Notes on Panofsky’s comments 3 – hesitations

Header Illustration: (left) detail from Bodleian Library MS Marsh 144  p.211 11thC ; (right) detail from Sassoon MS 823 / UPenn LJS 057 Catalonia 1361 AD.

Two previous


Notes 2, 3 & 4: dating and provenancing ‘shapely ladies’

second edition, edited and updated  – 15th. Feb. 2019

Anne Nill?  detail (reversed) from a photo posted at No source given.

Anne Nill wrote:

[on first seeing the copy, Panofsky]  became intensely interested and seemed to think the MS. early, perhaps as early as the 13th century2  … but as he came to the female figures3 in connection with the colours used in the manuscript4  he came to the conclusion that it could not be earlier than the 15th century!

*’colours’ – he was first shown worn black-and-white negative copies.

“13thC? …..15th? …. 

Anne Nill conflates a question about dating manufacture (‘range of colours’), with one about dating content ( ‘shapely ladies’)  though it’s true that both together had caused Panofsky’s hesitation.

Eight decades on, the revisionist can consider each item separately and  Panofsky’s original judgement appears justified on both counts: manufacture, 15thC;  matter gained from older sources.  Some of those sources may indeed have been thirteenth-century.



‘Colours’ – The manuscript’s Palette:

detail from: Bexur, Driscoll, Lemay, Mysak, Stenger and Zyats, ‘Physical Findings’ in the Yale facsimile edition pp.23-37. original caption slightly edited but not altered.

Panofsky’s first dating manufacture of the manuscript to ‘not earlier than the fifteenth century’ would eventually become the consensus among persons whose work was in evaluating manuscripts.  By the early 1960s, as d’Imperio recorded:

“Helmut Lehmann-Haupt..stated in a letter to Tiltman dated 1 November 1963 that “there was near agreement on the date of the cipher manuscript as around, or a little after, the year 1400”.

Elegant Enigma (1978) p.8

These unnamed specialists, and Panofsky before them, were validated finally in 2011 by the vellum’s radiocarbon range : 1404-1438.

I’ll leave the subject of pigments for a later post, where I’ll compare Panofsky’s statement with Dr. Carter’s descriptive list of the palette  (recorded by d’Imperio), and by reference to a scientific study which was included in the Yale facsimile edition. Since the 1930s, and indeed since 1954 –  we have developed more precise techniques for analysis and identification.


Comment – Shapely figures

Panofsky was quite right to say that ‘shapely’ women (whom we’ll define by their swelled bellies) would not become a Latin fashion until the fifteenth century, but with more medieval manuscripts known today, we can say his original opinion may not have needed second-thoughts on this account, for research into the imagery in Spanish-and-Jewish manuscripts indicates that the form does occur there earlier, though interestingly only to represent metaphorical or allegorical ‘bodies’.  The closest comparison  found so far – since we must take  both stylistics and apparent subject into account  – is the ‘Gemini’ in MS Sassoon 823 (now: UPenn MS LJS 057).  The remarkably close similarity suggests a need to revise much of what has been generally assumed about the Voynich ‘ladies’.

(detail) f.77r

As our header shows, the ‘swelled belly’ emerged as an effort to imitate drawings in the first (pre-Ulugh Beg)  illustrations found in copies of al-Sufi’s Book of the Constellations. Those images in turn had reflected the traditions of  pre-Islamic peoples, including but not limited to, those of the Greeks and Romans. The rounded belly was most characteristic of an Indo-Persian style and we must consider that the works of al-Biruni may have had some part to play in  first formulation of the drawings illustrating al-Sufi’s tenth-century composition.

That remains to be seen.  However, the header for this post illustrates the progression of the style; the left panel shows a detail from  the ‘Gemini’ in an eleventh-century Iraqi copy of al-Sufi’s text; the centre shows the Gemini from MS Sassoon 823, whose content is a compilation of astronomical works, out together in 1361 in Catalonia, and the third panel is from another compilation, in a manuscript made (as we know) during the early decades of the fifteenth century.

