This is to be read in front of any conjecture I made about the script. Does it convey uncertainty and mistrust with a hint of frustration? Good. Because that is how I feel whenever I get deeper in the code cracking business. There is a good reason why it resisted all efforts until now.
For a survey of the attempts made to decipher it, please see theWikipedia page.
It is not just an elaborate code (if it is), but it mocks you. Whenever you think you found some (common) sense in it, things change. Whenever you think you are finally familiar with the signs, the signs change. They appear and disappear like the blue tigers in the short story by Jorge Luis Borges (it starts with “A famous poem by Blake …”). And yet, it was deciphered as many times as I quit smoking last year. Some did it reading it from left to right, top to bottom, some from right to left, bottom to top, some found not only the key but also a completely new old language. But most of those who got involved with it probably felt like I do at times, that somebody, somewhere is laughing, so pleased to have fooled so many. MUHAHAHA.
And then I go back to the drawings, and they look so clumsy and candid…
One Saturday afternoon, in May 2010, I decided I will decode the manuscript. I had several printed pages from the Codex, the Latin-Romanian edition of St. Augustin’s Confessions (it seemed appropriate) and E. A. Poe’s Gold-Bug (for the cryptology know-how). What can I say? If you want to be successful in this business you will have to be far better equipped than that. So were Konstantin Jireček or Gyürk Ottó(who cracked the Gardony Geza diaries) and still failed.
After that I promised to keep away from the script. It was the reasonable thing to do, considering my (lack of) expertize in paleography or cryptography.
Despite all the sound arguments, eventually I couldn’t resist the temptation to give it a go. Every now and then.
Here is a summary of what I found:
- Fluxus – about the direction of writing
- Similarity to other alphabets
- Charting the signs
- Minor cracks in the code
- Concordance between text and drawings
- Some thoughts, or rather doubts, concerning the script
It is generally accepted that the script goes right to left, but there is no consensus on whether it goes top to bottom or bottom to top. Miklos Locsmandi suggested that it should be read from the bottom cover. M.K. Singh deciphered 24 pages reading it left to right while Viorica Enachiuc did it to the entire book reading it right to left, bottom to top, starting at page 1.
At first I did seriously consider the bottom to top variant, mainly because of the first page. The “word” at the bottom right looks like a good place to start the text.
Another argument is on the second page, where a glitch on the eighth line from the top seems to cause a large gap in the line above. This would have supported the Enachiuc flux.
But a closer look reveals a different situation. In two successive drawings the same string turns up but, as in the second drawing it did not fit in one row, the left half was written underneath.
Or, here, one of the often recurring strings,
which on page 100 shows up on the left side in one row, while on the right side is written on two rows:
I think this closes the case: the writing is right to left, top to bottom.
But the CR still has a surprise. Almost by accident, I found that the page order is somewhat particular. Bear with me for a moment.
Here is the right side of page 85, with a long string I highlighted:
Now, here is the left side of the the same page. At the bottom of the page I highlighted the signs that appear to be the exact match for the starting part of the string above.
But does it continue? Check with the top of the right side. No match. No match on the previous page either. But on the next page, 86, there it is, the rest of the string:
I am confident now that the script goes:
I verified it on several pages, and it appears to be consistent. Of course, for this I needed to find strings I was familiar with.
It was a very depressing finding. I was elated at first and went through page after page, “reading” them in the right order, expecting them to suddenly start making sense. They didn’t.
(I rechecked the Wikipedia page and Gyurk Otto found the same thing. Darn, I thought I am the only one who knows this. Well, at least I stand proof for him.)
However, it made a significant change in the order of the drawings. This is the reason why I labeled them with Right or Left instead of any other sort of numbering. Therefore, the very first drawing of the manuscript is on the right side of page 5, the evangelist Matthew with his symbol the angel, or winged man.
The second drawing is the one on the left, the Sacrifice of Isaac, maybe an illustration to the genealogy of Christ, with which Matthew’s Gospel starts. Could this be the first part of the Gospel according to Matthew?
