Regular Cipher Mysteries readers will know that I’m pretty good at digging historical things up, at shining lights under long-unmoved archival rocks. Well… my challenge this week was to find some mid-Quattrocento Milanese enciphered letters, and though I’ve possibly got most of the way to an answer, I’ve ended up a bit stuck, and would really appreciate some help from all you good people!

The starting point was that I was sure that the Archivio di Stato in Milan contains a vast number of documents from the period I’m most interested in (Milan’s Francesco Sforza era, i.e. 1450-1465), so that ought to be the first place to look for these. But (as is normally the case) relevant manuscript catalogues are few and far between online, so I initially drew a blank.

Then I (somewhat luckily) stumbled across a 1995 book online called “Fifteenth-century Dance and Music: Treatises and Music” by A. William Smith. Page 6 of its “Fifteenth-century Italian Dance Sources” chapter mentions a letter: “26 July Archivio di Stato milanese. Potenze estere. Napoli 1455. in cifre from Albrico Maletta, Sforza ambassador in Napoli to Duke Francesco Sforza in Milano.” Interestingly, f20v of the famous Tranchedino cipher ledger is marked as “Cum Francisco Maleta” (though this is sandwiched between a 1458 cipher and a 1459 cipher, so might well have been entered into the ledger later than 1455): all the same, it would be interesting to compare the two. But how to find the manuscript reference for this?

The first thing to note is that “Potenze Estere” is actually the name of a large set of documents within the Milanese Carteggio Sforzesco archive. Obviously, I then searched like crazy for (I’d guess a scan of a 19th century) inventory of this, but without any luck. So where next?

Then I remembered Aloysius Meister’s “Die Anfaege Der Modernen Diplomatischen Geheimschrift” (1902): p.30 contains a (surprisingly complex, I think) Milanese cipher key and nomenclator dated 14th March 1448, with the reference “Mailand, Staatsarchiv. Pot. Est. Cifre Fasc. 2 Nr 5.” There’s also a 1483 cipher (p.31) noted as “l. c. Fasc. 1 Nr 15” (interestingly, this contains a “4o” composite character (for ‘z’) but with the ‘o’ attached to the downstroke of the ‘4’), and a 1530 cipher key (p.32) listed as “l. c. Fasc. 4 Nr. 53 Grofs 4o”.

(I should add that Meister 1902 also lists ciphers for Modena, including one [p.35] dated 23rd June 1435 “In Milano” which fascinatingly contains “4” for ‘Q’ and “4o” for “Qua”. [“Canc. duc. Arch. Proprio Mappe II. Nr 1.”]: and for Florence, he lists the Cifra di Galiotto Fibindacci da Ricasoli 1424, which similarly uses “4o” for “Q” [p.50])

So there you have it: it seems that the Carteggio Sforzesco’s Potenze Estere archive contains several specific bundles of cipher documents (“Cifre Fasc[iculus/-i]”) that sounds like what I’m looking for. But then again, Meister was writing over a century ago and much may well have changed there: specifically, here’s a link to the best listing I could find for the pre-1535 part of the Potenze Estere archive, but note that there is no obvious cipher bundle or subset to be seen. And that would seem to be the end of the line – though I’d expect the 1455 letter from Naples listed by Smith is probably filed in the Napoli section of the Pot. Est. archive (which is more or less entirely arranged geographically).

At the end of all that, I don’t know whether I’m really close or really far away. Are the cipher bundles Meister referred to still in the Potenze Estere, and what do they contain? Or have they been moved, split, stolen or lost at some point during the last century? Regardless, where do I need to go to see them and what should I ask for? Any pointers you can turn up to help me answer these questions would be much appreciated! [Please leave comments on the page below, or email me at the normal address]. Thanks!

Update: I subsequently found a more detailed listing on p.927 of this sizeable inventory: it says that the Atti Ducali (1392-1535) section of the Archivio Sforzesco contains “Cancelleria segreta 1450-1535, scatole 11. Raccolta di documenti relativi all’attività quotidiana della cancelleria: sommari, cifrari, occorrenze (carta, inchiostro ecc.), archivio, documenti relativi alla biblioteca del castello di Pavia.” So perhaps the cipher documents Meister saw were later moved over to this Atti Ducali section?

Alternatively, the Carteggio Sforzesco’s Potenze Sovrane archive also holds a section marked “Cancelleria segreta – Chiavi e cifrari (scatt. 1591, 1597 – 1598)”, which is what Lidia Cerioni relied on for her book “La diplomazia sforzesca”, and might instead be what I’m looking for (it’s hard to tell). Oh, and just a few bundles away, the same archive has the intriguing-sounding scat. 1569: “Miscellanea, astrologia, occultismo, superstizione etc.” Really, what historian of mysteries could resist sneaking a peek? 😉

158 thoughts on “Milanese enciphered letters, call for help…

  1. Dennis on June 30, 2011 at 4:49 am said:

    Hi Nick! Just a comment in general. Have you ever contacted David Kahn? He has a web site, and it’s possible he’s studied some of this sort of thing. And what about contacting Cryptologia journal for possible contacts?

    I have a few FB friends in Italy, most likely the same ones as you, who might conceivably be helpful. it’s good to have correspondents on the spot, or as nearly as possible, like news services do.

  2. Not exactly enciphered but I think relevant to the Vms
    A Yemeni tax list of the Mamluk era found in Milan.

    Cahen, Claude and Serjeant, R.B., ‘A Fiscal Survey of the Medieval Yemen Notes ..(etc.)’ Arabica , T. 4, Fasc. 1 (Jan., 1957), pp. 23-33.
    -available through JSTOR –

  3. Laurent on July 7, 2012 at 8:59 am said:

    I see I’m not the only one studying milanese ciphers. I’m writing a thesis for my Master 2 in italian language and litterature about the work of Lydia Cerioni “La diplomazia sforzesca…” and i need some information on the author, who is she? where is she working? etc. If you have any information to share about these books and about the milanese dutchy in the 15th century in the Sforza’s court in general i’ll take it, I can share information too if you need but i’m in the beginning of my thesis…I’m french by the way…

  4. bdid1dr on July 13, 2012 at 11:44 pm said:

    I may be completely off-track, but: several weeks ago I came across a piece of art-work that portrayed one of the Sforza family Popes. In the background of that portrait painting was anoither painting that supposedly portrayed that particular Pope as being the hero that saved the Italian naval ships in the battle that took place in the harbor (Venice?Milan?).

    I hope I’m not leading you all astray!

  5. bdid1dr on July 13, 2012 at 11:52 pm said:

    Oh, I just now recall the name of the website: Rome Art Lover

    Don’t be put off by the rather “commercial” venture of the website’s tour-guide. He does a fair dialogue for each of his presentations.

  6. The subject of Aloys Meister has cropped up in the Voynich mailing list.

    Hope you don’t mind but I’ve referred everyone to this post.


  7. Mark Knowles on October 21, 2017 at 2:29 pm said:

    Nick: Another interesting blog post. I am only just starting to get to grips with the daunting task of chasing up the information that I am looking for or specifically one or two nuggets of gold. Any assistance if you have any time based on your experience would be of great value.

  8. Mark Knowles on October 31, 2017 at 11:42 pm said:

    I was googling for Milanese ciphers from the era of Fillipo Maria Visconti. I could not easily find anything, although I came back to a review I had seen before, on a different search, by Doctor Catherine Fletcher of “Carteggi degli oratori sforzeschi alla corte pontificia I: Niccolo V (27 febbraio 1447–30 aprile 1452). Gianluca Battioni,ed”. I quote from this:

    “Beginning with the death of Pope Eugenius IV in February 1447, the letters recount the process through which Sforza and Nicholas maneuvered toward their respective political triumphs of 1450: the former’s entry into Milan and the latter’s successful jubilee. These were early days for resident diplomacy in Rome, which was still not officially tolerated, though among Sforza’s representatives, Nicodemo Tranchedini (1413–81) soon came to take a central role. As with any series of diplomatic letters, this carteggio gives us a sense of the ruler’s priorities. Early on, the concern is with establishing Sforza’s position, and there is a wonderful juxtaposition of the envoy of Marcolino Barbavara’s part-ciphered dispatch suggesting to Sforza ways of encouraging “these priests” to contribute to his expenses with Sforza’s letter to the College of Cardinals, describing his desire for the health, well-being, peace, quiet, defense, and growth of the Papal States. Other correspondence underlines the importance of secrecy. One entry in a diplomatic memorandum notes that certain matters are not to be disclosed to the vicechancellor. Filippo Maria Visconti, sending a cipher to his agent in Rome in a letter a couple of months before his death, reassures the man that he will not be tasked with anything that might weigh on his conscience.”

    In this instance the letter from Filippo Maria Visconti sounds interesting to me as part of an attempt to fill in the gaps in the Milanese cipher record.

  9. Mark Knowles on November 1, 2017 at 9:01 am said:

    Nick: I have downloaded Meister. As you say there are ciphers from 1424 and 1435 which use the 4o character.

  10. Mark Knowles on November 2, 2017 at 4:45 pm said:

    Nick: Unsurprisingly it would seem that given the destruction of the Milanese records most examples of Milanese ciphers prior to 1450 are best found in amongst correspondence with other states and so a good place to look is in amongst the archives of letters in Florence, Venice, Vatican as well as others. I think some other people have carried out some of this research and hopefully they should be able to give me some idea of their findings.

  11. Mark Knowles on November 17, 2017 at 4:05 am said:

    Nick: One simple a probably very obvious thing->

    What was the procedure in practice for cipher communication?

    Presumably in person or by letter the rules of a particular cipher were agreed upon for subsequent communication or would the cipher be different for each communication and so obviously transmitted separately each time? I would imagine that there would be different cipher rules dependent on who was being communicated with, so one cipher for communicatiion witb Venice and another for communication with Savoy. Was the cipher the same both ways? i.e. messages to Savoy used the same cipher as messages from Savoy. I would guess that every say 2 years the cipher was changed. I would assume the message would have been hand transmitted by a different trusted individual each time. How would generally a messsge be intercepted, a disloyal member of staff or the physical capture of the messenger? Before the days that there was a cipher secretary would one secretary, in practice be responcible for the production of all ciphers amongst their other duties? Or would this work be spread over many secretaries for each different state one was communicating with? To what extent were cipher used for communication within a state?

    There are probably other questions to ask, but these are the first that occur to me.

  12. @Mark
    Is that a general question? Because with the VM it would not help.
    Since the cipher is useless without a book, and without a cipher, you can not read it.
    This is different from a normal conversation where information tiled back and forth

  13. Mark Knowles on November 17, 2017 at 10:53 am said:

    Peter: It is a general question which only relates to the Voynich indirectly. I am talking about diplomatic cipher communication not the Voynich itself. However it relates to my Voynich research.

  14. Mark Knowles on November 21, 2017 at 11:20 am said:

    Nick: When it comes to enciphered letters do we tend to know who wrote each letter or normally is it the case that it is written on behalf of someone without any clear about who prepared or implemented the cipher.

    Do we see->

    Your Sincercely,

    John Smith on behalf of Filippo Maria Visconti


    Filippo Maria Visconti

    So we may or may not know who had the cipher expertise.

  15. Mark: as far as I know, we almost never see the name of anyone who wrote a letter for someone else, whether enciphered or not. So this whole area remains murky, with only occasional flickers of light. 🙁

  16. Mark Knowles on December 7, 2017 at 4:18 pm said:

    Nick: What do you think the fate was of intercepted letters? So supposing a Milanese enciphered letter was captured on route to its destination then would we find the letter in the archives of the intercepting power? I wonder if this is somewhere else to look for Visconti era ciphers.

