I’ve had a nice email from Marius-Adrian Oancea, asking me if I would look at his interesting Rohonc Codex site. While working for the EU in Fiji between 2009 and 2011 (it’s a dirty job, but someone’s got to do it, I suppose), Marius-Adrian filled his spare time making notes on the Rohonc Codex, and has now written them up in a series of web-pages.
For example, he sets out some persuasive arguments that the text is written from right-to-left (along each line), from top-to-bottom (for lines within each page) and from back-to-front (from page to page).
What is interesting about this is that because I don’t currently believe that the folios (folded pairs of pages) have ended up in the correct order (simply because the chronology of the Biblical pictures seems strangely out of order), it may be possible to identify some candidate facing pages based on matching incomplete half-phrases on the bottom of the right-hand page with incomplete half-phrases on the top of the left-hand page. (I’m not sure that anyone has done this yet.)
He has also found a pair of intriguing repetitive word-skipping sequences on 162L and 162R, where the two instances are each padded out with a different filler “letter” / “word”.
However, I think there is a far simpler explanation for this problematic text than his conclusion that “existing paragraphs were repeated or repeated with insertions […] to create a larger book without making the effort to produce original, non-repetitive text“. What if these inserted filler shapes are both cryptographic nulls? They certainly don’t seem likely to be meaningful, so perhaps they are purely meaningless (nulls): and the fact that the phrase without the nulls also appears on 162L also seems to point that way. It would be interesting to revisit the stats if those two (possibly) null characters were excised from the text stream.
Alternatively, the apparent presence of nulls in the Rohonc Codex’s text might instead mean that the author was trying to duplicate the page structure of an existing manuscript, and that those pages didn’t originally have much text on (and hence needed padding). We’re still not necessarily looking at an enciphered document: we still have no definitive proof of that, but the presence of nulls would seem to be a very strong indication of the presence of encipherment (to my eyes, at least).
Similarly, on the same facing page pair (162L/162R), the author seems to repeat a block of text: though I should also point out that a straightforward explanation for this could be that the encipherer lost their position in the text and ended up enciphering the same block twice in a row. It’s certainly easy to do if you’re not hugely experienced at enciphering.
All in all, I’m not (yet) convinced that “The Codex is written in Hungarian, or at least transliterates words in Hungarian, using a version of the rovásírás (Old Hungarian Alphabet) also known as székely rovásírás or székely-magyar rovás.”. Up to that point in his pages I was feeling quite comfortable with his overall argument, but decomposing symbols into pieces and then anagramming them to get transliterated old Hungarian is a bit more than I was personally able to chew on without choking. Even so, there are plenty of tasty things on Marius-Adrian’s site to get your teeth into. 🙂