The fourteenth-century Catalonian-Jewish figure has more in common with the Voynich manuscript’s unclothed figures than just the quirk which sees many of the bellies given a slightly-angular form.

They also have in common their curiously-formed ankles, flat feet and boneless-looking arms –  none of which elements appear in extant Islamic copies of al-Sufi’s constellation-illustrations, and none of which mars the later, more literal, fifteenth century ‘shapely women’ of Latin (i.e. western Christian) art.

There are differences evident, too.  A majority of the Voynich figures have heads disproportionately large, as the Catalonian figure does not.  More importantly (because even rarer ) many are drawn with overly large thighs in combination with bone-thin shanks, something shown most clearly in the Voynich ‘bathy-‘ sub-section, and which again is present neither in the Catalonian figure, nor in any remaining copies of al-Sufi’s work of which I’m aware.

That stylistic habit is not absolutely unknown, though  since it speaks more to the route by which the material had reached the west than our present subject, I leave it aside.

On the matter of proportions, which topic I’d brought forward quite early for its significance, the general indifference saw it ignored at that time, but more recently we have had a  lucid ‘revisionist’ post on the subject by Koen Gheuens, which I recommend:

. . . . . . . . . .

the chief point to be taken from this is that Panofsky’s judgement of ‘southern and Jewish’ content again finds support in the style of that drawing in a manuscript  predating the Voynich manuscript’s manufacture by at least forty years, and perhaps as much as sixty.

The possibility that its precedents could date from as early as the  the reign of Alfonso X (1254-1282) relies on the context in which the Sassoon ‘Gemini’ occurs, for even in Sassoon 823 its style of drawing stands apart.  To clarify, I rely on a paper by Fischer, Langermann and Kunitzsch, describing in detail the sections comprising the compilation of Sassoon 823/LJS 057.   The optional Preface clarifies another ‘ground hog day’ issue but skipping it will not lose anything from the main topic.



Optional preface: History of Sassoon 823/LJS 057 in Voynich studies.

I came across a notice of sale and acquisition for MS Sassoon 823 in 2011 in the course of my principal (non-Voynich-related) research.

Its description contained a greater-than-usual number of points in common with the Voynich manuscript – though at that time I was still the only person in the second mailing list to hold that Beinecke MS 408 was also a compilation from several earlier sources. (Today, I daresay, most would claim it general knowledge, and some would assert having known it all along.  Perhaps, if so, they might have lent a word of support at the time.)  Hunting more details of the manuscript, I had only an abstract of the article by  Fischer when I posted a note (in my old blogger blog, Findings) on Nov. 21st., 2011, listing the features I considered it had in common with the Voynich manuscript.  (At the time, a couple of the ‘German’ theorists were disputing use of the term ‘vellum’ and claiming the material could just as easily be described as German parchment.. which isn’t so, but they’ve come right on that matter since.)

A codex – probably fourteenth century – from the Iberian peninsula or thereabouts (Ceuta?) contains illustrations with human figures drawn short, and with distended bellies. One of these illustrations (for Gemini) is shown on p.288 of the article cited below.   That same article, written in 1988, provides the few details about the ms…

Article: Karl A. F. Fischer, Paul Kunitzsch and Y. Tzvi Langermann, “The Hebrew Astronomical Codex MS. Sassoon 823” The Jewish Quarterly Review , New Series, Vol. 78, No. 3/4 (Jan. – Apr., 1988), pp. 253-292

which says that the ms in question is:
*Inscribed in an ‘early’ Spanish hand.

*A florilegium – i.e. a collection of extracts.
*Vellum (?) rather than parchment.
*Total number of pages is greater than the Vms… but
*quires are also 8 pages each.


There is also apparently a  book [which could be an intro. plus facsimile, at 292 pages]: Karl Adolf Franz Fischer, Paul Kunitzsch, Yitzhak Tzvi Langermann, The Hebrew astronomical codex ms. Sassoon 823, Center for advanced Judaic studies, University of Pennsylvania, (1994) – 292 pages.