2. Similarity to other alphabets
Each of those who studied the manuscript saw similarities with another writing system. Let’s see: Hungarian Runes, Cyrillic, Brahmi. Did anybody mention Hebrew? I guess I saw on a forum.
The first ones to see the manuscript noticed the characters alike to the Hungarian Runes.
Here they are, on my husband’s T-shirt. He got it from Magyarorszag (Hungary).
Find out what it says, and remember, right to left.
There are obvious similarities, but so are with Cyrillic, Brahmi, Old Turkic, Greek, Latin, Iberian, Hebrew, and some other.
This is an image of a minute Albanian manuscript, written in an alphabet invented probably in the 18th century named the Elbasan script, after the Albanian city. The pages are 10/7 cm and the alphabet is original, adapted to the Albanian language, showing some Slavic and Greek influences. Doesn’t it look a bit familiar? However, the Elbasan script has a “normal” amount of signs, and it is well understood, in fact on the site where I found it there is a full transliteration and translation.
Here are the links to some of the sources I used for comparing scripts:
An Introduction to Latin and Greek Paleography by E. M. Thompson
Warning: Should anyone indulge into comparing scripts, be aware. It is like reading a medical encyclopedia. You end up suffering from 42 unrelated and rare diseases.
Best proof, the following resemblance I found, thanks to Omniglot:
In case you have not recognized it, the top line is Klingon (language and alphabet).
And in the end, is there any benefit in comparing the Rohonc script to known writings? Are these similarities relevant in any way?
The following example would suggest that it’s a futile work. Here is the little 8, present mainly in the second half of the manuscript and also present in a number of alphabets:
From Edmund Fry’s Pantographia here are several phonetic values for the exact same character: in Chaldeean 17 is “S”, in Charlemagne 3 – “L”, in Coptic 1 – “I”, in English 19 – “G”, in Etruscan 1 – “PH” while in Etruscan 2 – “H” and Etruscan 3 – “F”. Take your pick. Not to mention that in cursive Glagolitic writing, “I” is a standing 8 (sort of) when it’s at the beginning of a word as seen in the Istrian Book of Boundaries (Istarski razvod).
Probably many of the signs from the Codex have twins in some alphabet. But are any of these similarities deliberate or mere accidents? The simplest explanation for the coincidental letters is that if anyone invents a writing system with so many signs, inevitably, some of them will have lookalikes in existing (or yet to be invented) alphabets.
Szabo Karoly, the same one who purported the counterfeit hypothesis, commented on the script “If I want them to be Cyrillic they are, if I don’t want, they aren’t”. Now, here is where I beg to differ.
Whereas resemblance to various alphabets might be fortuitous, two characters in the Rohonc script are not just alike Cyrillic letters but they are actually the Cyrillic letters mirrored, with their original phonetic value. I am talking about И (if you do not see it in your browser – the inverted N) standing for “I” and C standing for “S”. I’ll get to this a bit later in 4. Minor cracks in the code.
And then there are a large number of graphemes strikingly similar to Cyrillic letters,
or to Glagolitic letters.
First, notice the letters highlighted in yellow. They are both “M”s, showing that there was fairly common to use letters from different alphabets within the same text.
Then, take a look at the reddish sign, which according to the transliteration by Darko Zubrinic , stands for the sound śća. And here is its CR twin (also in red). The greenish Glagolitic letters from the Baska tablet have several possible similar signs in the CR, oriented the same way or mirrored, while the blue one is identical.
The Glagolitsa on page 169 is my favorite. It is very neat and very rare in the Codex. While on page 169 is carefully drawn, check it out on page 204. It looks like the same sign but traced by a hand familiar with the trajectory of the letter. A hand that couldn’t have become familiar with it from the CR, since it only appears twice (with a third maybe).
And an excerpt from another famous Glagolitic rounded script, Codex Zographensis, comes with the same letters (the purple sign, which is an A, looks like a plus (+) sign, very common in CR).
