I’ve waited a decade to find anything good on the Rohonc Codex (and don’t get me started on Wikipedia yet again), so it is with great delight that I read Benedek Lang’s April 2010 Cryptologia article “Why Don’t We Decipher an Outdated Cipher System? The Codex of Rohonc” that he kindly mentioned in a comment on this site a few days ago.

Despite the slightly clunky title, I think it is fair to say that Lang’s piece utterly replaces pretty much all the previous writing on the subject, and arguably moves the Rohonc Codex very nearly on a par with the Voynich Manuscript. Really, it is almost unnerving to find out that the RC suffers from precisely the same issues bedevilling VMs research:

  • wide possible date range (1530s [from the Venetian paper] to 1838 [when it was donated by Count Gusztáv Batthyány])
  • uncertain provenance (one possible mention in a 1743 inventory, but that’s it)
  • inability to narrow down the plaintext language (Old Hungarian? Latin? or what?)
  • apparently unhelpful drawings (probably representing a life of Christ, but offering very few cribs)
  • non-trivial cipher nomenclator / shorthand combination (in my opinion)
  • dominant hoax narrative (but which is at odds with the early dating of the support medium)
  • unsubstantiated links to murky historical figures (forger Sámuel Literáti Nemes rather than Dee & Kelley)
  • inadequate codicological and palaeographical analyses (by modern standards)
  • multiple hands contributing to the object’s construction (two in the case of the RC, it would appear)

To me, the RC and the VMs (and their complicated mad ecologies of attempted decryptions) seem like two expressions of the same underlying historical pathology – when the aspirational desire to reconstruct the what overwhelms the grounding need to look for the how. Hence I asked Benedek Lang the same kind of “Voynich 2.0” questions I try (in vain) to start from these days, to round out the parts of his article that are less obviously cryptological (yet still important). Here are his responses (very lightly edited)…

* * * * * * *

[NP] (1) Has anyone done a codicological analysis of the Rohonc Codex? That is, how confident should we be that the bifolios remain in their original gatherings/quires and nesting order and that no bifolios have been lost, and when was the cover added, etc? Are there any signs of multiple rebindings? Are there any fingerprints?

[BL] No fingerprints, but basically anyone can touch it in the library, and some people in the 19th century even made notes in it. There had been a little research regarding the watermark, which I largely confirmed with my own research, though this however says nothing about the writing itself (which might of course be a later addition). The beginning and the end of the book are quite destroyed, to the point that the first and last 20 pages are no longer bound into the book, hence their (19th century) numbering might well be wrong. I think the book is in its original binding, which is not a real binding, just a piece of leather.

(2) Has there been a systematic study of any apparent corrections by the author(s)? For example, I notice a line apparently crossed out in Figure 5, or is that just boxed for emphasis?

No, nothing. My impression is that the corrections do not say anything that makes sense to me, but I should perhaps pay more attention to this.

(3) Has there been a palaeographic study of the text itself? For example, might it have (Leonardo-style) been written right-to-left for convenience by a left-hander? And have the palaeographic differences between the hands been described carefully? For example, did all the hands form the letters in the same way?

No, nothing, although it would be good to know whether there really are two hands – as it appears to me – and whether the text was written by one left handed person (or two), or just in the other direction by a right handed person.

(4) Has there been a palaeographic study of the marginalia and (what appear to be) interlinear notes? As with the VMs’ 15th century quire numbers and marginalia, dating the folio numbers might give a far more limiting (if pragmatic) terminus ante quem – really, there ought to be _some_ internal evidence that can help improve on 1838, which in historical terms is practically yesterday.

These marginalia were made by one of the less clever late 19th century “scholars” who believed that they were able to decipher the text.

(5) Apart from the introduction of new symbols, are there any signs of evolution or development of the core writing system through the 450 pages? As new symbols are added, are they progressively more ornate (which would argue for them being improvised, rather than as part of a pre-existing system)? Furthermore, are there any places where a new symbol is added in a left-right textual context which recurs around a word earlier in the document? (This would again argue for a nomenclator being improvised during the writing process).

There are certainly some occasional changes – for example, one of the symbols (the winged one) becomes less ornate – but apart from this I do not see any systematic changes. It is also true that new signs are introduced when there is a new person in the text (Pilate, for example). But I have not done serious research into that question.

