hands are to be discerned also in the illustration, which has a generally Bolognese flavor. The volume was in Bohemia by 1326, as an obit on the last folio indicates.

Provenance: “Tomazlaus notarius Mylewfensis],” 1326 (obit, 6851:). Cathedral of Olomouc (Czechoslovakia), 16th century (fol. 4). [bought from] Fritz Paneth, Königsberg. Gift of the Yale Medical Library Associates in 1955.


  • Walter Cahn and James Marrow, ‘Medieval and Renaissance Manuscripts at Yale: A Selection’,  The Yale University Library Gazette, Vol. 52, No. 4 (April 1978), pp. 173-284.