As you’ll see, some of those details were mistaken: the provenance is now established as Catalonia and the library presently holding it is clear about the date: 1361.

The next year, still unable to get hold of a copy of the larger study, and with the manuscript not (yet) online, I put out the word again –  through my still-fairly-new wordpress blog, voynichimagery (‘Curiosities’, Friday, Nov.2nd., 2012)

Still no response from any of the thousand or so who read that post.

By  2013, I was about to give it up, but because I had not found anywhere a drawing so like in both form and style to the Voynich ‘ladies’ as the Sassoon manuscript’s ‘Gemini’, I followed that manuscript’s progress after its purchase by the University of Pennyslvania (where it would be re-classified  Lawrence J. Schoenberg Collection, MS 057).

By 2013 I was also frustrated that no other Voynich researcher had yet investigated exactly where and when ‘swelled belly’ figures begin appearing in Europe’s Latin (western Christian) art, so I set out to investigate both topics in parallel and in earnest.  I acquired a photocopy of Sassoon 823/LJS 057… which was later digitised by UPenn.

Some of my research and results I shared in the context of posts about Beinecke MS 408, published at voynichimagery through 2013-2014.  Two, for example, are:

  • D.N. O’Donovan, ‘The ‘beastly’ Lombardy Herbal Pt1 – female forms’ (22nd July 2013); and then (e.g.)
  • __________________,   ‘ Talking about art and codicology’, 26th October 2014).

I referenced the paper of 1988 which I’d first read in 2011 – and from which I quote again further below.

  • Karl A. F. Fischer, Paul Kunitzsch, Tzvi Langermann, ‘The Hebrew Astronomical Codex MS Sassoon 823’, The Jewish Quarterly Review, LXXVIII, Nos. 3-4 (January-April, 1988) 253-292.

The most important discovery, in my opinion, was that the Sassoon ‘Gemini’ served as support not only for Panofsky’s location and character for the Voynich manuscript, but also for my own conclusions (published as early as 2011) that it is not only the ‘ladies’ in the calendar’s tiers, but all of them (and associated males) which were intended for celestial/immaterial ‘bodies’/souls.  To some extent, Nick Pelling (among others?) had sensed something of this in calling the figures “nymphs” – but it was also understood or intuited as early as 1921, by Professor Romaine Newbold, albeit he had interpreted that idea within the terms of  late-classical  neoPlatonist philosophy, rather than those of pragmatic astronomies.  (Some years later, Koen Gheuens would do something of the same, but in terms of the Latin mainstream and its standard texts: For the record, my own view is that we are seeing an older, more pragmatic tradition whose closest ‘cousins’ in the western Mediterranean are those of the navigator and chart-maker, whose terrestrial and celestial grids are constantly superimposed on one another.  However…

Having followed the trail of Sassoon 823 after its sale, corresponded with the new owner, written about it in posts (which were then still online and with the blog’s ranking, highly likely to turn up on any search),  I was disappointed to see that Darren Worley failed to refer to the precedent when, in 2017, he left a comment at Stephen Bax’ site announcing  the existence of Sassoon 823/LJS 057 in a way suggesting it a newly idea within the study.

At that time I  had a manuscript – a set of twelve essays – in the last stages of preparation for publication at that time;  and since academic editors do prefer no doubt should exist about the originality of work they have commissioned,  I asked Darren to acknowledge the precedent for form’s sake.  He did not.   No-one wants to be put in the position of being asked, in effect, why if their work is original, the same material is now seen everywhere (including and wikipedia) with not a mention of one’s own name as the first to have contributed the research, conclusions or insight.

Given that this relatively minor incident was only one of the great many similar – and worse instances that I’d had to deal with over almost a decade, I had no option but to stop sharing original material online, and to close voynichimagery from the public  – which I did soon after.  The issue has nothing to do with money, or copyright; it has to do with transparency and the honest mapping of the subject’s development over time.  (see the ‘About’ page)

On a brighter note, Worley’s comment itself had value.  I recommend it for his observation about the  quire signatures which I have not seen made before.