(6) Did the Battyhany family ever compile inventories of their library? Has anyone looked for provenance in this kind of way?

Yes! There are several partial inventories of this very large library, and some earlier Rohonc Codex scholars thought that a book entitled “Hungarian prayers” in a 18th century inventory referred to this book. However, I remain skeptical, for I would be more satisfied by an inventory entry along the lines of “a book with unknown signs”. Such a description, however, is absent from the catalogues, the last one of which is dated exactly 100 years before 1838, when the codex first appeared.

As a general comment, I’d say that the lacuna in your account of shorthand is between Tironian notae and Bright’s Characterie. In Italy, Quattrocento scribes built up local traditions of abbreviations, with “underbars” and (macron-like) “overbars” for contraction and abbreviation (there are even some of these in Alberti’s facade for Santa Maria Novella). Isaac Pitman’s history of shorthand also mentions (p.6) a (probably 16th century) “Mr Radcliff, of Plymouth” whose version of the Lord’s Prayer – “Our Fth wch rt n hvn : hlwd b y Nm” – looks rather like modern SMS txtspk! What links many of these, then, is that they were ugly systems of abbreviation mainly intended to capture charismatic sermons as they were spoken: and so Bright’s innovation was to make the strokes easy to write, rather like Greek tachygraphy (which, though it was used in antiquity and in the Byzantine Empire, never seems to have crossed over into Europe).

Thanks! I was not aware of that.

In this context, then, the Rohonc Codex’s awkwardly angular letter forms seem to me quite independent of the many post-Bright shorthands: and also seem to have nothing structurally corresponding to the characteristic underbars or overbars of Quattrocento scribal practice. Hence to my eyes, it seems unlikely to fall within any known shorthand tradition, save that of pure abbreviation / contraction.

Yes, I agree.

As with the Voynich, I think the most likely scenario for the Rohonc Codex is that it is formed of a combination of (specifically abbreviating  / contracting) shorthand and non-polyalphabetic cryptography (though it seems very likely that the VMs’ cryptographic aspect is many times more sophisticated than the Rohonc Codex’s): and it is this pairing when also combined with the lack of knowledge about the underlying language that makes it impractical to crack in a conventional way. In both cases, I suspect that the necessary first step will be to crack the history first!

Yes, but what can be done when almost nothing is known about its history? The Batthyány family might well have purchased it anywhere. In my mind, I imagine that it is a combination of a shorthand and a cipher, though lately however I am convinced that it is a consonant writing (due to a possible Turkish or Hebrew origin) and a cipher applied to that consonant language. (In fact, this is almost the same as saying that it is a cipher and a shorthand, because shorthands are usually composed of consonants.) I do not believe that it is a hoax because it is an ugly book, and I do not really know of any similar hoaxes from the pre-19th century period. I was, however, convinced that the Voynich Manuscript was itself a late 19th century hoax until I learned about its new dating. Hence I remain puzzled!

 PS: do you have a picture of yourself I could include in the post? Thanks!

Benedek Lang

* * * * * * *

So there you have it – the Rohonc Codex is very probably, as Lang’s piece implies, just as uncertain as the VMs. Yet where are the massed ranks of me-too US documentary-makers clamouring to go to Budapest to view it? Why can’t we hear William Shatner’s voiceover ringing in our ears? 

To me, the central mystery of the Rohonc Codex is therefore why its ‘ugly duckling’ cousin [the Voynich Manuscript] gets all the mad heresy theories when it’s the Rohonc Codex that has all the pictures of Christ. (Note to novelists & film companies: Budapest is much prettier than New Haven). Go figure!

23 thoughts on “Benedek Lang’s Rohonc article in Cryptologia…

  1. Thanks for the abstract and exchange. I’d like to see a follow-up on this:
    “New signs are introduced for example when there is a new person in the text (Pilate, for example).”

    A question that might have been examined in the article is whether the writing system was in existence before the Codex was created. It appears the vocabulary was exhausted, or nearly so. Likewise, in the VMS, the word list may have been expanded with modified systems in preparation for additions to the text. Assuming the bound sequence is correct, word overuse (as opposed to change in subject matter) in Quire 13 demanded invention of new words for Quire 20.