  17. Mark Knowles on December 8, 2017 at 7:53 pm said:

    Nick: The “4o” character is regularly mentioned for understandable reasons. In general to what extent are characters reused in new cipher alphabets? And to what extent are completely new characters invented each time? The reason for my question is that I am interested in seeing whether it is possible to find another distinct Voynich character in a non-Voynich text.
    I am in the process of trying to track down various Milanese enciphered letters from the letters of Marcolino Barbavara to the earlier intercepted letters of Guarniero Castiglione to the Duchy of Milan dated 1432.
    My hope is that with examples like these and with luck earlier Milanese enciphered correspondence I might find some identifiable character of feature which parallels something we see in the Voynich.
    It is true that the author could have created all the significant characters excluding the “4o” from his/her own imagination without drawing on other known characters.
    So I wonder if inventing completely new cipher alphabets is a standard thing to do or if it is very unusual. Obviously the Voynich falls into the category of very usual, but I like to believe it has some similarities to ciphers of the same time and place even if only superficial rather than technical. Do tell me what you think?

  18. Mark Knowles on December 9, 2017 at 6:39 am said:

    Nick: I wonder about the evolution of cipher alphabets. So for example a character in one alphabet may resemble a character in an earlier alphabet even though it is not the same. Obviously I am trying to establish links betweem the symbols of one cipher alphabet and another and thereby looking to see if I can establish a link between the Voynich alphabets and that of a different cipher. Again my focus is much more on purely visual similarities rather than functional similarities. Again, do tell me your thought. Thanks

  19. Mark Knowles on December 9, 2017 at 12:02 pm said:

    Nick: I should say that I don’t believe the identification of common cipher characters will necessarily make it any easier to solve the cipher. However it clearly has another value in that it helps pin down more precisely time periods and potential location.

    At the moment the “4o” character, which is also in the Croatian Glagolitic alphabet, is a loose common thread between the Voynich and Northern Italian ciphers.

    In addition, if we are honest, I don’t think we can be confident associating the “4o” with specifically Milanese ciphers.

    I am not saying that I am doubting the Milan connection or that we are dealing with a cipher, merely that the connection is more tenuous than I originally thought.

  20. Mark Knowles on December 10, 2017 at 2:59 pm said:

    Nick: A question I ask myself and perhaps you is: Why is the “4o” character the character we see in the Voynich rather than others? Was it a particularly commonly used character? Are they other similarly complex characters that were as commonly used?

  21. Mark Knowles on January 1, 2018 at 6:16 pm said:

    Nick: What do you think was the level of knowledge about ciphers outside of the diplomatic world?

    My understanding is that ciphers were really only used for diplomatic communication.

    I would have thought that prior to Cicco Simonetta’s text there would have been widespread ignorance of ciphers as there would have been a significant effort to prevent the dissemination of such knowledge for fear it fall into the hands of the enemy, so that innovative cipher techniques remained hidden from rivals as long as possible. I think the same is true of intelligence agencies today who I am sure guard their cryptographic methods extremely carefully.

    I know it is hard to guess how widespread cipher knowledge was, but I would have thought that given it was a tool with very narrow practical appilcations to the field of diplomacy it would be largely unknown to people outside of that world.

    Your thoughts would be of value. Thanks!

  22. Mark: ciphers were widely known throughout Europe during the Middle Ages (Roger Bacon had famously described them), and were used in trade as well as in diplomacy. Florentines were famous for preferring codebooks to ciphers, but both were probably present in just about every North Italian city around 1400 and beyond. They became even better known after the Treaty of Lodi, but that’s another story. 🙂

  23. Mark Knowles on January 1, 2018 at 7:42 pm said:

    Nick: I ask, because I have been researching the who’s who of the Visconti government, in particular the government of Filippo Maria Visconti. I have noted that I can find references to 4 people who look to have been involved in cipher work in this period. 2 of these people are very close relatives of the person that I am interested in. 1 was possibly a close aquaintance of his.

    Given that my identification of authorship was not done by connecting him to ciphers it makes me want to ask what the probability a person in his position would have had such a connection.

    I could look at many people in his position at this time and calculate the frequency of similarly strong connections to the cipher world, however this seems like a task I don’t particularly want to undertake as it feels like a bit of a digression, though I would if absolutely necessary. I would of thought the people who were precursors to the cipher secretaries would be the most advanced writers of ciphers at that time. I suppose much of the public would be familar with the Caesar cipher, but presumably the diplomats would ensure that they were using the most advanced techniques.

    I may be reaching, but it doesn’t feel like it.

    As far as sources for Visconti ciphers go I have 2 separate lines of enquiry, though I don’t know quite when I will set my eyes on the different Visconti enciphered letters that I seek. I have other relevant lines of enquiry that I am pursuing simultaneously which are progressing slowly.

  24. D.N.O'Donovan on January 2, 2018 at 6:13 am said:

    I’ve been reading your comments here and find it difficult to work out the questions which are driving your research – could I ask you to clarify. I mean, are you working on the question of where the object was made? or what language informs the written text? or the origins and/or provenance of the imagery? or the intention behind the last? Clarity about cultural/religious context? Or perhaps a name that could be attached to an author? attached to an author?

    Or all the above? 🙂

  25. J.K. Petersen on January 2, 2018 at 8:07 am said:

    Mark, I am not convinced that “4o” is a character in the VMS. The 4 occurs by itself without the “o”, and “o” is frequently at the beginning of tokens, which means the 4 is perhaps prepended to “o” words in a way that makes it look like “4o” when it fact it might be 4 + o-word.

    Also, 4o is regularly found in alchemical and astrological manuscripts as an abbreviation for quarto and for 4th grado (degree), so the shape itself was generally known. It’s not common until the mid-15th century and later, but it does exist in earlier manuscripts back to about the third quarter of the 14th century (I have examples somewhere in my files, perhaps even 1 or 2 from earlier in the 14th century).

    So “4” or “4o” occurs in various medieval sources (including ciphered documents) and may be inspired by more than one line of thought.

    The 4o shows up in the mid-16th century as a combined character in a simple substitution code in the letter of Gabriel de Luetz d’Aramon. I would love to think that maybe the originator had seen the VMS, but the de Luetz 4o has no positional component or other characteristics comparable to the VMS, and thus may have been inspired by the abbreviation for quarter (4o) which was common by this time rather than any connection to other ciphers.

  26. Mark,

    much has been written about the topic. Philip Neal has made a summary of one of the important sources on this topic on this page:

    This gives a good indication just how widespread the use of ciphers was throughout the middle ages and throughout Europe, even just outside of diplomatic use.

  27. Mark Knowles on January 2, 2018 at 9:58 am said:

    Nick: I should make it clear that I am interested in the family connections, as I wonder, if I happen to be correct about authorship, how the author began learning about ciphers.

    An older brother already working for the Visconti writing diplomatic ciphers would be an ideal person to convey that information.

    Also given that there is the “4o” in the Voynich doesn’t that make an influence from the diplomatic world more likely than other different kind of cipher usages, as it is in diplomatic correspondence that this character frequently occurs?

  28. Mark Knowles on January 2, 2018 at 10:50 am said:

    Rene, J.K and Diane:

    Thanks for the comments!

  29. Mark Knowles on January 2, 2018 at 10:52 am said:

    I should note that of course I made my earlier comment before reading your replies to the comment preceding it.

  30. Mark Knowles on January 2, 2018 at 11:36 am said:

    If I pick 20 randomly selected Northern Italian/Italian bishops or Abbots of around the period that the Voynich was written in and research the connection between their surname and ciphers of the period to determine the extent to which my authorship candidate stands out then I hope that might be a useful test. I must admit I do not relish it as it will be some work without an obvious side benefit, but if it is necessary then I guess that is that. I guess I could choose those in other professions than bishops, however selecting random individuals would be more difficult. I would be happy to list those individuals who I selected, so others could verify my results. I could randomly select bishops outside of Italy, but then I would have to wonder how far to cast my net. A sample of 20 should be large enough to give meaningful results. Many more than 20 would be too much work to make it feasible. I suppose if there is a cipher connection the nature of that connection would be important.

    It seems to me that in the case I am considering there is a disportionately close connection to the world of ciphers, but a test such as I have described unfortunately seems a more rigourous way of ascertaining whether that is the case than an anecdotal approach.

    Any suggestions for methodical improvements would be appreciated.

  31. D.N.O'Donovan on January 3, 2018 at 12:16 am said:

    It is true, as Rene says, that other people have noticed, or speculated, or offered informed comments on the ‘4o’ combination, but take heart. If you provide new insights to the linguists and cryptographers, they’re likely to respond with intelligent interest.

    I daresay Rene assumes you’ll know some of the better-known opinions, such as Nick’s, but Nick has also fairly represented views opposed to his own and they’re worth considering too e.g. (Tony Gaffney’s opinion, quoted within Nick’s post here ( June 14th., 2009).

    I have no opinion on the question, only more questions which I hope the cryptographers and linguists and statistics guys may answer one day … such as whether the frequency of ‘4o’ in the Vms is close to the frequency of a similar looking but reversed ‘o4′ shape in medieval Latin texts – which last, I must add, is an abbreviation and not actually meant to be read as “o”-and-“4’.

  32. Mark Knowles on January 3, 2018 at 7:15 am said:

    Rene: From this list can we not exclude many of the items as not being akin to the Voynich i.e. to start with all ciphers not using invented alphabets which would reduce the list a lot.
    Then I would imagine that looking at the result list we can shorten it further by other sound criteria.
    Also, presumably diplomatic ciphers would be the most advanced ciphers of the time and so would be the closest parallel to the Voynich.

  33. Mark Knowles on January 3, 2018 at 7:19 am said:

    Rene: Do you know if the list is available as an excel file, csv or similar? This would make the process of pruning the list much quicker.

  34. Mark Knowles on January 3, 2018 at 7:21 am said:

    Does anyone have any suggestions for improving the test that I have described in the “January 2, 2018 at 11:36 am” comment on this page?

  35. Mark,

    the manuscripts listed by Bischoff all include simple ciphers.
    Whether the text in the Voynich MS is a cipher remains open.
    It certainly is not a simple cipher.

    Some of the ciphers in Bischoff’s list (but certainly not all) are based on ‘invented alphabets’.

    Invented alphabets were not only used for ciphers. They were also used for invented languages or for meaningless text.

  36. Mark Knowles on January 3, 2018 at 12:33 pm said:

    Rene: From your list it seems that “Invented alphabets” is the applicable section, of course other features could be incorporated, but this is an absolutely necessary feature and we have no idea if any of the other features listed are applicable.

    So this it seems is the list to work with.

  37. Mark Knowles on January 3, 2018 at 12:44 pm said:


    Wolfenbüttel Weiss dates from the 8th – 9th century, so is hardly relevant. So I think all manuscripts like this should be excluded.

    I would also think manuscripts with very simple ciphers should be excluded.

    In fact ciphers of similar or greater complexity to the diplomatic ciphers of the early 15th should be the central focus.

    This should trim down the “Invented Alphabets” list a lot I would imagine.

    Then I would think the remaining cases could be compared with Voynich and diplomatic ciphers of that time to see if there are any valid alternative sources.

  38. Mark Knowles on January 3, 2018 at 12:48 pm said:


    I note that Nick’s focus has been almost exclusively, from what I have seen from his posts, on diplomatic ciphers having been the predominant influence on the Voynich. Would you agree Nick?

  39. Mark Knowles on January 3, 2018 at 12:56 pm said:


    As is often the case your comment was posted whilst I was writing mine, so I did not see it until afterwards.

    Obviously there is the language versus cipher debate. Other people have tackled this debate to my satisfaction, so I won’t revisit it.

    So if all the ciphers are simple then given the Voynich cipher is clearly complex does it not follow that by far and away the most likely influences are diplomatic ciphers which have a much greater level of complexity?

  40. Mark, to avoid that this discussion goes round in a circle, let me get back to the point where I brought up the list of Bischoff.

    You wondered whether the fact that some relatives of your candidate had to do with ciphers was too much to be a coincidence.

    My purpose was just to give an example of something that Nick had also said in that context, namely to show that ciphers were ubiquitous.

    Whether the Voynich MS should actually be considered a cipher remains an unanswered question. This is not at all a given.

    The safest thing that can be said is that the MS includes some symbols that have also been used in diplomatic ciphers. However, the overlap is not great.

    What’s more, the textual statistics of the MS do not at all fit with the application of a standard diplomatic cipher à la Tranchedino. This already fails the most basic statistic, namely the count of different symbols.