Another similarity, this time with the Hebrew alphabet is with the signs that clumsily imitate the Tetragrammaton. The intention is to make them look like the letters for YHWH and they mean YHWH (see in Divine Designators).
This pattern of “mock” letters, interspersed in the script raise serious doubts about the script as a whole.
3. Charting the signs
I tried to make an inventory of the “letters”, but gave up after the first 100. Signs, not pages.
Then I found on-line Nemethi Kalman’s book (which is no longer available but here is a copy I made of his list ). He identified 800 separate signs, however, some of them seem normal variations of the same sign.
They are awfully many for us, used to alphabets consisting of 25 to 30 signs. But cryptology books mention nomenclators, a system combining both code and cipher (go figure, these are not synonyms), consisting of substitutes for each letter of the alphabet plus special signs that could mean entire words, names or parts of words (the definition I found in Code Breaker – The History of Secret Communication, by Stephen Pincock and Mark Frary). You could, for instance, decipher the “heavily” encrypted text below if you knew the key (which is to replace each letter with the following one),
Xf uppl pvs gsjfoe # up uif ipvtf pg gbuifs $ ofbs @.
but you still need my nomenclator to find out who or what #, $ and @ stand for.
Such nomenclators, which appeared in the 14th century, grew from several coded words to several thousand by the 18th century, the heaviest ones being created in Russia. It is up to the cryptographer to select how many words will have a specific designator. 800 signs don’t seem so many anymore, do they?
Identifying and isolating the signs is a problem in itself.
I found Miklos Locsmandi’s article on his site and worked my way through it, struggling with my poor Hungarian, a better fit for reading goulash recipes than computer based analysis of an old manuscript, as he did. He recreated the signs on the computer but stopped at the first five pages. He also ranked the graphemes by frequency, identified the most frequent strings, determined that it is not gibberish, and that the language is unlikely to be Hungarian.
But charting the signs of only five pages does not cover the disturbing diversity of the “letters” in the Codex. And by the way, what qualifies as a “letter”? This puzzles me still. Some of the graphemes clearly stand on their own, other seem to aggregate to form more elaborate units.
When I counted and identified the signs for some statistics of my own, I regarded the little circle as one of the five letters in the string but as part of a complex sign in the next one.
Another difficulty in charting the signs is the careless handwriting. In many cases is difficult to tell the signs apart. Are these different “words” or the same “word” in a faulty transcription?
So here is an arbitrary classification:
d. pictograms which probably work like this
Certain signs look more graphic, or they stand out from the lines, these were the criteria I used in selecting them. In the table they are accompanied by the page number where they are located.
The first one (from page 44) appears opposite to the flogging scene and could be a whip. The second is obviously a cross. The ones on pages 84 and 181 could be the sun, while the ones on pages 142 and 163 could be the moon. Calling them black suns is so tempting, but I have examples suggesting that the darkening means lesser light, like the candle in this drawing.
The one on page 96, if it is not Kermit, could be the Hellmouth. No idea what the rest of them are.
e. neumes – the text might include indications on how to voice the words. The first combination from below might be accentuated or the sound might be pronounced longer when written this way:
I also had the confuse and confusing thought (trying to find a justification for the diversity of signs) that maybe musical notations are mixed among letters. Since the waving dots on page 212 look pretty much as some sort of chanting indication, this idea could be put to rest. And when I think about all the neumes, Hebrew and Slavonic cantillation marks I stared at (I can’t say I studied).
This is how musical notations look in a beautiful illuminated manuscript
And here they are in a Russian manuscript
The book written by Viorica Enachiuc (Rohonczi Codex – Dechiffrement, transcription et traduction) has a CD attached with the recording of the song reconstructed based on the fragment from page 212 of the CR.
Unfortunately, I did not find it, but here is the next best thing: a song recreated from a knife.
f. starters and enders – Miklos Locsmandi proposed as a sentence starter the quite frequent
I am not convinced but I don’t have a better idea.