    Conjecture re. “history first”: The Codex was written in an area in which a religion (or religions) had been introduced as a political expediency. The creator had only enough knowledge of the religion to convert the inhabitants. We need to look to a place and time in which the artist was not familiar with an iconology that placed feathery wings on angelic mammals.

    I posted a sequential view of Rohonc Codex low-resolution illustrations apart from the script. Unfortunately the site is so slow as to be useless. Occasionally it isn’t. You are welcome to try.

    http://syndrome8.netcipia.net/xwiki/bin/view/Main/

  2. Well, the Rohonc certainly has one advantage over the VM (in terms of cryptanalysis), and this is that it’s character set is fairly clear. While with the VM we still struggle to find out whether two characters are really different or the same, or whether a set of strokes forms a single, two or three characters, this appears to be more obvious in the Rohonc.

    But what makes the VM more appealing to a wide audience is certainly the illustrations. I mean, what’s a bunch of dudes (no, kings! — Big deal…) walking up a hill in the Rohonc, compared to a set of naked chicks doing a square dance in someone’s innards…?

  3. Elmar: it seems to me that the cultural appetite for ‘enciphered heresy’ is so large that many people want this to be the de facto explanation for the VMs – I couldn’t count the number of times I’ve read that the VMs ‘must surely’ contain something dangerous to the Church, la-de-da. And this in a document which has at most one cross in it (and even that is somewhat dubious)! By way of comparison, the Rohonc Codex gives every impression of being an enciphered / shorthand life of Christ, so why pick on the poor old VMs? *sigh*

  4. To me the rohonc looks like an old embalming manual, perhaps Christianised at some time. Intriguing similarities to some figures in the Artemidorus papyrus too. I’d be quite prepared to accept it as a book of “Hungarian prayers”. In fact, I’ll blog a bit about the curious ruling on exchanging an older Hungarian script for Latin script in the tenth century. It’s always assumed that the only earlier script was Orkhon or rovas runic, and maybe so, but I believe that there’s room for doubt.

  5. Dennis on August 26, 2010 at 7:25 pm said:

    The VMs is a lot more esthetically pleasing to me than the RC. Many people have said the drawings in the VMs are crude, but those in the RC really are. Elmar is right, the content is a lot more appealing, too. OTOH, the likely underlying languages, such as Hungarian, Romanian, Czech, etc. are less accessible to Westerners than the likely ones for the VMs.

    The ‘enciphered heresy’ hypothesis is really a lot more plausible for the RC, given its overt religious context. One of the striking things about the VMs is the almost complete lack of religious symbolism, pace D’Imperio. More characteristic of a herbal from those times, I’d think.

    Why don’t we send Dan Brown a copy of the VMs and the RH each, and see what he does? Just try to get a cut of royalties! 🙂

  6. rootlesscosmo on September 17, 2010 at 6:39 pm said:

    My generation (I was born in 1942) of offspring of Communist families may well have recognized the name Voynich, as I did, as that of the author of “The Gadfly,” a melodramatic novel of Irredentist and anti-clerical tone set mostly in 19th century Italy. (That author, Ethel Voynich, was the daughter of the logician George Boole, though this, while arresting, sheds no light [as far as I know] on the Manuscript itself.) There are three Soviet-made film versions of “The Gadfly;” the second has a score by Shostakovich which is sometimes played on orchestra programs. Could the dissemination of “The Gadfly” partly explain why the VM is better-known than the RC?

  7. Sleepangel on March 28, 2011 at 9:53 pm said:

    Imagine, if you will, a language that is ever-changing…like the strings out of a random number generator.
    To us, who have been disabled, it would appear to be nothing more than a faint impenetrable echo of lost ability’s…
    but to those that remain intact, a reading would be child’s play.
    ps. Time itself, would be embedded in the text.