  41. Mark Knowles on January 3, 2018 at 4:54 pm said:

    Rene: I only think trying my experiment is what is left. I will generate a list of all the persons of the same profession in Lombardy, excluding the person that I am interested in, who were in this profession during a period at least partially overlapping 1404-1438; there were not so many in this profession as I am sure you know. I will select the first 20 in alphabetical order of their surname. I will then research their connections to the world of ciphers and count those who have a similar or greater link. I will also look at the degree they would have been connected to herbal and/or astrological manuscripts. I will not be putting in a lot more effort to search for connections than I put into my candidate as that seems reasonable, because I must set a time limit to how long I spend on each person.

    I would be happy to provide you with a list of the individuals that I am considering.

    I will assume we are dealing with ciphers here, for the purpose of the experiment. If it turns up positive results then I will consider the need to justify that it is indeed a cipher. For me this is like the “Is it a map?” or “Is it a hoax?” questions, which I already have a strong opinion on.

    The purpose of this experiment is as much for my benefit though I would welcome the opinion of others on the results. It is useful for me to see at this stage of my research if my identification of a close correlation between my candidate and the Voynich, independent of the 9 rosette foldout, is justified.

    Given that my identification was not based on a connection with ciphers and so independent of it I would be interested in looking at whether this individual stands out. I could look at other attributes such as their connection to herbals and astronomy. Asking directly, do they fit better as a candidate for the Voynich is a more nebulous and non-specific question, so is not easily tackled.

    As I think you know, I prefer the numerical approach, where possible and as rigourously as it can be made, over the more subjective approach. The Voynich is not an ideal framework to work within and there is often some subjectivity left over, but keeping subjectivity to a minimum through numerical methods is better I think than not.

  42. Mark Knowles on January 3, 2018 at 5:09 pm said:

    Rene: When you say “This already fails the most basic statistic, namely the count of different symbols.” can you elaborate?

  43. Mark,

    the Italian diplomatic ciphers tend to obfuscate language characteristics by using several different symbols for frequent (or all) plain text characters, additional characters for nulls and doublets, and characters for some nomenclator words. The result is a character set of well over 50 for each cipher.

    There are a few (not many) scans of Tranchedino’s ledgers online, but D’Imperio (which is fully online) has some clear examples in Fig.39 (p.117).

    The Voynich MS only has some 25-35 different characters, depending how one counts the composites.

  44. Mark Knowles on January 4, 2018 at 10:59 am said:


    Thanks a lot for that. It is really appreciated!

  45. Rene & Mark: don’t forget that as well as documenting the Milanese Tranchedino (diplomatic) cipher ledger, Lydia Cerione also documented the Milanese state cipher ledger, which consisted of hundreds of non-diplomatic ciphers from the second half of the fifteenth century.

  46. Mark Knowles on January 4, 2018 at 1:41 pm said:

    Nick: When you say “non-diplomatic” ciphers what exactly do you mean? Are you talking mostly about ciphers used for sending messages such as between political or religious institutions/individuals or are you mostly talking about something else? If so what kind of ciphers are you refering to?

    Nick & Rene: Regarding the test I proposed, I noted when researching the many individuals connected to the Filippo Maria Visconti government from the period the Voynich has been dated to that only a tiny minority seemed to have any recorded or implied connections with the writing of ciphers. 4 individuals were recorded as having any connections with ciphers where 2 of them were closely connected to the person that I am interested in.

    My hypothesis is that were I to choose a randomly selected sample of people in the same profession from the same period in the same geographical area as my suggested author I will find it rare for the individuals to have the same level of connection to the cipher or obscure language world.

    Do you think that can serve as a valid test? Can you suggest any methodological improvements?

    I would be happy to list the people in question, so that anyone can check the results.

    I am not excited about carrying out this test as it will involve some work on my behalf and I am not sure what other enlightenment it will provide.

    It is important at this stage for me to have some confidence in what I perceive as a pattern really is one, especially as my research looks like it will more and more rely on me putting in considerable effort in tracking down primary sources.

    However if you think the test is flawed then that leaves me back where I started.

    I am not claiming this can serve as a proof, but it can serve as a useful support for my authorship hypothesis.

  47. Mark Knowles on January 4, 2018 at 4:46 pm said:

    Nick: Where did you see the Cerioni book? It appears not to be available for purchase or held in any libraries in the UK, but I could be mistaken.

    I believe it is held in the library of Tortona and I could certainly get them to scan pages and email them to me, but that could work out quite expensive for the whole book. The library of Alessandria already emailed me some pages from a different book I was interested in, but I knew which 4 pages I wanted, so this didn’t cost much.

  48. Mark Knowles on January 5, 2018 at 12:03 pm said:

    Nick: Looking through the Tranchedino I noticed the following character occurs often as it does in the Voynich. It may be incidental, but I thought it worth mentioning it, if you haven’t already done so youself ->

    (The top line should be touching the ^)

  49. Mark Knowles on January 5, 2018 at 12:27 pm said:

    Nick: The character also occurs in Meister. I wonder if one could look at the frequency of common Voynich characters like the 4o and the above character and even more simple common characters in ciphers to determine within which region, era and sphere they are most commonly associated, given available sources. So for example if these kinds of symbols are found most commonly in Milanese diplomatic ciphers of the mid 15th century that would tell us something even though it wouldn’t prove anything directly.

  50. Mark Knowles on January 5, 2018 at 12:29 pm said:

    Nick: What the frequencies are of early 15th century Milanese ciphers has yet to be determined.

  51. Mark Knowles on January 5, 2018 at 2:31 pm said:

    We also see the:

    Character in the “In Milano” cipher mentioned above.

    In addition we find:

    “a”, “8”, “8 minus bottom like simple and symbol”, “? minus the dot”, “S reflected on the horizontal”, “cc possibly a coincidence”, “^” and “One other difficult to describe which may be a coincidence”

    I haven’t checked all other Meister or Tranchedino ciphers for the letters, but in this cipher there does seem to be quite a bit in common.

  52. Mark Knowles on January 5, 2018 at 2:41 pm said:

    I think the ? without the dot is reflected horizontal, so this may be less significant.

    A closer examination of the symbol may point out other similarities.

    It would be interesting to compare the commonality of the symbols with other ciphers. They may all be relatively common or particularly noticeable in this cipher.

    I very much look forward to the time when I will see some Visconti ciphers to compare with the Voynich.

  53. Mark: the difficult thing about ciphers is that whereas a few of the cipher ledgers have survived (e.g. the Tranchedino ledger, which was actually the third iteration of a well-used single ledger), hardly any cipher letters have – once decrypted, they were usually destroyed. So if you want to find any Visconti-era ciphers, the first place to look would perhaps the Urbino cipher ledger, which I’ve written about several times before, e.g. .

  54. Mark Knowles on January 5, 2018 at 3:22 pm said:

    Nick: Thanks a lot for that link, really appreciated as always. I have downloaded it and I will certainly study it in more detail.

    The examples I am looking for at the moment are:

    1) The intercepted letters address to Guarnerio Castiglione who was a Milanese papal envoy. The interception of the letters caused a bit of a diplomatic incident. I hope these might be somewhere in the Vatican Archives.

    2) The letters of Marcolino Barbavara who was a later papal ambassador under Filippo Maria Visconti and then under Francesco Sforza which are referred to in various places, so hopefully avoided destruction somehow. There are other reasons why I would like to see these letters.

    Finding these is non-trivial and I think it will be easier with the Marcolino Barbavara. However the Castiglione letters date from about 1431 whereas the Barbavara letters date from the 1440s onward. I am not yet aware of other candidates for suitable letters. I would guess intercepted letters is an area to further research, but not the only area.

    I have a possible idea of how I might, if lucky, find some letters in the hand of the author I am interested in. However again this probably won’t be easy, but I would like to do a handwriting comparison.

    Do you have an idea of how many pages the Lydia Cerioni is? Have you seen it? I am contemplating requesting a library in Italy to scan and email me it, possibly only the pages with ciphers on. However depending on the number of pages it might be rather expensive. I would naturally be happy to make the scans available to whomever would like them.

  55. Mark Knowles on January 5, 2018 at 5:08 pm said:

    Nick: I can find references online to->

    “La Diplomazia sforzesca nella seconda metà del Quattrocento e i suoi cifrari segreti” by Lydia Cerioni

    However I haven’t yet spotted a reference to a list of non-diplomatic ciphers. Are they in a different book?

  56. Mark Knowles on January 5, 2018 at 5:15 pm said:

    Nick: I see volume 2 of this book. I found one of your email threads.

  57. Mark Knowles on January 5, 2018 at 5:21 pm said:

    Nick: Your thread says the book is in the British Library, so the internet lied to me or more specifically Worldcat did. I wonder if it is in the Bodleian. Otherwise I guess a trip to London may be in order.

  58. Mark Knowles on January 5, 2018 at 5:33 pm said:

    Nick: It seems one version of it is available in Oxford and London and all the other versions are not available in the UK.

    Sorry for the hassle. I will do my best to scan as much as possible, so that I can share it with you all in electronic format. I hate paper books, but unfortunately not everything is available as an Ebook or equivalent.

  59. Mark Knowles on January 5, 2018 at 6:31 pm said:

    I have spotted a symbol which looks like a lowercase greek letter sigma, which is in the Voynich, in one of the other ciphers. I haven’t checked if there are more ciphers which contain this. Again something like this could be coincidental, but put together the appearance of multiple of the simple symbols in the Voynich in specific ciphers could act as some kind of pattern.

    I could be curious to see how similar the non-diplomatic ciphers are to the Voynich.

  60. Mark Knowles on January 5, 2018 at 7:07 pm said:

    It is worth noting that the characters that represent a specific word or thing in the non-Voynich diplomatic ciphers appear to almost always represent a place/person like Milan, Florence, Pope, Modena etc. For standard diplomatic ciphers these are relevant. Given the different subject matter of the Voynich these would be very much superfluous and the motivation for having them would not exist.

  61. Mark Knowles on January 5, 2018 at 8:47 pm said:

    Out of curiousity I looked at Hildegard of Bingham alphabet and Rudolf IV alphabet in Meister and I could see 0 symbols in common with the Voynich. I looked at for example the “In Milano” alphabet and I see lots of symbols in common not just the “4o”.

    I have not seen a wide range of non-italian non-diplomatic ciphers alphabets, but it seems questionable if there are far fewer similarities in these with the Voynich that there is no evidence of a pattern here.

  62. Mark Knowles on January 6, 2018 at 12:23 pm said:

    Nick: I downloaded and looked at the Urbinate ciphers. However there does not seem to be quite as much in common with these alphabets and the Voynich compared to the “In Milano” cipher. I probably ought to look through all the other ciphers and record their similarities.

    I suppose I could investigate what exactly “In Milano” means. Does it mean a cipher sent from or to an envoy in Milan? Does it mean a Milanese cipher?

    Ideally I would like to find a Milanese cipher from the 1430s i.e. the era of Fillipo Maria Visconti. However any Milanese cipher from the reign of Filippo Maria Visconti would be great. I would be very curious to see the amount of differences and similarities of these ciphers and the Voynich.

    Milanese ciphers from the reign of Gian Galeazzo Visconti would also be of interest though significantly less interesting.

    It is a lot to ask, but I will do my best to find examples of these kind ciphers.

  63. Mark Knowles on January 6, 2018 at 3:07 pm said:

    I have looked at the “In Milano” again and I noticed there is also a “u” shared with the Voynich and also another character which is difficult to describe is shared. I have not yet looked carefully through the extended Voynich character set, so there may be even more commonality.

    Now characters like “a”,”u” and “^” are not particularly remarkable characters. However these characters combined with the more distinctive shared characters make for what I think is a significant amount of commonality. I will certainly scour the other Meister and Tranchedino ciphers for alphabets with a greater amount of commonality. Nevertheless from my researches there seem to be significantly less commonality amongst natural alphabets or non-diplomatic cipher alphabets that I have seen. I should see the Lydia Cerioni soon to check what the level of similarity there is between the Voynich alphabet and the non-diplomatic cipher alphabets.

  64. Mark Knowles on January 6, 2018 at 3:15 pm said:

    There could be a case for giving each alphabet a score as to how much commonality there is with the Voynich. This could either be a simple count of the number of common characters or a weighted count with the more distinctive characters given a higher weighting.