  The TEXTS IN MS SASSOON 823 AND THEIR PICTURES: Bar Hiyya, al-Sufi and anonymous. NON-LATIN LINEAGE.

Sassoon 823/LJS 057 was made almost forty years earlier than the posited ‘1400’, and fully half a century before the Voynich manuscript was made.

Whether Panofsky was right or not in first attributing the content in Beinecke MS 408 to the thirteenth century, its ‘swelled-belly’ figures offer no objection to a ‘southern and Jewish’ character ‘with Arabic influences’  – for that is precisely how the manuscript is described which offers our closest-known comparison for the unclothed  Voynich ‘ladies’.

Of the astronomical drawings in Sassoon 823, Fisher comment:

 the figures found in the Sassoon manuscript cannot have been copied from a manuscript of the Sufi latinus corpus, and equally not from an Islamic celestial globe. The only remaining possibility is that they were selected and copied from an Arabic manuscript of al-Sufi’s treatise.”

The text accompanying the Gemini figure (p.225) comes from an unnamed source, and the ‘Gemini’ image itself is not drawn in a way closely akin to any other, even in that manuscript.

The content in pages 195-228 is  described altogether as  “Astronomical Tables by Abraham bar Hiyya and others” and  In bold letters at the top of page 195 is written:  “From here onwards, from the Jerusalem Tables of the Nasi’  R. Hiyya the Spaniard, of blessed memory”

Kunitzsch adding his comment:

‘I know of no medieval astronomer by that name; however, the Nasi’ R. Abraham bar Hiyya is, of course, very well known, and in fact the tables in this entry up to page 214 are indeed his tables. On the other hand, I know of no other reference to Bar Hiyya’s tables as the “Jerusalem Tables.” … 

The ‘Gemini’ image (p.225) belongs to the additional, anonymous, material occupying pp. 215-28  which “deals mainly with astrology. Some of these tables are found in at least two other manuscripts which contain Bar Hiyya’s tables: Chicago, Newberry Library Or. 101, and Vatican Heb. 393.   Other items are unique to our manuscript…

ibid. p.272.

Chicago, Newberry Library Heb.MS 2 (unfoliated)

The ‘Gemini’ image may then have been brought into the Sassoon compendium with its anonymous(?) tables, not designed by Bar Hiyya but  found with his in at least two other manuscripts.   What is not known is how early the sources were joined – nor where – though ultimately the ‘Gemini’ (which we accept as deriving from an Arabic manuscript copy of al-Sufi’s ‘Book of the Constellations’ ) has to derive from the versions made before the time of Ulugh Beg, after which Gemini is differently represented.

Bar Hiyya  was known to the Latins as Abraham Judeus, and was born  three generations after al-Sufi’s death. (Al-Sufi  903-986; Bar Hiyya 1065—1136 AD).

Those manuscripts cited as containing the same tables, together with Bar Hiyya’s are not both presently accessible online, and Vatican Ebr.393 (1497 AD) though  digitised contains contains no constellation drawings. (Catalogue entry here.)  The Newberry Library informs me that the article by Fischer is mistaken. They have no ‘MS Or.101’, but they do have Heb.MS 2, whose content appears to be as described in that article. There are no constellation drawings in this copy.  At right, a reduced copy of one of the images very kindly sent me by the library.

Sidenote – ‘Jerusalem’.   David King demonstrated that in al-Andalus some at least had knowledge of Jerusalem latitudes;  an astrolabe  dated c.1300 has all its inscriptions save one in Arabic, the exception transliterating into Hebrew script the Arabic ” لعرض بیت المقدس لب li-ʿarḍ Bayti ‘l-Maqdis lām bā’” –  “for the latitude of Jerusalem, 32°”.