  8. SirHubert on December 5, 2012 at 5:49 pm said:

    Please excuse what may be a very stupid question, but has anyone actually attempted a transliteration of most or all of the Rohonc? A few hours searching online has yielded images of about 220 pages and quite a lot of speculation about what the script and pictures resemble, but I can’t find anything that looks like good, solid, dull statistical analysis. I’m almost tempted to have a go myself, but if someone better qualified has done this already I’m happy not to do a quite bad reinvention of the wheel…

  9. SirHubert: to be honest, I’ve been waiting for a little bit more to emerge from Benedek Lang (he’s written a book on it in Hungarian) – as I recall, there are a number of partial transcriptions, but no single one. I’ll drop him a line to see what’s available, I like a nice bit of “good, solid, dull statistical analysis” myself too. 🙂

  10. SirHubert on December 8, 2012 at 12:00 pm said:

    Nick: thank you for following this up for me. Anything which saves me having to work through all 450-odd pages, or even which gives me something to check my readings against, would be a godsend. I would be very surprised if nobody else had done this, given that a fairly casual glance suggests quite a bit of repetition within the text, ranging from groups of three or four characters (words?) to much longer strings (sentences?) but I can’t find anything like this in English (and I don’t read Hungarian). If nothing else, it will be a nice change from the Guardian crossword and a sudoku for whiling away a long commute to work!

  11. SirHubert on May 29, 2013 at 9:26 pm said:

    I’ve now spent some hours scribbling on grainy printouts of the first 70 pages with a biro (appropriately Hungarian) and orange highlighter, and have puzzled quite a few fellow-commuters while doing so.

    The symbols seem to fall into two broad categories. Some are really quite simple in form and also, in my opinion, would be easy to write quickly. This might fit with the comments by both Nick and Benedek Lang about shorthand. But others are far more complex to the extent that they are really pictograms, and it is hard to believe that these could have been designed for speed or ease of writing.
    Some of the simple characters could have been created by something like a PigPen system, using semicircles facing up, down, left or right. These semicircles can be empty or contain one or more dots, Another common character is made up from two facing semicircles with a central pellet (a bit like (.) ), apparently a variation on this theme. A similar symbol with a straight line instead of a semicircle on the left hand side (a bit like I.) ) may also be a single character similar to the last, rather than a combination of a vertical line (apparently also a symbol in its own right) with the very common ‘backwards C and dot’: .).
    I’ve seen most (not yet all) of the theoretically possible arrangements of these in the pages of the manuscript I’ve examined thus far, and there are several other symbols which might derive from a similar system.
    I would therefore suggest that the script used in the Codex is likely to be an artificially generated cipher alphabet rather than an obscure kind of proto-Hungarian or Sanskrit-derived natural script embellished with additional symbols.

  12. SirHubert: you have to bear in mind that if the dating of the Rohonc Codex is broadly in line with its Venetian paper watermark (i.e. 1530s), then this is still 50 years before Timothy Bright’s “Characterie”, the first ‘modern’ shorthand. Even so, Bright’s system was clunky and not really suited to fast writing: so this was before modern shorthand had even begun to take shape, it would seem.

    Hence you have to be careful about placing it in a cipher camp or a shorthand camp, as these categoric pigeonholes were as-yet-unformed back then. Just saying!

  13. SirHubert on May 30, 2013 at 8:13 am said:

    Nick: absolutely. Looking back I think I misunderstood your suggestion that:

    ‘I think the most likely scenario for the Rohonc Codex is that it is formed of a combination of (specifically abbreviating / contracting) shorthand and non-polyalphabetic cryptography’

    I don’t think for one minute that the script was devised by a Hungarian ancestor of Pittman, Bright or anyone else. All I was suggesting was that quite a few of the commoner characters in the script are quite simple and could be written with only two or three penstrokes.

  14. SirHubert on May 30, 2013 at 8:36 am said:

    By the way, is there anything more concrete on the proposed ‘decipherment’ alluded to by the completely infallible Wikipedia:

    As a result of Gábor Tokai’s and Levente Zoltán Király’s recent cooperation, the code of the codex seems to have been broken. The major publication of the two authors is still forthcoming, however media reports of their first results have already appeared in print in Hungarian.[when?] According to Tokai and Király the script is a code system that does not indicate the inner structure of words, and the language of the text is most probably artificial, as optionally proposed by Benedek Láng. They claim that the codex contains the date 1593 CE as a probable reference to its writing.

    If (and it’s a big if!) Wikipedia is right about the proposed decipherment involving a ‘most probably artificial’ language which ‘does not indicate the inner structure of words’, I can hear alarm bells ringing. But who knows?

  15. SirHubert: let’s just say that I don’t (yet) share Lang’s optimism about Tokai’s and Király’s claimed decryption. It would be interesting to be proved wrong on this one… but I don’t currently think I will be. =:-o

  16. Lang is bit of a star – in the nicest way. I watch for his every paper.

    BUT – the point is that going through my first blog ‘Findings’ I see I noted a piece of papyrus found in England and dated to the 1stC. I have always thought the Rohonc looked like a work about the Egyptian embalmer’s religion, from about that time, but ‘Christianised-up’.