Two slash lines at the end of the row turn up throughout the manuscript, but not at the end of each row and not to mark the end of a sentence. The already familiar string below is separated
Or the longer repeated string on page 93.
If it were a sentence or paragraph ender, it shouldn’t be placed in two different positions of the same string. It is more likely a void filler.
g. numbers – as identified by Gyurk Otto – see Locsmandi’s article, page 15.
h. firlefantz – the signs of the manuscript usually fit between two parallel lines, with some of them emerging above the upper line by a little, but never descending under the bottom line.
There are some singular signs though, which are unexpectedly exploding in the interlinear space, like the gesture of a twitching hand. Here are some of these firlefantz.
i. signature – yes, the work is signed, on page 222, not quite at the end, there are three more pages after the signature, a colophon maybe?
In the enormous diversity of the signs which appear and disappear, I tried to see which ones are constant. To this purpose I took ten random pages (with no drawings) and listed all the signs for the left and the right side of each. I compared the results for each page and then compared the ten short lists resulted. This is the final product.
Six characters, one of them the omnipresent inverted c. Apart of which remain five signs, like in the basic five vowels? And, as we will see, one of them is a proven vowel.
4. Minor cracks in the code
The obvious place to start the code cracking (or cipher deciphering) was the standardized inscription (Titulus Crucis), always the same (or so I thought), the acronym that appears on crucifixes: INRI (Iesus Nazarenus Rex Iudaeorum – Jesus from Nazareth, King of the Jews), that would provide three certain letters, since the first and the last should be the same.
They are not. The acronym appears to be, as I composed it with larger signs taken from the text:
How can this be? If this script makes any sense, and the Titulus Crucis consists of four different letters, I needed to find a plausible explanation. Actually, I found two:
a. In Eastern Christian (Orthodox) countries, the Titulus might use the initials of the text in Greek:
which, sometimes is augmented to:
INBK – Iésous o Nazóraios o Basileus tóu Kosmou (king of the World)
INBD – Iésous o Nazóraios o Basileus tis Doxis (king of Glory)
The acronym for both will have four different letters.
b. What if it comes from the translation into some vernacular?
Hebrew – Yeshu’a HaNatserat Melech HaYehudim
Polish – Jezus Nazarejczyk Król Żydowski
Bulgarian -Исус Назаретски, Цар Юдейски
Croatian -Isus Nazarećanin, kralj židovski
Lithuanian – Jėzus Nazarietis Žydų Karalius
Hungarian -Názáreti/Nazoreus/Nazireus Jézus, a zsidók/júdeaiak királya
Romanian – Isus din Nazareth, Regele Iudeilor/Jidovilor/Evreilor
In Hebrew the first and last word start with the same letter, I think. The rest of the languages would provide four different initials. I must admit that I never saw a Titulus written in this way before. But then again, I have never really looked before.
All these variants (except for Hungarian) support the idea that the first letter is “I” or “J”. The second should be an “N”, but the letters in the drawing are nor clearly readable.
The other drawing that can give some hints is also partly readable and confirms only the first sign.
This one letter was the hint that helped me in recognize the Christogram I wrote about in the Divine Designators. Another identifiable inscription is on the drawing with the Sacrifice of Isaac. It stands most likely for Abraham, Isaac or Sacrifice.
It appears again on the Last Judgment drawing above the head of one of the companions of Jesus
It should be Abraham. He was the first patriarch, the first one to make a covenant with God, the one who started worshiping a single god and made a major difference in the life of his people by doing so. His name appears 74 times in the New Testament, in 27 different places, often in phrases like “the people of Abraham” or “the God of Abraham”. In the Codex, the string appears 48 times: on pages 5 – 5x (1 drawing), 6 – 4x, 27 – 3x, 62 – 6x, 79 – 1x (drawing), 85 – 1x, 90 – 9x, 91 – 2x, 106 – 5x, 107 – 1x, 209 – 5x, 215 – 6x (and 2 maybes).
