The Rohonc Codex
Partially explainedSummary
Nearly four hundred and fifty pages, in a script nobody recognises, with almost ninety illustrations, many of them plainly religious: crosses, crescents, scenes that look like the Gospels.
It has been called a hoax for most of its existence, and by very good scholars, and they may well be right.
And in 2018 two Hungarian researchers published a substantial partial decipherment, and appear to have read parts of it.
Both of those things can be true at once, and that is the interesting part of this record.
What is documented
The object. A codex of roughly 448 pages, densely written, with some 87 illustrations. It is written right to left.
The script. Counts of distinct symbols range from some 150, the figure given by Benedek Láng, to 792 in an 1889 count, which is far too many for an alphabet and points instead toward a syllabary or a partly logographic system.
The illustrations. They are unmistakably religious, and they are unmistakably Christian in their principal imagery, though crescents also appear. They show what look like scenes of teaching, of crucifixion, of battle.
The provenance. It appeared in the library of Count Gusztáv Batthyány, who donated his collection, including the codex, to the Hungarian Academy of Sciences in 1838.
The paper. It bears a Venetian watermark dating to the 1530s or 1540s.
This dates the paper. It does not date the writing. Old paper can be written on at any later time, and a forger who wanted his work to seem old would seek out exactly this.
The forgery hypothesis, and why it is serious. The codex surfaced in the era of Sámuel Literáti Nemes, a Hungarian antiquarian who is a documented and prolific forger of historical documents.
Nemes manufactured false antiquities that deceived leading scholars of his day, and Károly Szabó attributed the codex to him in 1866. No documentary evidence connects Nemes to this codex. Benedek Láng argues that it does not fit the pattern of his known forgeries and that it is not a hoax.
The 2018 decipherment. Levente Zoltán Király and Gábor Tokai published a substantial partial decipherment in Cryptologia.
They identified the numbering system. They identified structural features. They argue that it is a Catholic devotional reader, mostly paraphrasing New Testament passages, chiefly from the Gospels, written in a code system rather than a simple cipher, that its underlying language is probably an artificial one, and that it contains the date 1593.
They have read parts of it. Their reading is not universally accepted, and it is a serious piece of scholarship rather than another claimed solution of the kind this archive collects.
Leading explanations
It is a nineteenth-century forgery. Long the majority view. A known forger was active when it surfaced, though nothing links him to it directly. The paper is old but the writing need not be, and no comparable script exists anywhere.
It is a genuine encoded religious text. Supported, in part, by the Király and Tokai work, which finds structure and content rather than noise.
It is both, and this is the point the registry wants to make.
These are not competing hypotheses. They are compatible, and the tendency to treat them as alternatives is an error that runs through a great deal of writing about this manuscript.
A forger can produce a text that says something. Indeed a good forger must: a fake that is gibberish is a fake that will not survive examination, and a fake with real content, real structure, a real numbering system and a coherent religious argument is a far better fake and a far more valuable one.
If Király and Tokai have read the codex, that establishes that it is not nonsense. It does not establish that it is old.
The two questions are separate. Does it mean anything is a question about the text. Who wrote it and when is a question about the object, and it is answered by paper, ink, provenance and history, not by decipherment.
What the popular version gets wrong
"It has been deciphered, so it is genuine." Reading a document tells you what it says. It does not tell you when it was written. A nineteenth-century forgery of a medieval religious text would, if it were any good, be readable.
"It is a forgery, so there is nothing in it." If the 2018 work stands, there is a great deal in it: an encoded religious text of some length, with an internal system. That is worth reading whether it was written in 1540 or 1830.
"It is another Voynich." It is not, and the difference is instructive. The Voynich has resisted every attempt at reading, and its language, if it has one, is unknown. The Rohonc Codex has yielded structure, a numbering system, and an identifiable subject, and the live dispute about it is one of provenance rather than of meaning.
Current status
Partially explained. Substantial progress has been made on the script and the content, and the authenticity question is open and may remain so.
The registry keeps this record for the distinction it forces, which is one the archive needs and does not often get to state so cleanly.
Is it real? and does it say something? are different questions, and a great deal of confusion in this field comes from treating them as one.
Sources
- Király, L. Z. and Tokai, G. (2018). "Cracking the code of the Rohonc Codex." Cryptologia 42(4), 285 to 315.
- Hungarian Academy of Sciences, catalogue record and provenance of the codex, donated 1838 by Count Gusztáv Batthyány.
- Analysis of the paper and its Venetian watermark, dated to the 1530s or 1540s.
- Documentary record of the forgeries of Sámuel Literáti Nemes, and Károly Szabó's attribution of the codex to him (1866).
- Láng, B. (2010). "Why don't we decipher an outdated cipher system? The Codex of Rohonc." Cryptologia 34, 115 to 144.
- Láng, B. (2021). The Rohonc Code: Tracing a Historical Riddle. Penn State University Press.
- Published assessments of the codex's script, sign inventory, and illustrations.
Last reviewed: September 2026. Records are provisional. Where the evidence changes, the entry changes. Found an error? Tell us.