When I was writing “The Curse of the Voynich” a decade ago, my friend Philip Neal very kindly translated Cicco Simonetta’s Treatise on Decipherment (BNF Fonds Italien 1595 ff. 441r-442r) into English for me. This was a huge help, because this is one of the few accounts of fifteenth century code-breaking we have.

With Philip’s permission, I posted this onto the Cipher Foundation website earlier this year: it’s a straight-down-the-line, properly accurate translation.

But all the same, the source document is – as indeed is most writing of the period – somewhat verbose. So I thought it would be useful to extract the core details of what Simonetta’s document is describing and to then re-present them in a more modern idiom.

So here’s a very much stripped-down modern version. Enjoy!

Cicco Simonetta’s Treatise on Decipherment

(1) If the words in a ciphertext have five or less different word endings, the plaintext is probably Italian (or if not, then it’s Latin). Alternatively, look at all the single-letter words: Latin normally has only one kind (‘a’), but Italian tends to have more.
(2) If the ciphertext has many two- or three-letter words, the plaintext is probably Italian.
(3) If the plaintext is Italian, then you already know what letters are vowels (because they’re the last letters of words). If one of these often appears as a single-letter word, it’s probably ‘e’.
(4) Two letter words in Italian very often begin with ‘l’: lo / la / li / le.
(5) The most common three letter word in Italian is ‘che’.
(6) However, if the plaintext is Latin, the letters that appear at the end of words are vowels, s, m, or t. (Apart from ab, ad, and quod, which are very common).
(7) In Latin plaintexts, single-letter words are normally ‘a’ (but possibly e, i, or o).
(8) In Latin letters, the most common two-letter words are et ut ad si me te and se. Less common A fuller list of two letter words would be: ab ac ad an and at; da de and do; ea ei eo et ex and es; he hi id ii in ir is and it; me mi na ne and ni; ob os re se and si; tu te ue ui and ut.
(9) Latin three-letter words where the first letter is the same as the third are: ala, ama, ara, ede, eme, ere, ehe, ixi, iui.
(10) Any Latin letter that appears three times in a row within a word is ‘u’, as in ‘uvula’. [Though Simonetta writes ‘mula’]
(11) Latin letters that appear doubled, and particularly in four-letter words, are probably ‘ll’ or ‘ss’, e.g. esse and ille.
(12) A final rule that holds true for both Italian and Latin: if you see a letter that is always followed by only one possible other letter, then this is ‘q’ followed by ‘u’: moreover, the letter following the ‘u’ will be a vowel.

However, these codebreaking rules can be defeated in many ways, e.g. using a mix of Italian and Latin; inserting nulls, particularly into one-, two-, or three-letter words; by using a mixture of two completely different cipher alphabets; and by using an extra cipher for ‘qu’.

13 thoughts on “Cicco Simonetta’s Treatise on Decipherment, in modern English…

  1. Ah, the simple days of monoalphabetic ciphers.

    BTW what does he mean by a “mixture of two” ciphers? A homophonic cipher?

  2. James R. Pannozzi on December 18, 2016 at 8:12 am said:

    Quite interesting….many thanks for sharing !!

  3. SirHubert on December 19, 2016 at 9:24 am said:

    Ken: he means that you use two different monoalphabetic cipher alphabets simultaneously, so that each plaintext letter has two possible equivalents. The Latin reads cum duobus alphabetis zifrarum omnino diversis, which translates as “with two cipher alphabets which are different in all respects.”

  4. Nick,
    It seems from your summary that Simonetta only meant to indicate a way to tell if the plaintext was either Latin, or Italian.

    Does he mention any others? Did he know any others? Would the linguists like to comment on whether his statements which run “… if… [then] .. it’s Italian” are only true relative to Latin or are absolute?


  5. SirHubert: thanks very much, correct in every respect. 🙂

  6. D.: that’s exactly correct. Simonetta was for many years head of the cipher department in the Sforza-era Milanese administration, where almost all of the ciphertexts that arrived were in Latin or Tuscan: and so that forms the context in which this treatise should be interpreted.

  7. Thomas F. Spande on December 20, 2016 at 11:17 pm said:

    Dear all, Hats off to Nick for a useful Latin-Italian comparison.

    Can we push this further by noting that Latin does not use definite or indefinite articles but Italian does? If the same glyph pairs appear often before a multiglyph expression (assume that some are nouns), then maybe we are looking at Italian? If the single “e” (Italian for “and”) appears at all, then Italian is involved along with Latin, as I feel certain that “et” appears frequently in the guise of “89”.

    I plan to chug along looking for more Latin pronouns as I think many are in the VM but disguised by ending in that “inverse gamma”. I think some of the double-stemmed gallows may turn out to be what they look like “ll” and appear in the pronouns “that” and “which”.

    Cheers, Tom

  8. Josef Zlatoděj Prof. on December 21, 2016 at 1:02 pm said:

    Dear, Bee and ant.
    Your Christmas, and so will give you a gift. 🙂

    Manuscript ( MS 408 ) no herbarium.
    You see the plant. Ant see the plant. Bee see the plant.
    It is bad.
    Search in manuscript flowers and so is wrong.

    Author manuscript you the pulling the nose. ( Tugging at his beard ).
    Search in manuscript plants is unnecessary work.
    I know that, I you will clestroy this joy.

    You see the plant. Root, leaf and flower. And strive for identification. It is the goal of the author. He shows you plant. And you can see it.
    ( This is called image suggestion. This is called visual suggestion . )

    As I wrote, so this is called the influence of mind.
    Manuscript no herbarium.

    Cheers Josef Zlatoděj Prof.

  9. It would be interesting to know the ratio of latin to Italian in the Voynich. It will tell us something about
    the literacy of the author. It is possible he was doing a transliteration of older texts like Tibetan or Mesopotamian.

  10. Thomas F. Spande on December 22, 2016 at 3:25 pm said:

    Professor, I am overwhelmed at your generous Christmas gift! In exchange, the bee in me, won’t sting you with why you haven’t worked “Old Czech” into the Latin-Italian discussion above?

    Cheers, Tom

  11. Josef Zlatoděj Prof. on December 22, 2016 at 5:52 pm said:

    Tom bee. I’m glad you’re amazed. Maybe it amazed even an ant. I’m writing this here’s why. That the information is important. Many bees and ants, scientists and linguists are trying to identify plants. So I try to help them.
    The manuscript has with Latin and Italien nothing.

    Merry Christmas. Josef Zlatoděj Prof.

  12. Tom: it’s a long way from A and B to Z. 🙂

  13. Josef Zlatoděj Prof. on December 22, 2016 at 7:19 pm said:

    Nick : Long way ?? 🙂
    I already wrote several times. So that characters are published manuscript. And I’ll write their significance. Alphabeth Eva is bad.

    Ten years looking for the meaning of characters is wrong.
    Write a letter and I will write their significance.

Leave a Reply

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

Post navigation