Now, take a look at the character on the left. He is probably Moses, the other great Hebrew who made a covenant with God and who lead his people to the promised land. (I explained in the chapter The Story in the Pictures, pages 69 through 91). He is also identified by a string, the same as the one smudged in the drawing with the (three!) tables of the law.
And there we have a second inverted Cyrillic letter: the c standing for “S”. (Of course, I noticed the resemblance of the first letter with the Hebrew “Shin”).
Moisi or Moisie is not uncommon in Slavic or Romanian inscriptions of the name.I was a bit disappointed by the outcome for the little triangle. Based on its high frequency, I was convinced that it stands for “E”.These are my minor cracks in the code/cipher, here is the seed of an alphabet/nomenclator
The signs for I and Ij might be reversed.For the graphemes signifying Jesus Christ and YHWH, please see my post about the Divine Designators. And I forgot about the birdie that means Jerusalem.
Back to the “mock” letters theory. I was wondering what might have inspired the CR author in “designing” the very frequent triangle letter. So, here is an inscription with the name of Moses (Moyses) in a Danish church painting.
Just a thought.
5. Concordance between text and drawings There seems to be a match between several drawings and the surrounding text.The script above the Sacrifice of Isaac drawing shows up on the same page four more times
and five times on the next page
In the text opposite to the flogging scene, the pictograms suggest a scourge. This particular pictogram is nowhere else to be seen in the Codex.
In three drawings, the cities in the distance are identified by certain signs. You might think that hovering over the city on the right is a bird, but is the code sign for Jerusalem. The “birdie” occurs in the script in several places (pages 11, 12, 13, 22, 28 …) just not opposite to this image.
Neither to the next one.
But there is a flock of them in the third (and last) place where it appears above a city (or building).
6. Some thoughts, or rather doubts, concerning the script
These are some of my observations pertaining the script. There are so many more issues, like the special page with underlines,
or the insane number of different Z’s.
Despite the fact that I found the equivalents for several signs, I am skeptical about the chances to decode the entire Codex. The more I look into it, the less it looks like a meaningful script. A large part of the characters change every several pages. Not to mention the paradox of the insane diversity of characters associated with the insane number of repetitions of one sign or one string of signs within the same page (see also in What Makes the Puzzler Sore).
The script is believed to be a complicated ciphertext. Yet, the few letters we know the meaning thereof are ordinary Cyrillic letters (for I, S, O), mock Latin letters (the triangle) or mock Hebrew letters (the Tetragrammaton). And they are not mere decoys, like in the Copiale cipher, they have a central place in the manuscript. I don’t know much about ciphers, but this looks a bit suspicious.
However, if this script has a meaning, I find it hard to believe that the creator of the Codex is also the creator of the cipher. We have on one hand a notebook with painfully unsightly drawings, clumsy handwriting and unskilled binding work. On the other hand we have a complicated and sophisticated cipher. How likely is it to be produced by the same person?
Therefore, if it is not the product of a feeble mind, I think the script might be copied from another source. And if it is a copy, we cannot know how accurate it is. We know so far that the scribe made corrections by either deleting or adding signs.
If you have a cipher or code, or a combination of both, of such complexity, it is serious work to transliterate an ordinary text in ciphertext. But how about reading it? Imagine that you have the cypher, the code, the meaning of each and every one of the 800 signs. It would still be a painstaking effort to read the text, hardly a lecture to be made during the everyday prayer and impossible in front of an audience.
Maybe the Codex was not meant to be read the way reading is understood, letter after letter, sign after sign. Maybe the Divine Designators work as benchmarks (as Christopher Tuckett implies about nomina sacra), catching and guiding the eye of the reader through the text he already knows so well. The Designators, in conjunction with the drawings and the pictograms may work rather as a map of, than a thread through the labyrinthine script.
In my opinion, the Codex wasn’t meant as much to be read as it was to be written. The act of affixing the signs to the paper transformed a bunch of ordinary leaves into an extraordinary object with meaning but also with power, empowering its creator in the process.