    Having a measure for comparison could be useful.

  65. Charlotte Auer on January 6, 2018 at 3:28 pm said:

    Mark: since I don’t have the time to read your numerous blog posts in in detail I’d like to ask you some few questions.

    You seem to be convinced of a certain authorship, a candidate of Milanese origin connected with some of the ruling families. Who should that be and why?

    Wouldn’t it be much easier for you to just name your candidate and see what others can contribute to that idea? To compare diplomatic and non-diplomatic medieval cyphers is not the way to the authorship unless a previousely undiscovered cypher would be revealed and could be linked to your candidate beyond doubt. I’s not absolutely impossible, but highly improbable.

    Btw.: it was Hildegard von Bingen (and not “of Bingham”) who invented her own alphabet (lingua ignota) for her mostly private use and contemplation.

  66. Mark Knowles on January 6, 2018 at 3:29 pm said:

    In theory if one calculated a comparion score for each cipher alphabet or non-cipher alphabet with the Voynich, one could use these statistics to study the patterns of commonality by geography or era or type or anything else of alphabets.

    In the event that some ciphers have different distinct characters in common with the Voynich from others we could use this to hypotheses about unknown parent alphabets from which the two may be partially derived.

    I believe that many of the characters in the Voynich are of the author’s own invention, most probably the most complex characters; despite this I think there is still a lot of scope for looking into commonalities amongst the other characters.

  67. Mark Knowles on January 6, 2018 at 4:12 pm said:

    Charlotte: As far as the name of Hildegard, I apologise; to be honest I have not given her much thought as I don’t see any connection with the Voynich, so you will have to forgive me for getting the name wrong.

    As far as authorship I would not use the word convinced as it is a very strong word. However I have had a hypothesis for some time which I haven’t been able to shake off and which to me makes more sense than any other that I can conceive or anyone else has to my knowledge proposed.

    I think you have not completely understood my motives behind the comparison of cipher alphabets. I am not proposing in my case that comparing diplomatic and non-diplomatic medieval cyphers is the way to the authorship; though rather a way of checking whether the link I see holds up. Having said that I do think comparing the similarities of cipher alphabets could help to narrow down the geography and the kind of influences of the author.

    I have my own specific hypothesis, but I am also interested in exploring broader questions such as those of geographical origin.

    My authorship hypothesis derives solely from my analysis of the 9 rosette foldout most of which I have presented in great detail on this website. Recently I have been focused on exploring and looking for evidence which supports or undermines it. I think I have seen evidence of the author’s connection to the world of ciphers which is beyond coincidental.

    I believe the author to be Italian, but as is clear from my 9 rosette foldout I believe the author had travelled in Southern Germany and Switzerland.

    I think your area of research, if I remember correctly, is on the basis that the manuscript is solely German in origin; correct me if I have got your hypothesis wrong. So I would think there would be little commonality between our research paths as we having quite different starting points. My theory and line of research is probably closest to Nick’s, although there are so very significant differences.

    If you are interested in collaborating drop me an email at:

  68. Charlotte Auer on January 6, 2018 at 6:11 pm said:

    Mark: thank you for your answer.

    My working hypothesis is of course on the basis that the VM is of German origin, but this doesn’t mean the German language or Germany in general. I also see my starting point in Northern Italy, namely in Tyrol, under the historic circumstances as they were in the early 15th century. This is quite different.

    One point in your approach irritates me a lot: what do you think is the underlying language? The world of cyphers was the world of Latin, and Italian (i.e. in Northern Italy) as a written language was not very developed or common those times. As a consequence from your point of view, Latin would be the most probable language to be encyphered by your special candidate. Is that your assumption?

  69. It is perhaps worth pointing out that english ‘irritate’ and german ‘irritieren’ are so-called ‘false friends’. The words sound the same but have rather a different meaning. I strongly suspect this affected Charlotte’s wording. German ‘irritieren’ basically means ‘to confuse’.

  70. Mark Knowles on January 6, 2018 at 9:40 pm said:

    Rene: Thanks for the reference.

    As you have no doubt read I am very interested in how one might measure the degree of similarity between any 2 alphabets.

    Questions such as:

    What is the likihood that the similarity between any 2 alphabets are coincidental?
    What the probability that they have a common root?

    Now these are clearly difficult questions to answer. However I think they can be approached.

    As a question, if I grabbed someone off the street and I asked them to produce a list of 40 symbols imposing some kind of limit on the complexity of the symbols then how similar to the Voynich alphabet would those symbols be?

    Is there some way of modeling the process of creation of symbols?

    Again these questions become more difficult and could really be developed more precisely and I don’t propose to answer all of them. However I believe there is sound way in theory at least of measuring the similarity between human generated alphabets and whether they are independent of one another and how related they are.

    I am inclined to the view that the similarity between these Italian diplomatic ciphers and the Voynich are significant. The question becomes how can one best attempt to prove this.

    I am obviously not an expert on human visual symbol formation psychology, in fact I don’t know what the proper name of this area of research is.

  71. Mark Knowles on January 6, 2018 at 10:00 pm said:

    Charlotte: I appreciate your reply and interesting questions.

    First of all at this stage I wouldn’t like to say what the underlying language or languages was. It is worth noting as it states in Cicco Simonetta’s paper on this invaluable website that more than one language could be used to confuse the person trying to decipher the message. So there could be more than one language.

    You are right Latin would be a plausible choice, but there are other languages the person I am interested did or may have known.

    Unfortunately I am far from being at the stage of identifying the underlying language; in fact I would think that this would be done at near the very end of the process.

    At this stage I have an authorship hypothesis that I would ideally like to prove or disprove. If I have correctly identified the author I hope further research may provide some clues as to some aspects of the workings of the cipher, so making the process decipherment a bit easier.

    I have some ideas as to lines of research I would like to pursue into the Voynich cipher, but at this stage my main focus is on authorship.

  72. Charlotte Auer on January 7, 2018 at 11:57 am said:

    Dear Rene,

    thank you very much for your kind help! Your explanation of the confusing ‘false friends’ in my wording is absolutely right.

  73. Mark Knowles on January 7, 2018 at 12:46 pm said:

    In general I am really keen, where possible and as far as positive, to move from subjectivity to objectivity.

    So rather than being faced with a question where person A says connection X exists and person B says it doesn’t, to, if possible, trying to approach the question in an objective way and so get closer to a definitive answer.

    Now I know “the Study of History” is not an exact science, but I think, certainly, in the context of the Voynich there is a scope for the application of scientific/numerical/statistical methods not just in the anslysis of the ciphertext.

    I, personally, get very frustrated by the morass of claims about the Voynich which are very much one person’s perspective against another’s without any move towards resolution and consensus. And I firmly believe there is scope for escaping this situation though there will envitably be some questions for which this is not possible.

  74. J.K. Petersen on January 8, 2018 at 4:44 am said:

    Resolution might be a good goal but I have my doubts about consensus. Most human achievements and most of human progress were not arrived at by consensus.

  75. D.N.O'Donovan on January 8, 2018 at 11:21 am said:

    -Jkp –

    I’m with you on one point, at least. The notion that we can ‘decide by consensus’ what is, and isn’t historical fact is fairly bizarre if there’s no sign that the pronouncers have the slightest knowledge of historical studies or the slightest desire to open a book and study.

    We’ve seen the utter fiasco which can result from pushing this notion that ‘consensus’ is all there is to historical study: I mean the positively crazy decision to issue an ‘authoritative Voynich herbal’. Utter madness which endured for almost two years.
    No matter how many people agree that it says ‘Moo’, a frog will never be a cow. (Genetic-engineers apart).

  76. Diane: why read a book when you can click to vote in an online poll? 🙁

  77. Mark Knowles on January 8, 2018 at 3:18 pm said:

    Maybe consensus is too much to much to hope for. I like to think consensus amongst intelligent reasonable people is feasible i.e. the people who think the Voynich was written by aliens are beyond hope. There is an overwhelming consensus on certain questions already, so it clearly has been achieved in some areas.

    We all have a propensity to be reticent to change our mind in spite of the evidence as I have discussed elsewhere.

    It is the case that the physicist Max Planck said, “a new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.” Which is a rather depressing outlook, but people can be very much invested in their theories and unwilling to give them up despite the evidence to the contrary. Unfortunately whatever the truth is we all have to accept it like it or not; who wants to believe a lie?

    I do think questions such as is alphabet(I) more similar to alphabet(II) than is alphabet(III) is a reasonable question to answer and amenable to a rigourous scientific approach, it only becomes very difficult in the case of alphabets that are very close in their degree of similarity to the comparison alphabet.

    As someone with a Mathematician background I think a consensus on a theory which has been rigourously proven is unavoidable amongst people with an ability to understand the proof or respect for those who do. Whilst in practice many historical questions are not open to the same level of rigourous formal arguments I think it has to be something that is worked towards.

    I must admit I have not studied historiography, but I have some basic familarity with it.

    Those who are interested in reading books might be interested in:
    “An Introduction to Quantitative Methods for Historians” by Roderick Floud
    Whilst I am not deeply immersed in this area of research my sister-in-law is a political historian and almost all her work revolves around the use of statistical techniques. That is not to say I don’t value other approaches providing they are as rigourous of they reasonably can be.

    So as Nick points out well, amongst the general public there is no consensus on basic questions such as Evolution, Climate Change or even in some cases whether the world is flat, but I like to think that we as serious people can do better than that. And amongst academics there is a lot of consensus if even if not so with the general public.

    I would like to say that the carbon dating of the Voynich was I think the most important advance in the Voynich research, which is of course a technique that comes straight out of science. I have discussed elsewhere other potential scientific advances in our understanding of the Voynich.

  78. Mark: it may well have been someone other than me who pointed out the lack of consensus “amongst the general public”, that doesn’t sound like one of my arguments. 😉

    As far as Voynich research goes, I’ve long given up on the idea that any kind of meaningful consensus could be formed – even the radiocarbon dating range seems to me to be slightly inconsistent with the kind of Art History dating I’m personally most comfortable with (though only by a couple of decades), so even the hardest of hard sciences isn’t yet enough to force an absolute consensus.

  79. Mark Knowles on January 8, 2018 at 4:24 pm said:

    Nick: Sorry I was trying to reflect what I may have misunderstood you to be saying when you mentioned voting in online polls. I thought you were pointing out that the results of online polls taken by the general public are not a reliable source of truth and that this kind of approach to consensus forming is unlikely to lead to enlightenment.

    Well my raising this subject came about from discussing the relationship between alphabets. Rene mentioned, I believe, that similarities between the Voynich alphabet and, for example, certain Italian diplomatic cipher alphabets are not really significant. I believe this question is amenable to analysis and we don’t have to be left with an argument where one person is saying “Oh, yes it is!” and other person is saying “Oh, no it isn’t!” back and forth.

    Otherwise what is the point in advancing an argument at all if there is no scope for persuading anyone?

  80. Mark: the specific difficulty of historical persuasion is that it has (in recent years) come to rely almost entirely on close reading of evidential texts, which is the one thing we don’t have here.

    Putting everything else aside, your primary challenge would seem to be to build up some kind of picture of the Visconti regime’s cryptography, something which I suspect currently has effectively no literature on it at all. In some ways, this is a good thing, because it means you get all the glory at the end: but it can also be a hard slog. :-/

    You might ask Professor Andrea Visconti (a cryptography Assistant Professor in Milan) if he knows of anything, it might possibly be something he has previously heard of or stumbled over, or he might happen to know someone. Asking is free! 😉

  81. Mark: p.94 of has an enciphered 1422 letter from the Archivio Capitolare di San Lorenzo di Genova (ACSL) [which is now at the rather better-known Biblioteca Medicea-Laurenziana Archivio Diocesano di Genova (housed in the Museo Diocesano on Via Tommaso Reggio) ], cartella 391, n. 37:

    Matteo del Carretto a Pileo
    Subiaco, 27 dicembre ( 1422 )
    Originale cifrato (sistema sostitutivo monoalfabetico con lettere nulle, omofoni e repertorio), decifrato nell’interlinea in A.C.S.L., cartella 391, n. 37. Sigillo placcato.
    L’indicazione dell anno si ricava dalla presenza a Roma di Bartolomeo Capra, arcive­scovo di Milano (A.S.V., Reg. Lat. 231, c. 5 r.), inviatovi per trattare il matrimonio
    tra Filippo Maria Visconti e Caterina Colonna, nipote di Martino V (L. O sio cit., I li, pp. 98, 105), dall’accenno al concilio di Pavia-Siena e dall’invito a recarsi a Roma. La provenienza della lettera ne denuncia chiaramente il mittente.