    Point is, that the signs on the same papyrus from England look not unlike one of those Lat.Pal. germ manuscripts – can’t recall which but an alchemical thing, I think.

    I think the papyrus may explain symbols in that manuscript and could (just possibly) relate to the Vms too, though my two cents in on Voynichese being written in the script of 1stC Greco-Bactria/Indo Greek.

    My two cents.

    Nick, shall I send you my photo of that papyrus?

    Diane

  17. SirHubert on September 17, 2013 at 4:32 pm said:

    Diane – do you have any images showing first century Bactiran script which demonstrate the similarity with Voynichese?

  18. SirHubert on September 17, 2013 at 4:34 pm said:

    *Bactrian (oh dear…)

  19. SirHubert,
    I try to stay away from discussions of the script, since it’s wasn’t part of my brief (as it were), but there’s a stratum in the imagery – both forms and stylistics – which point to that region and like Voynichese it has a script which is strongly influenced by Greek forms, while lacking any ‘X’ shape. In addition the same region was part of the Khwaresmian territory until the rise of the Mongols, the Khwresmian astronomers being the most influential in the Islamic empire. The details in the Vms which refer to older imagery from that region mostly occur in the astonomical folios, and these again connect well with figures in MS Sassoon 823 – another astronomical work, produced by Spanish Jews in the fourteenth century. (Panofsky attributed the VMS as a whole to Spanish, ‘or at least southern’ Jews and no-one was in a better position to know what was or wasn’t a product of Latin Christian tradition).

    There are more reasons for my concluding that some at least of the source-works informing the Vms had come from early Hellenistic works maintained in hither Asia till about the mid-12thC, but I’m sure that’s enough for here.

    As I’m sure you know, works revered as ‘holy writ’ (in any sense) tend also to have their script – even their language preserved long after it ceases to be a iiving language – vide Coptic, Aramaic, Syriac, Sansrit, Latin etc.

    As to examples of script – most of what we have is from the time of Kanishka, I believe and on coins. But in any case even if some such script had been used in the Vms as a ‘sacred’ or (by now) magical script, there’s no guarantee it isn’t being used to record some other.
    It wouldn’t be the only example, either – versions of Aramaic script, from the adjacent region – were used this way by people in medieval Islam.

    Greco-Bactrian, or an early Pahlevi script aren’t so much a theory of mine, but neither are they random guesses. I suppose I’d say they seem to me a good bet, given the evidence which is offered by details in the imagery.

  20. SirHubert on September 18, 2013 at 9:11 am said:

    Well, one preliminary observation would be that Greek-derived scripts run left-to-right while Pahlawi and related scripts go from right-to-left…

    If you’re interested in scripts from that part of the world, Tom Mallon-McCorgray’s site is really good for the numismatic evidence: http://www.grifterrec.com/coins/coins.html. I can’t recall seeing anything which looks much like the Voynich script, though.

  21. SirHubert,
    My five years – should have been three – is about finished. The script and language are not my area, and it’s perfectly possible in one way or another the present text was first devised for the manuscript we have.

    The same certainly isn’t true for the imagery, one stratum of which belongs to hither Asia/transoxiana c. 1stC AD.

    So does the Graeco-Bactrian script. etc.etc. etc.

    I don’t have those intuitive ‘theories’ then hunt evidence for them. To me is seems a back-to-front way to to research, but perfectly intelligent people have disagreed on the point.

    Maybe Voynichese will prove to be so, maybe not. I might lay a couple of dollars on it, if pressed, but hardly more. Don’t much care about that side..

  22. SirHubert,

    Pictures – I tried to post a couple of links, esp. to the Taq-e Bostan Pahlavi though a bit late (6thC AD).

    Alarmed Nick’s spam filter, tho’
    🙂

  23. Šuruppag on November 14, 2013 at 3:11 am said:

    “New signs are introduced for example when there is a new person in the text (Pilate, for example).”

    Perhaps these signs are logograms that relate to the new subject matter. e.g. Perhaps Pilate’s throne is denoted with a *throne* character.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Post navigation