  • Abu Zayed & King & Schmidl, “From a heavenly Arabic poem to an enigmatic Judaeo-Arabic astrolabe” (2011), crediting  the Khalili Collection, London for the image.
  • David A. King, ‘Astronomy in medieval Jerusalem’ (Pt.2), revised and shortened 2018, available through

On Stephen Bax’ site (now in other hands) you will find various comments referring to Spain and to Spanish manuscripts, the work (chiefly by Darren Worley and Marco Ponzi) reviving and expanding the long-neglected opinion of  Panofsky, and later variation in Fr.Theodore Petersen’s work.

Checking the files of Reeds’ mailing list is always worthwhile; and I’d also suggest searching Nick Pelling’s long-running blog, ciphermysteries.   Running a search there before pursuing a ‘new line’ too far can often save you much time and effort – because even if Pelling has not looked at the subject himself, he may well mention that another researcher did.

A revisionist will want to revise past ideas and efforts, but it is always as well to begin by knowing what those were.

With regard to the ‘shapely ladies’ in Beinecke MS 408,  I should mention that the opinion of Fischer et. al. appears to preclude any close connection between them and the ’23 virgins’ which appear in a 9thC Byzantine diagram within Vat.Lat. gr. 1291.[Vatican City, Lateran Palace collection, Greek ms 1291]. The comparison has often – in fact continually – been re-produced since 2001 though without any effort to produce a formal argument, so far as I can discover.   It would appear to have been introduced to the study by Dana Scott in a post to Reeds’ mailing list (Mon. 12th. Feb. 2001), because ten days later (Thurs, 22nd. Feb 2001) Adam McLean refers to the diagram as if only recently mentioned.  The point remains a little uncertain because link to the image which Dana attached and labelled ‘Ptolemy’ no longer works.



Note: Swelled bellies in fourteenth century Bohemia.

Probably irrelevant to Beinecke MS 408,  I include this for the Voynicheros fascinated by Rudolf and his world.

The same essay continues:

To which globe are the (hemisphere) illustrations in the Sassoon manuscript most closely related? The answer is probably the globe of the King of Bohemia, Wenceslaus II (or III ?), which is now kept in Bernkastel-Kues and was first described by Hartmann.

The Spanish origin of the star catalogue in Sassoon 823 has already been established in Part I of this article (i.e. by Fischer, Kunitzsch and Langermann), .

Since the star illustrations in the Sassoon manuscript are similar to both Vienna codex 5318 [not digitised] which is considered to belong to the same family as Catania 87 [not found online] and the two hemispheres on pp. 112-13 of Vienna Codex 5415 [see Warburg database], and since both of these Latin manuscripts now located at Vienna originate from Prague, one ought to consider the possibility of Spanish influence on the manuscripts executed at Prague.

In the middle ages there were relations between the royal courts at Prague and Castile. The father of the present writer conducted research in Spanish archives before the civil war in that country which were destroyed in that conflict. He found there that the first known astronomer in Prague was sent as a “gift” by King Alfonso of Castille to Premysl Ottakar II King of Bohemia. … Previous scholars have frequently noted that Prague was the place of origin of many astronomical atlases.

 ibid. p.284

(Premysl Ottokar II was King of Bohemia 1253 -1278;  – D)

The Bohemian line of development shows an absence of some characteristics shared by the Voynich figures and those in Sassoon 823. Nor does the Voynich calendar show Gemini in this form But for the ‘ladies’ in the Vms’ bathy-section and for some of the surrounding figures in the calendar, we may suggest as one explanation, common emergence from that earlier, non-Latin al-Sufi textual tradition current in Spain,  the Bohemian works having been gained by second-hand exposure to them.  Of three examples illustrated by Fischer in another paper, it is only that  dated c.1350 which distinguishes the female figure by small, high breasts and none shows similar style for the limbs and hair as we see in the Sassoon manuscript.


Another section Sassoon 823 (pp. 25-29) contains extracts from Abraham Ibn Ezra’s astrological works – interesting in view of our earlier reference to the Voynich calendar’s month-names and their orthography.