  82. Mark wrote:
    “Rene mentioned, I believe, that similarities between the Voynich alphabet and, for example, certain Italian diplomatic cipher alphabets are not really significant.”

    This is perhaps not very accurate. These ciphers are one of four apparent sources that could have influenced the person creating the Voynich MS alphabet. Of these four, the “alchemical symbols” are the least likely.

    What I did mean to point out is that the text statistics of the Voynich MS text are not compatible with the product of applying one of the (say) Tranchedino ciphers to a European plain text in Latin or some well known vernacular.

  83. Rene: it is entirely true that fifteenth century ciphers known from the historical record do not help us decrypt Voynichese. Yet if you look at the way number shapes are used, these are very typical of Milanese ciphers from the pre-Arabic digit phase of cipher alphabets (i.e. ciphers that don’t have 10, 20, 30, as cipher shapes). As such, I suspect this comprises much of the similarity Mark is looking for.

  84. D.N.O'Donovan on January 9, 2018 at 12:35 pm said:

    Consensus among persons *capable* of being intelligent and reasonable has been responsible for most of the horrors in history – because against the individual’s capacity for reason is the very human drive to become and remain one of the group they consider the majority. It is not an urge to conformity, but for identity and is our very deepest human drive – for which individuals will surrender everything else, including their own lives. Why else would someone prefer to die than to be given a little white bird-feather? If you know – and who cannot – the response which has met all those who maintained opinions other than that the manuscript’s content was entirely the creation of some European Christian male, or that its language was other than the sort of language known to such a person, then you will perhaps see why Voynich studies not only attracts highly intelligent, competent and properly qualified persons – such as Stolfi, or ‘Sam G’ or any number of others – but why so few remain in the arena for long.

    If I may give you one pointer… when someone makes an authoritative sounding assertion, mentally put “in my opinion…” in front of it; then ask to see the body of their own, original research from which they drew that conclusion… and read it critically… and finally if there’s no such body of research ask the person the simplest of all questions, ‘How do you know?’.

    In the case of the Milanese ciphers, I understand that the research was actually done by Nick Pelling (though he may correct me on that). Certainly the only intelligent discussion of the pros and cons which I’ve read has been Nick’s. And though his book is now oop, some posts to this blog explain his evidence and reasoning.

    I’m not qualifed to comment on ciphers, and not much interested in them, but that’s not my area so it doesn’t much matter.

  85. Diane: please use your own blog for straw men arguments, I’m getting a bit tired of hearing about this supposed European Christian male theory hegemony.

  86. Mark Knowles on January 9, 2018 at 6:25 pm said:

    Nick: It is difficult that we do not have the kind of evidence that we would really like. However I think we must make use of what evidence we have and take it as far as we can go with it.

    You are right it looks like a hard slog, but as is always the case, if it were easy it would not have quite the same value. In general at the moment the different threads of my research are hard going, though progressing slowly, due to the lack of easily obtainable information and this is very much the case with Visconti ciphers, though without the internet everything would be 100 times harder. (I suppose increased digitisation of historical documents and better means of searching within them and through them over time will make things easier. Still even these days the wheel of progress in this area turns slowly.)

    As usual you have been very helpful in mentioning Prof. Andrea Visconti I will certainly follow up with him. Again the Genose cipher reference is really appreciated; I will follow that up too.

    I will be looking at the Lydia Cerioni non-diplomatic ciphers soon to make a comparison with the Voynich. Maybe I will find a non-diplomatic cipher alphabet more similar to the Voynich. I am perfectly prepared to change my mind if the evidence justifies it. It is often more interesting to change one’s mind and embrace a new perspective, so I am not averse to it. However of course one shouldn’t change one’s mind for the sake of it and at the moment I am most persuaded by my perspective as it is.

    I really appreciate your support and encouragement, it has genuinely been invaluable:)

  87. Mark Knowles on January 11, 2018 at 11:46 am said:

    Rene: I was responding to your comment “The safest thing that can be said is that the MS includes some symbols that have also been used in diplomatic ciphers. However, the overlap is not great.”

    Specifically your point that “The overlap is not great.”

    I was addressing how the question of whether the overlap was significant or not might be answered. So as to say what the probability is that the Voynich alphabet and the Italian diplomatic cipher alphabet(s) both had a common ancestor(s). I was discussing how methodically one might rigourously test such an assertion. If for example one could say the likelihood is very high that would be a very important breakthrough.

    Each time a new cipher alphabet was implemented a wholly original set of symbols was not generated, but rather some pre-existing characters were reused again and again though most probably with a different interpretation each time.

    I would think that each symbol was introduced by someone for the first time in a diplomatic cipher at a certain point in time and subsequently reused. And of course at some point that symbol ceased to be used in ciphers.

    On that basis it would seem reasonable to assume that the most similar cipher alphabet to the Voynich dates from around or not long before the creation of the Voynich.

    To say whether the cipher is similar or different is hard to say. There may be 1 or 2 significant innovations in the Voynich cipher which account for those statistics you observe. It is self-evident that the Voynich cipher is distinct from any diplomatic cipher we know, but I think it is fair to ask where we see the greatest similarity as it is also not like to any other cipher/language. The visual appearance of the alphabet is a simple straightforward point of comparison. Not only the question of the number of identifical unique and common symbols shared between the two, but also in the way the symbols are formed is I believe important i.e. things like whether they have smooth flowing lines of hard straight lines.

    I studied many alphabets before looking at ciphers when I first developed an interest in the Voynich. I spotted in the Croatian Glatolitic alphabet that there is also a 4o symbol. However I cannot say I saw any alphabets more similar to the Voynich than what we with the diplomatic cipher alphabets.

    It is important to note that the 4o character is not the only distinctive Voynich character we see used again and again in diplomatic ciphers. Once I have worked out the standard code used by the Voynich community to define each of the characters I will list the shared characters and my comments on each.

    My hypothesis is that there should be the greatest similarity with Milanese diplomatic ciphers from the years 1415 to 1435 and the Voynich alphabet. Unfortunately these ciphers are very difficult to locate as records from that period for Milanese ciphers are hard to come by.

    If my hypothesis is true by so doing I can establish a link to close family members of my suspected author which I personally think it would be hard to deny a connection, though in these circumstances I am sure many would. Frankly merely a link to Italian diplomatic ciphers becomes I believe highly significant.

    I should say that suspect that the gallows characters are a completely original creation of the author of the Voynich and to be found nowhere else. This may well be the case for other “families” of characters. I think one can expect the greatest commonality to be with the distinct isolated characters as I believe that they are the most likely to have been inherited from pre-existing cipher alphabets. So I very much doubt I will find any identical alphabet to the Voynich in any diplomatic ciphers.

  88. Mark: cipher alphabets were not identical to each other, that they were different was kind of the point – but it is likely that individual cipher makers reused shapes and preferred cipher tricks, so there may well be family resemblances out there waiting to be found. 😉

    Though I broadly applaud the direction in which you’re trying to move, History is at its best when it is able to stay a primarily evidential activity, and so far you haven’t yet collected together and analyzed any actual ciphers from the time and place you’re interested in. But this isn’t really your fault, it’s a lacuna in the whole Voynich research field – nobody has yet tried to collect together ciphers from 1400-1450 (aside from the Urbino cipher ledger, which I’ve already mentioned but have never actually seen for myself) in any obvious way. This might well be because researchers are only now starting to come to terms with the radiocarbon dating. :-/

    In fact, I’ve been thinking about trying to make this an activity all Voynich researchers could contribute towards: I’ll be posting about this very shortly.

  89. Mark Knowles on January 11, 2018 at 12:30 pm said:

    Nick: I think I must of not made myself clear, I was not saying that cipher alphabets were ever identical to one another and I fully appreciate that the purpose was to change cipher alphabets regularly.

    Having looked through the Tranchedino and Meister cipher alphabets I have noted some common characters with the Voynich other than the 4o and I was going to email you a list with the images for each character and comments. I wanted to investigate this out of curiosity and I think it helps to support the thesis that there is a relationship between the Voynich and diplomatic ciphers, though not exclusively my specific claims in terms of dating.

    I agree I have yet to analyse the ciphers that I am looking for and I fear it will take some time before I have, unfortunately, but I am on the case.

  90. Josef Zlatoděj Prof. on January 11, 2018 at 3:49 pm said:

    Ants. I do not want to take illusions. But on each side of the manuscript, it is written : that is written in the Czech language. 🙂

  91. bdid1dr on January 12, 2018 at 12:45 am said:

    @Nick and every other person who is really interested: I suggest you get a copy of Busbecq’s Letters . Publisher: Aramco World issue 201502 : “The Busbecq Letters” .htm
    Today, my toppling pile of downloads ended upside down. Just a few moments ago, my husband found my reference to a very interesting book I’ve been recently reading. I hope you will be able to find a copy of the book:

    “Lords of the Horizons” Written by historian, journalist and travel writer, Jason Goodwin — . Goodwin also wrote: “On Foot to the Golden Horn” . Mr. Goodwin, in 1993, received the John Llewellyn Rhys Prize in 1993.
    Fabulous reading — I’m on page 259 of a 336-page book which includes a chronology – and a glossary. I have reluctantly put down the book — so I can sit down at the dinner table (like a civilized person) and enjoy the meal my husband prepared for us.


  92. Mark Knowles on January 12, 2018 at 1:59 pm said:

    Nick: I will email you as soon as possible, but for the time being I thought I would mention something that you may or may not have noticed.

    When we see a character, including the gallows characters: if there is a top left loop we have the number 4, if we have a top right loop it is often, though not always, an italic “p”. This process of combining characters joined up to form new characters in this way is standard in the diplomatic ciphers, though the Voynich alphabet takes this approach further than any other cipher alphabets that I have seen especially when it comes to the gallows characters.

    We, of course, also see the combining of joined up characters in the case of the 4o character. Adding an italic letter “p” joined up to the right of a existing character is a very very common practice in cipher alphabets, this is particularly obviously when looking at the simplest Voynich characters with a top right loop, especially as these specific characters are particularly common amongst cipher alphabets unlike the unique gallows forms.

    I think both the 4 and italic “p” make the Voynich alphabet fit within the diplomatic cipher alphabets that I have looked at.

    I hope I have made this clear. More to say on cipher alphabet comparisons, but I wanted to put this out there as I think it is a particularly striking parallel with the Voynich.

  93. Mark Knowles on January 12, 2018 at 3:08 pm said:

    Nick: Worth noting the italic “p” may sometimes appear as a non-italic “p” both in diplomatic ciphers and the Voynich, I imagine this tends to be the case where the “p” at an italic angle does not fit as neatly.

  94. Mark Knowles on January 12, 2018 at 4:10 pm said:

    Nick: I think what makes the gallows characters stand out is that you have as an example a “4P” character combined with a cccc style character below to make one mega-character. With diplomatic ciphers for example you might have a “4P” character underlined or with a line through it or a number above it, but to have quite such a complex character is non-existent. Then given that the Voynich is, I think, a diplomatic cipher on steroids it is not surprising that it has to have the most complex characters.

  95. Mark Knowles on January 12, 2018 at 4:52 pm said:

    Nick: The four “italic” p ending characters that are common to the Voynich and diplomatic ciphers in order of frequency of occurrence in diplomatic ciphers are:


    I have written them in the best way I could using normal letters. Obviously all characters need to be joined up to the middle of the P. For the “HP” the right most line of the H is the same line as the vertical line in the P.

  96. Mark Knowles on January 12, 2018 at 5:13 pm said:

    Other common characters that we see in the Voynich and Diplomatic ciphers:

    c~ (Joined – straight top not curly. Can sometimes have a dot ontop.)
    8 on it’s side (The infinity symbol)
    co (joined at the top)
    ccc (joined at the top)

  97. J.K. Petersen on January 12, 2018 at 5:48 pm said:

    Mark, I think what you are seeing is Latin scribal conventions that found their way into diplomatic ciphers and into the VMS. There doesn’t have to be a direct connection between diplomatic ciphers and the VMS for these shapes to be found in common.

    For example, the shape that looks like Italic p is the Latin appreviation for “-is” (usually used as a suffix), which is also used to indicate “-em” for words like “Item”, and which is occasionally used in place of the more flourished “-rum” character.

    Also, the letter c, in languages that use Latin scribal conventions, is very commonly written as a ligature (joined to the next letter), probably moreso than any other letter. When you see two of them (sometimes three) joined at the top, this is the Latin scribal abbreviation for several different combinations, including cc, cr, rc, er, re, rt, ct, cer, ter, etc. This is because the old-style “t” was written almost the same as “c” and because there were forms of “r” and “e” that are written like “c”. Thus the cc bench character in VMS and in diplomatic ciphers is morphologically the same as a very common Latin general-purpose ligature which was ubiquitous to writing styles of the late 14th and early 15th centuries.

  98. Mark Knowles on January 12, 2018 at 6:44 pm said:

    I believe from looking at the extended Voynich characters that there is character which looks like the letter “a” to the power of “0”; this is typical of many diplomatic cipher characters. Also it appears there are characters like the 4o with a line ontop or 4o with dashes ontop this is typical of how diplomatic cipher characters are built up. Again putting a “.” ontop of a character as we see with some extended Voynich characters is not usual in many diplomatic cipher characters.

    Originally I was going to write that with diplomatic ciphers the similarity of two characters visually has of course no relationship to their meaning: so for example “c” and “ccc” whilst similar visually would have no relationship in meaning. However one thing that I noticed was that some cipher writers, possibly as a result of laziness could for example label consecutive letters as for example f = “1P” g=”2P” h=”3P” i=”4P” . It is conceivable the Voynich writer might have done something similar, though hard to believe given the seriousness with which he took his cipher.

  99. Mark Knowles on January 12, 2018 at 6:49 pm said:


    Can you provide examples of what you are refering for comparison purposes as I find it a little suspicious? If you can show me text that fits the same format as the Voynich characters that are shared with diplomatic ciphers then I will be very interested.

  100. Mark Knowles on January 12, 2018 at 6:59 pm said:

    JKP: I have looked at the page on your website on “Latin scribal conventions” and I see some very very loose similarity to the Voynich alphabet; there are much greater similarities with the Voynich alphabet and the Tranchedino cipher alphabets.

  101. Mark Knowles on January 12, 2018 at 7:10 pm said:

    JKP: I think you are confusing joined up writing and joined up characters. The “4P” is just a “4” joined to a “P”; well we all know the number “4” and the letter “P” in their own right, so what’s the big deal? Well “4P” is a character composed of letters joined together not joined up writing of separate characters.

  102. Mark Knowles on January 12, 2018 at 7:59 pm said:

    JKP: I have read your article carefully and it is well written and has some interesting things I didn’t know in it. Unfortunately it differs widely from my perspective, “Gujarati” sounds quite a romantic language for the Voynich to be written in, however it is a very different line of thinking from my own.

    I am a bit unsure whether you think each character is a word in its own right or more like a single letter.

  103. J.K. Petersen on January 13, 2018 at 3:19 am said:

    No, I am not confusing joined-up writing and joined-up characters. I am specifically talking about ligatures (I have professional expertise in this area and am well informed about the difference).

    Mark, I’m glad you looked at my blogs, but I have only documented about 5% of the most common Latin scribal conventions due to lack of time. There are many more that can be found in both the VMS and in diplomatic ciphers and also in several natural languages.

    Also, I never said that the VMS was written in Gujarati. My article is about scribal similarities and abbreviations found in *both* the Indic and Latin languages in the Middle Ages, that no longer exist. The overlap disappeared when modern Latin moved away from Medieval scribal conventions. These overriding concepts are important not only in researching glyph shapes but also in understanding how expansion of certain glyphs was a common practice in diverse medieval cultures.

  104. Mark Knowles on January 13, 2018 at 9:08 am said:

    JKP: All I can say is that what you have presented in your blog appears to have a rather limited correspondence with the Voynich. I can only comment on what you have written down not what you haven’t.

    It is good you have professional expertise in the area, but that doesn’t make what doesn’t resemble the Voynich begin to do so.

    I meant just to point out that from my perspective the links between Indic languages and the Voynich are tenuous.

  105. Mark Knowles on January 13, 2018 at 9:27 am said:

    Rene: One thing I don’t understand is that you said that the Voynich has fewer characters than cipher alphabets. However having looked at lists of Voynich characters that does not appear to be the case.

    As I pointed out elsewhere I think the habit of using a single character to represent an individual city state has little relevance in the context of the Voynich, so the habit of using a single character to correspond to a single word is probably unlikely to have carried over.

    It seems perfectly logical that some characters appear only rarely.

  106. D.N.O'Donovan on January 13, 2018 at 12:17 pm said:

    I hope your ‘suspected author’ had extraordinary knowledge of astronomical systems other than Ptolemy’s, or that your theory only has him compose or encipher the text’s written part.

    This because Crux is certainly pictured on one folio (though there the drawing is later than the rest, so with a bit of wiggling your might get through), and just as certainly referenced on two other folios.

    Pace JKP whose effort to re-work this information by claiming the cross is Gregory of Tours’ (in Cygnus), the testimony of the primary evidence is both plain and consistent: the figure is meant for the South marker and if it isn’t exactly our Crux, it can only be the ‘false crux’ which also remained wholly unknown in that form to European astronomers until about a century after our ms was made.

    It was a constellation very well known to the southern hemisphere, and especially to the mariners – in which context (just btw) I once illustrated a mariner’s loh that was inscribed in Gujerati. The author of the academic article which included that picture was kind enough to permit re-use, and to interpret the nearly incomprehensible script as a list of place-names. The problem about Gujerati is that it was a mercantile script which didn’t really come into use until after the Vms was made. I admit I was disappointed, but theories are meant to be disposable tools, not credos. 🙂

  107. Mark,

    I put numbers in the relevant post. It is not clear from your message which ones you don’t agree with.

  108. Mark Knowles on January 13, 2018 at 1:06 pm said:

    Rene: I suppose I should start by saying that I have spent time studying the cipher alphabets in the Tranchedino in great detail to look at the variety of characters shared with the Voynich, of which there are quite a few examples. I have also looked at the way that those characters are composed out of simpler characters. I say this so you understand my perspective.

    When you include the typical as well as the rare characters which appear on different lists for the Voynich alphabet there seem to be a sizeable number of characters indeed. I haven’t attempted to count them all, but I would be interested in you telling me how many characters you think there are.

    Coming from the perspective of the diplomatic cipher alphabets, clearly we know there are differences in the ciphers, so this is just a parallel, characters can correspond to rare symbols or rare double letters or infrequently used null characters or infrequently used letter substitutes. So I would think one could potentially expect there to be quite a few rare characters.

  109. Mark Knowles on January 13, 2018 at 1:14 pm said:

    Rene: You say “The result is a character set of well over 50 for each cipher.” when refering to diolomatic ciphers. However when referring to the Voynich you say “The Voynich MS only has some 25-35 different characters, depending how one counts the composites.”

    The number of characters you give for the Voynich seems a gross underestimate from what I have seen; maybe I have missed something. In fact a figure of over 60 seems plausible.

  110. Mark Knowles on January 13, 2018 at 2:24 pm said:

    Nick: If we were talking about diplomatic ciphers the much more frequent occurrence of the “4o” at the beginning of words could merely reflect that it maps to a letter that is more likely to be found at the beginning of a word or I suppose it could be a null the author often placed at the beginning of a word.

    As I have repeated there are clear differences with a diplomatic cipher, but that would be my interpretation looking through that lens.

  111. Mark: the piece of the puzzle you seem to have missed out is that there are a number of different ways the Voynichese shapes can be pieced together (e.g. is EVA cth 1, 2, or 3 letters? etc), and there is no consensus about which way is likely to be the correct one – indeed, I think it would be fair to say that if there was a consensus, we would probably have solved the mystery a century ago.

    So when you present your thoughts based on the particular way of piecing the Voynichese shapes together that seems best to you, you’re assuming that everyone else buys in to that particular way, which they don’t. From one perspective, Rene Zandbergen would be right to say that Voynichese uses too few letter shapes compared to typical diplomatic ciphers of 1440 onwards (which is essentially what we have evidence for, from the Urbino cipher ledger onwards): but from a different perspective, if you accept the radiocarbon date range as being very close to the truth, you’d be more interested in comparing it to ciphers in the 1400-1440 date range, of which we currently have only a few. My own perspective is different yet again. 😉

  112. Mark Knowles on January 13, 2018 at 3:31 pm said:

    Nick: I am sure other people may not share my perspective, especially if they doubt the diplomatic cipher connection. I am really only giving my interpretation based on a close examination of diplomatic cipher alphabets and how the characters are constructed.

    In the “In Milano” cipher alphabet, which I pick as one that has common characters with the Voynich though there are many in the Tranchedino that do as well, we see a large number of characters in the cipher alphabet and of course this is dated 1435. So the question becomes before 1435 rather than 1440, I guess. Certainly broadly speaking I think it is fair to say that earlier ciphers one would expect to have smaller numbers of characters in their alphabet. My point is that I doubt that the Voynich alphabet is as small as Rene says, this statement is not based on any assumptions of the number of characters in cipher alphabets prior to 1435, but rather the different Voynich characters listed by a variety of people and their consistency with my own framework for character construction.

    I am telling it as I see it at this time with no expectation that others will automatically agree with me.

    As a quick review I would say many characters in the Tranchedino are the same as the Voynich. Few characters in the Urbino are shared with the Voynich. In the Meister there is only one cipher alphabet I believe that has a few characters in common with the Voynich characters and that is the “In Milano”.

  113. Mark Knowles on January 13, 2018 at 3:42 pm said:

    Nick: The 1424 Florence cipher alphabet in Meister has a quite a lot of characters though only a few in common with the Voynich.

    Nevertheless I am still focused on finding early 15th century Visconti ciphers as this, as you know, is where I expect to see the most in common with the Voynich cipher alphabet.

  114. Mark,

    if you look at the Eva tables in , then keep in mind that all ‘extended Eva characters’ occur less than 10 times, in a text of around 160,000 characters. They play no role in the statistics. It is as if a Tranchedino-type cipher was used in such a way that for each plain text character, always the same cipher text character was used. This effectively reduces the cipher to simple substitution, and that won’t work.

    That is only one reason why this principle was not used for encoding the Voynich MS text. The next simplest statistic is the single character frequency distribution, which also should look very different if a Tranchedino-type ciper were used, and the same holds for character pair statistics.

    The best one can say is that these tables *may* have inspired the creator of the Voynich alphabet.

    However, there is another fundamental point, which is quite hard to quantify, namely that the Voynich alphabet allows to write the text as a flowing script. This looks like the consequence of a deliberate design effort.
    Most (but indeed not all) encrypted texts using ‘funny’ symbols look like a string of symbols, not flowing text.

    If you browse through this (well-known) manuscript:
    you get some good examples.

  115. J.K. Petersen on January 13, 2018 at 5:24 pm said:

    Mark wrote: “I meant just to point out that from my perspective the links between Indic languages and the Voynich are tenuous.”

    Oy. Again you missed the point.

    I was pointing out the correspondence between Medieval Latin scribal abbreviations and Medieval Indic scribal abbreviations and how they were written and expanded in the middle ages, not positing a direct link between the VMS and Indic languages.

    As an example I illustrated the Latin “ra” and the Indic “ra”, which are the same glyph with the same general sound correspondence, AND one cannot ignore this because the VMS also uses the same glyph *and* its variations and might, in fact expand in the same way as was characteristic in the 15th century, in both Europe and the Middle East (and parts of western India).