Ibn Ezra, who also translated Ibn al-Muthanna’s commentary on the tables of al-Khwarizmi,  is recorded – in the Parma version –  as saying “The tables in the Almagest are useless”

  • above quoted from p.255 of Fisher, ‘Hebrew Astronomical Codex….’

and just to show that the eastern ‘swelled belly’ was often difficult for Latins to interpret, here’s what was made of it c.1300 by a draughtsman in Paris: the belly becomes a rib-cage, twisted sideways.



Prague 1350 AD



Few heeded the distinction between dates of composition and those of manufacture

The point is that this distinction between dates for manufacture and for content, when considered in concert with other items of evidence, (some of which have already been mentioned in these posts) obliges us to take seriously the possibility that our manuscript is a fifteenth-century copy of material gained from sources which may date to the thirteenth century – or earlier.

This is something which had been suggested even while the cryptanalysts were involved, half a century ago. In 1969 Tiltman seems to attribute to both Panofsky and the keeper of manuscripts his saying:

… the manuscript as we have it may be a copy of a much earlier document.

Quotation above from [pdf] John Tiltman, ‘The Voynich manuscript: “the most mysterious manuscript in the world”‘(1968).



Other than John Tiltman, the record of d’Imperio’s Elegant Enigma shows that the cryptanalysts around William Friedman evince a lack of regard for opinions of specialists in their own subjects.  Friedman is recorded complaining of the ‘naivety’ of university men and his behaviour towards Newbold and towards Panofsky reinforces this impression.

That curious indifference may be due partly to the diversity of those opinions, partly to individual bias, and in the case of Erwin Panofsky partly his uncooperative response in 1954, but more than those – so it appears to me – was the dichotomy presented by those opinions versus the cryptanalysts’ confidence that they had a role, and an important role, to play in the manuscript’s study.

Had they accepted the opinion of early fifteenth century date, they would have had to abandon their fixed belief that the written part of the text was ciphertext – one so resistant to their cryptological attacks  that they must presume it the invention of a highly sophisticated Latin, one having access to techniques not attested until the …  late fifteenth century…  early sixteenth century…   late sixteenth century… early seventeenth century…

Marcus Marci’s reporting the Rudolf-rumour had one clear benefit for this study. It set a definite limit on such rovings. Rudolf’s death occured in 1621.

Today, the ‘cipher-or-language… or other’ question remains unresolved, but the date for manufacture is set within narrow limits and obliges us to date the content, therefore, before that period 1404-1438.

And the content, like the ‘shapely ladies’ may derive from sources considerably earlier – as two of those specialists had pointed out.

In sum: Panofsky dated the pigments – and hence manufacture – in the fifteenth century.  He was right.  By reference to the ‘swelled belly’ figures, Panofsky felt his initial view of the content as “early… perhaps as early as the thirteenth century” could not be correct, and since he had no knowledge of that custom in art of the western Mediterranean before the fifteenth century, so he felt he must shift the date for content to co-incide with than of manufacture: 15thC.   Given the resources available today, we are able to say he was right about a pre=fifteenth-century date for composition,* since the ‘Gemini’ in Sassoon 823 is in a manuscript dated 1361, and made as he said by Jews of ‘Spain or somewhere southern’.

*the ‘pre=‘ dropped out during editing. Replaced today (15th Feb. 2019) with apologies to readers.

Moreover, that image occurs in a manuscript whose matter dates to a much earlier period and some of which is, in fact, dated to the thirteenth century and the time of Alfonzo X, a court in which (again as Panofsky said) you find influence from Islamic art in Jewish – and in Christian – art.


Notes 6 & 7  ... shows strong Arabic 6and Jewish influences.7 “

So far  little to oppose, but much to support this part of Panofsky’s original assessment.



Next post:  Notes 8 ‘Kabbala’; Notes 12 & 13:  Salomon and Liebeschutz.