    To understand *Medieval Latin glyphs* (which were used to write more than 20 different languages) and how scribes THOUGHT about glyphs and ligatures and abbreviations, it is wise to acknowledge how common this mode of thinking was across diverse cultures and alphabets.

  116. J.K. Petersen on January 13, 2018 at 5:41 pm said:

    Mark wrote: “I am really only giving my interpretation based on a close examination of diplomatic cipher alphabets and how the characters are constructed.”

    I’m sorry, Mark, but comparing the VMS glyphs to diplomatic cipher glyphs without learning Latin scribal conventions is like comparing two similar-looking people and trying to determine if they are brothers without looking into their family histories.

    I have a copy of Tranchedino, by the way. Even if the VMS borrowed glyph ideas from these ciphers (which is not at all certain since the glyphs almost all derive from common sources such as Latin, math and astrology symbols and, to a smaller extent, Greek), there is no resemblance whatsoever between the structure of the VMS text and the diplomatic ciphers.

  117. Mark Knowles on January 13, 2018 at 7:47 pm said:


    You say “Comparing the VMS glyphs to diplomatic cipher glyphs without learning Latin scribal conventions is like comparing two similar-looking people and trying to determine if they are brothers without looking into their family histories.”

    Frankly that is nonsense.

    To say that the symbol which looks like a:


    for example, in the Voynich exactly resembles symbols looking the same way in diplomatic cipher alphabets requires no knowledge of Latin scribal conventions. Moreover it even requires no knowledge of the Latin alphabet, maybe even no knowledge of any alphabet.

    Seeing a symbol in the Voynich alphabet and observing a symbol which looks the same way in a diplomatic cipher alphabet requires reasonably good eyes nothing more.

    The brotherhood analogy is a flawed one as genetic inheritance works quite differently and gene manifestation is very complicated and genetic difference may not be obvious. Of course you can determine if two people are brothers by comparing their DNA without looking at their family history.

  118. Mark Knowles on January 13, 2018 at 8:04 pm said:

    Rene: I think you have misunderstood me; I am not saying that the Voynich cipher is the same as any diplomatic cipher I am aware of. What I have focused on is the parallels between the Voynich cipher alphabet and the diplomatic cipher alphabets. I think is important at first to separate those two things:similarities in the alphabets and similarities in the cipher.

    If you want my current guess, it is that we have a diplomatic cipher with 1 or 2 significant innovations which are responcible for the statistics you observe. So the author had knowledge of diplomatic ciphers and familarity with their alphabets and had learnt how to write them; then he/she developed these ciphers further to corporate his/her own encipherment ideas.

    However what I have been focussing on here is the significant parallels in the alphabets; this can be treated as a separate question I think.

  119. Mark Knowles on January 13, 2018 at 8:15 pm said:

    Referring to the character which I describe as “4P”, but others might think looks like a “qp”. When I first saw this character in the Voynich alphabet I guessed that it was supposed to be symmetrical about the vertical which it pretty much looked as though it was, without staring carefully at it. So when looking for shared characters with diplomatic cipher alphabets I naturally looked to see if I could find this character; it does occur in some cipher alphabets though it is less common than some other shared characters. When looking at this character in the diplomatic cipher alphabets I thought it looked more like a “4P”than a “qp”; this also seemed to make much more sense as prefixing a “4” to the start of a character was much more normal than having “q” at the start in the characters in diplomatic cipher alphabets. So a “4” at the start of the character made sense visually and logically to me. So I returned to look at the character in the Voynich cipher alphabet and was surprised and gratified to see that the start of the character was clearly angulated exactly like a “4” and not curved like a “q”; in fact on closer inspection it was a “4” just as it was in the diplomatic cipher alphabets. This further affirmed in my mind my notion that the Voynich alphabet is related to diplomatic cipher alphabets.

    Not only was it clear to me in the “4P”, but also looking closely at the gallows characters it was clear the top left character was also angular like a “4” and indeed was a “4”. Similarly some people may view the “4o” as a “qo”, but in the context of diplomatic cipher alphabets that makes little sense.

    I should make it clear that this is my analysis and that I am aware it is not universally held.

  120. Mark Knowles on January 13, 2018 at 8:23 pm said:


    If you want to make your case that all the commonalities between diplomatic cipher alphabets and the Voynich can be found in Latin Scribal Conventions then you need to provide examples. On your blog there are very few examples. I know you said that you had provided only 5%, but without more your case looks weak to me.

  121. J.K. Petersen on January 14, 2018 at 3:00 am said:

    Win a lottery for me, Mark, and I’ll post them all. I’m running two businesses. I work 90+ hours a week.

    And looking only at the shapes and comparing them, once again, means nothing.

    If diplomatic ciphers and the VMS took their inspiration from Latin/math/astrology/Greek, then there is not necessarily a link between them, and studying the original inspiration (Latin scribal conventions) may be more enlightening than studying the superficial similarities between them.

    Either way, ignorance of the underlying Latin conventions does not give one a full deck from which to evaluate whether the similarities between the ciphers and the VMS might be causal or coincidental.

  122. Mark Knowles on January 14, 2018 at 5:56 am said:

    JKP: You think the key common ancestor of the Voynich and the diplomatic cipher alphabets is the Latin Scribal Conventions, that is why you think they are so important and crucial to understanding the Voynich. However you don’t provide the evidence to support your assertion. Hence this is why I do not find an understanding of Latin Scribal Conventions important.

    If you thought the source of the Voynich was the Georgian language then that would be crucial to understanding the Voynich and you would be telling that I need to understand that to understand the commonalities between diplomatic cipher alphabets and the Voynich. Since I don’t think Georgian was the source, understanding Georgian becomes of not so much less importance than understanding Latin Scribal Conventions.

    I wanted to convey the idea that for me the subject of Indic languages was very much a digression from the question of the Voynich.

  123. Mark Knowles on January 14, 2018 at 6:13 am said:

    Rene: When you say:

    “The best one can say is that these tables *may* have inspired the creator of the Voynich alphabet.”

    That is exactly the issue that I was raising: “How do we determine objectively if there is a relationship between the Voynich alphabet and the diplomatic cipher alphabets other than coincidental?”

    You use the word “*may*”. However one could use that word with so much to do with the Voynich. The more interesting and challenging question is that of the likelihood that this or that explanation is correct as this forces one to analyse the hypothesis more carefully and thoughtfully. To say something “*may*” be the case in this context is so noncommittal as to be almost superfluous.

  124. Mark Knowles on January 14, 2018 at 6:29 am said:

    I thought I would give a few examples of how I would represent some Voynich characters including gallows characters:




    These are obviously only visual representations not capable of being spoken. For me personally this could serve as useful in determining what a specific character is when it is unclear and so a step towards creating a framework for all character representation based on diplomatic cipher character representation. Naturally these are my opinions.

    I have largely finished my analysis of diplomatic cipher alphabets until I have some more material to work with.

  125. Mark Knowles on January 14, 2018 at 10:55 am said:

    JKP: Repeating that the key is Latin Scribal Connections isn’t any more persuasive than repeating that the connection is Sanskrit, repetition doesn’t make it so, evidence is the key and I guess until you win the lottery we are not going to get any.
    If you don’t have the time to work on the Voynich don’t do so, but don’t use it as an excuse to justify a flimsy theory.

  126. Mark Knowles on January 14, 2018 at 11:42 am said:

    JKP: I am reallly not trying to be rude, but you commented on my theory saying you are an expert in this area and that I know nothing about this subject. But you present a theory with next to know evidence which unless you win the lottery or someone else win’s the lottery on your behalf you are not going to provide any more of.

    I find this more than a little exasperating. Rather than commenting on my theory I would advise you to put the spare time you have available to work on the Voynich on getting your evidence out there.

    Sorry again for my brusqueness it seems possible to me that Latin Scribal Conventions had a loose influence on some of the character set it the Voynich and diplomatic ciphers. However I think you are very far from demonstrating that everything they have in common is also shared with Latin Scribal Conventions.

  127. Mark,

    you were in the process of arguing that the observation that some relatives of your candidate MS writer had to do with Visconti ciphers, was a potentially important support of your theory.
    So, the logical step is to wonder how much these diplomatic ciphers have to do with the Voynich MS.
    The fact that the Voynich MS was certainly not encrypted using such a method is highly relevant for this question.

    My observation that such ciphers *may* have inspired the creator of the alphabet isn’t saying very much, but it is exactly as far as I would go.

    Finally, I am afraid that you desperately underestimate the accumulated knowledge of JKP with respect to this topic….

  128. Mark Knowles on January 14, 2018 at 1:47 pm said:

    The point for me is this one: Amongst which alphabets cipher or otherwise are there the most commonalities with the Voynich?

    If the answer is diplomatic cipher alphabets then that implies there is more likely to be a connection between the Voynich and those familiar with diplomatic cipher alphabets than anything else, everything else being equal, so we can approach the question without any knowledge of the precise details of the Voynich cipher.

    This would imply that there would very likely be a connection between the Voynich and those writing diplomatic ciphers as they would be the most familiar with diplomatic cipher alphabets.

    Therefore this would support the idea of a very close connection between the author and those writing diplomatic ciphers, especially if the cipher alphabets they were writing were the most similar, to that we see in the Voynich, of all the cipher alphabets. So for example Milanese cipher alphabets look to be generally much more like the Voynich alphabet than those of other Italian States, from what evidence I have seen, implying potentially a link with Milan.

    My hypothesis is that of Milanese cipher alphabets we will see the most commonality with Visconti cipher alphabets of very roughly around 1420 to 1430.
    If my hypothesis proves true this would appear to imply a significant link between the Voynich cipher alphabet and the very close relatives of the person that I am interested in.

    Given that my identification of authorship had nothing to do with ciphers this would seem to be very strong support of my authorship hypothesis.

    To reiterate I see no need to get bogged down by the specifics of the Voynich cipher, which nobody understands, in order to potentially establish a very strong link to diplomatic ciphers or more specifically their alphabets.

    As far as JKP goes he may have a vast accumulated knowledge, but I have seen little relevant evidence of it in this specific context. I am much happier with appeals to evidence rather than appeals to authority. If JKP can provide evidence to support his assertion which is, as best I understand it, that the commonalities between the Voynich and diplomatic ciphers are shared with Latin Scribal Conventions that’s great. If however I am supposed to take it on trust on the basis of his supposed expertise that is not how I operate; I apply that perspective to anyone working on the Voynich as I would expect others to apply to me.

  129. Mark Knowles on January 14, 2018 at 2:21 pm said:


    Additionally if the Voynich author has a connection to those writing diplomatic ciphers it is not unreasonable to think that as well as being influenced by the diplomatic cipher alphabets the author could quite likely have been influenced by diplomatic ciphers themselves.

    Regarding JKP, I am sorry if I seem a bit rough, but appeals to authority over argument or evidence I find really frustrating. When it comes to the Voynich I prefer to analyse things myself and decide on that basis. That is not to say I am not prepares to listen to others when they present arguments and evidence, but if they only say “Believe me I am an expert.” I am much more reluctant.

  130. Mark,
    I has always seemed to me a basic mis-step in the history of this study that no-one had ever done what you seem to be doing: that is, to seek if one can discover the most elegant explanation for the choice of glyphs (if it was a choice). Obviously all alphabets are likely to have an ‘o’ in them, but such universal forms apart, if you can account for all the Voynich glyphs using just a couple of alphabets and/or cipher systems it is likely to be a helpful pointer to where (and when) the written part of the text was formulated.

    I’ve often wondered why this seems not to have been done, given the computing power and enormous amount of available data and technology. Perhaps it’s science fiction, but I’ve often wondered what would happen if a few hundred manuscript-pages from different times, language and so on were fed in as a sort of ‘face recognition’ data-set and then a scan made of some pages of text from the Vms. To adepts that might seem a foolish exercise, but you never know; it might even work. 🙂

  131. Mark, you are asking for advice and opinions, so you get them…

    The entire topic of the writing has of course been looked at over many decades, by many people. Not all of that has been written down, and parts have been written down in places that are not easily accessible.

    D’Imperio already summarised this in a way in the late 1970’s. Nothing conclusive of course. Both the diplomatic ciphers and the Latin abbreviations are equally likely sources.
    In the context of the former you have looked at Tranchedino, but have you looked at Capelli?

    I don’t see any “decision” or conclusion coming out of this general area of investigation, but lots of opinions. Maybe not even after (if ever) the text can be read.

  132. Mark Knowles on January 14, 2018 at 8:56 pm said:

    Rene: Of course advice and opinions are always welcome, I just can’t promise that I will automatically agree with them.

    I would be very happy to look at Capelli. I have tried to find the characters online, but all that I can see is a dictionary with abbreviations, but not showing any actual images of the text corresponding to those abbreviations. And also I have seen one image with what look very vaguely like extended “4P” or “qp” characters.

    However to make a real assessment I would think it is vital to see more real images of characters illustrating the Latin abbreviations. It would be great if someone has the link to where I might find them, so I can make my own analysis and comparisons with the appearance of the actual Voynich characters.

    I am perfectly happy to compare evidence in favour of the two hypotheses, in fact I think that is important.

  133. Mark: I know you’d be “perfectly happy to compare evidence in favour of the two hypotheses”, but it’s not hard to notice that that would currently only highlight your current lack of any Visconti ciphers. Having said that, the Voynichresearch world seems to be almost entirely populated by stone-throwing glass-house-dwellers, so please just take it as a provocation to try to reduce the glassfulness of your walls. 😉

  134. J.K. Petersen on January 14, 2018 at 11:00 pm said:

    Mark, René made a very good point when he asked if you have looked at Capelli.

    I second his suggestion. It would give you have a broader base from which to assess whether the diplomatic ciphers might be directly or indirectly similar, but make sure you look at the full version, not the abbreviated version.

  135. J.K. Petersen on January 15, 2018 at 2:28 am said:

    I just discovered someone has indexed the Capelli abbreviations to make them easier to search:

    Capelli did an excellent job for his time, it’s a must-read book in paleography, but even this doesn’t cover all the common abbreviations. It is, however, a good starting point.

  136. Mark Knowles on January 19, 2018 at 10:30 am said:

    Nick: Quick update on the Genovese enciphered letter.

    I contacted the archive you mentioned and it has been moved again. I was told that Sandra Machiavello from the University of Genoa would know where the letter was, so I have emailed her. I have also contacted Storia Patria Genova to see if they also can help me track down the letter. I explained that I would like a scan or photo of the letter.

    As far as finding more examples of Visconti era enciphered letters I wonder how you found out about this letter as I might give me clues as to where to look for other enciphered letters. Is it just a question of systematically looking at listing of letters from various archives? It doesn’t matter if it is laborious I am happy to it. Do you have any other thoughts as to where to look? I have other avenues to explore, but the more the better.

    I have emailed Andrea Visconti, but haven’t heard back yet.

  137. Mark Knowles on January 19, 2018 at 10:39 am said:

    Nick: I note the letter was an invitation to the marriage of Filippo Maria Visconti. I wonder if you think that other wedding invitation letters are likely to be something to look for.

  138. Mark: I just Googled italian cipher keywords, nothing too clever. The following should find a few more for you, particularly if you search in Google Books:

    “gian galeazzo visconti” OR “filippo maria visconti” “in cifra”

  139. Mark Knowles on January 19, 2018 at 7:04 pm said:

    Rene & JKP:

    So returning to the question->

    Are the Latin Scribal Abbreviations a more likely direct source of the Voynich characters than the diplomatic cipher alphabets?

    The first thing that strikes me is that the convention unsurprisingly in most of the Capelli abbreviations is to have a shortened word, maybe with some symbol, superscript or equivalent added, which is obviously very different from what we see in the Voynich. I would have thought if Latin Abbreviations were the root then we would have seen at least one Voynich “character” which appeared as a shorten word, yet done are like this. Now not all the abbreviations are of this form how, but I wonder if the author really only selected abbreviations that were represented in a way like that which we see in the Voynich.

    If the Voynich “characters” corresponded to real abbreviations then translation should be easy as we should in some cases be able to identify the words. If the abbreviations have no relationship to the meaning then in what sense are they abbreviations? It would seem strange to take abbreviations with a definite meaning and assign a completely different meaning to them.

    By contrast the diplomatic cipher characters have no intrinsic advance meaning. The meaning is assigned largely arbitrarily for a given cipher by the encipherer.

  140. Mark Knowles on January 19, 2018 at 7:07 pm said:

    Rene & JKP:

    We have no idea as to how much or how little the Voynich cipher has in common with diplomatic ciphers as we can’t decipher it. Clearly it seems markedly different from any diplomatic cipher we know certainly in its output if not its mechanics. However despite this there may actually be a lot in common; we just don’t know.

    It seems someone with the mind/brain and interest to write ciphers whether similar or not would be far more likely the kind of person who would have written an enciphered document than someone with a knowledge of latin abbreviations to write an untranslatable document.

  141. Mark Knowles on January 19, 2018 at 7:17 pm said:

    Rene & JKP:

    Obviously we could get into precisely which “characters” are in common between the Voynich and Latin Abbreviations relative to the “characters” in common between the Voynich and Diplomatic Cipher Alphabets. We could count the number of these and possibly their relative frequencies amongst the overall set of which they are a part.

    We could discuss how characters are constructed from smaller parts in the diplomatic ciphers and how abbreviations are normally formed.

    This could be done though getting into this detail may be time consuming.

  142. Mark Knowles on January 19, 2018 at 7:22 pm said:

    Rene & JKP:

    To say that diplomatic cipher alphabets and the Voynich alphabet were both influenced by Latin Scribal Conventions is plausible, but to say that everything they have in common is shared with Latin Scribal Conventions is something quite different. In fact to say Latin Scribal Conventions have more in common with the Voynich than diplomatic cipher alphabets is a third thing. I doubt both the second and third statements are true.

  143. Mark Knowles on January 19, 2018 at 7:32 pm said:

    Rene & JKP:

    In conclusion I am unconvinced that Latin Scribal Conventions were the direct source or influence on the script we see in the Voynich. Clearly the Latin alphabet and the Arabic numerals play a significant part, but there are characters which do not come from either of these and I believe were essentially made up characters with no direct origin.

  144. Mark Knowles on January 19, 2018 at 7:35 pm said:

    Rene & JKP:

    I should say that regarding made up characters, I don’t mean that these were necessarily made up by the author of the Voynich, though some characters may have been, but rather that were made up by earlier diplomatic cipher writers and were inherited from them.

  145. Charlotte Auer on January 19, 2018 at 8:23 pm said:

    I’d just like to add a little comment on the Capelli, and a question about the Visconti cipher(s).

    The so-called Capelli is a huge collection of Latin and Italian abbreviations (and ligatures etc.) and as such an invaluable help for paleographers in the field, but has nothing to do with ciphers per se. That doesn’t necessarily mean that a single abbreviation would never have served as a single glyph in an enciphered text, and some of them may appear here and there in diplomatic and non-diplomatic ciphers, but as a system it’s more than unlikely. Otherwise it would be well known and documented. So far, the study of the Capelli might be a good introduction for amateurs to get familiar with late medieval writing systems, although it’s not of much help for the comprehension of the VM.

    If it comes to the supposed enciphered Visconti letters: as far as I konw there is no physical evidence left of them, they are only mentioned by some early 20th century authors (as an outcome of their studies of the Visconti). Am I wrong / badly informed? If yes, I would appreciate it very much if someone could name me a reliable source for the existence of these ciphers.

  146. Mark Knowles on January 19, 2018 at 9:55 pm said:


    As far as the Capelli, I agree with you, I merely researched them, because Rene and JKP stated that they might be/are the source of the Voynich script.

    On the subject of the enciphered Visconti letters, most Visconti records were destroyed when the Castello di Porta Giova was destroyed. However I know of some letters intercepted by the Papacy. I have read of other letters.

    But more importantly Nick has directed me to a letter in the Genoa archive, see Nick’s link above, which I have contacted and with luck they will email me a photo/scan next week of this particular enciphered letter on the marriage of Filippo Maria Visconti. I have also been researching other archives to contact. If you are interested I would certainly value someone else’s assistance.

  147. J.K. Petersen on January 20, 2018 at 4:45 am said:

    The VMS includes glyphs that are morphologically consistent with Latin letters (a, o, i, possibly u, c…).

    It also includes glyphs that are morphologically consistent with Latin abbreviations and ligatures (EVA-m, EVA-r, EVA-s, EVA-y, EVA-ch, gallows-k and the tail on dain).

    Scribal abbreviations were an integral part of a scribe’s mindset and of the Latin “alphabet” in those days. And not just Latin scribes… Indic scribes applied many of the same concepts, as well. It’s difficult to find a page and sometimes even a paragraph or sentence that does NOT use Latin abbreviations in 15th-century manuscripts.

    Thus, I think Capelli is very relevant to the study of the VMS. If the scribes who penned it were familiar with Latin letter-shapes and Latin abbreviation-shapes (which they clearly were), then it’s possible some of the glyphs are intended to be expanded in the same general way, and no amount of 1-to1 substitution is going to yield useful information about the text.

    Researchers who don’t understand the concepts in Capelli are not going to recognize which glyphs are part of the common Latin repertoire. They will see only the VMS vowel-shapes and EVA-e as Latin and will wrongly assume the others are foreign characters when most of them are not.

  148. Mark,

    not sure if I can answer all the questions, but Charlotte is right. The point of referring to Capelli is that, similarly to the Tranchedino (and other) cipher alphabets, the symbols that have been used for these scribal abbreviations may have inspired a person into putting together the Voynich alphabet.

    Whether the cipher symbols were more important than the abbreviations, or the other way around, is something I would not want to argue. For one thing, there is an even larger group of characters in the Voynich alphabet that look like ‘plain old’ letters or numbers.

    At one point you asked for an example of a text with abbreviations. My favourite one is a MS in the Vatican called Vat.Lat.869:

    I think it is Latin but I can’t read anything.

  149. Helmut Winkler on January 20, 2018 at 10:12 am said:

    Vat. Lat. 869 is a very good example what Beinecke 408 is, highly abbreviated and using its own set of abbr., but B. 408 is a scientific, not a theological ms., you can compare it with the abbr. sets the legal mss. are using., it is the same technique.

  150. Charlotte Auer on January 20, 2018 at 2:51 pm said:

    Rene: yes, is in Latin. It is a philosophical discourse of the soul (de anima) and the animated body, and I can read it. No, to be honest, I have a transcription of some of the most important parts of it before me. I couldn’t read anything either without a little help from old Capelli, and even with the Capelli a good part of it would remain unreadable to me since I’m not trained for that kind of mss. At least it would be a tremendous piece of work to get it transcribed myself.

    JKP: on the contrary, it is absolutely not “difficult to find a page and sometimes even a paragraph or sentence that does NOT use Latin abbreviations in 15th-century manuscripts”. There are thousands of mss not using Latin abbreviations just because they are not wirtten in Latin, but in German, French, Italian and other languages.

    It is clear that most of the scribes got their education in monastic scriptoria, and were of course trained in Latin as the main language not only for books, but also for documents, correspondence and administration. Nevertheless, with the increasing amount of vernacular mss they also got trained in writing their native language as well – by still using the Latin scribal system/alphabet. From a serious paleographic point of view it is therefore not possible to generalize the use of Latin letters, symbols and/or abbreviations and ligatures.

  151. Mark Knowles on January 20, 2018 at 6:53 pm said:

    Rene & JKP:

    I think unfortunately we need to break this down:

    The source of the Voynich alphabet could be medieval algebra or the result of someone musing as they look at the different shapes their spagetti makes as they twirl it on their plate. However the question is what is the most likely source for the shapes we see? I, of course, unlike Rene think this is a question that we will know the answer to.

    I think systematically someone would need to go character by character of the Voynich alphabet and try to explain the origin of each character in terms of either the Latin Scribal Abbreviations or the diplomatic cipher alphabets; I also see no reason not to include the rare characters as part of this process as long as they are clearly readable. In the case where there is no direct analog with the Voynich alphabet there needs to be an explanation provided as to how that character might have been arrived at. I have done and can do this for the diplomatic cipher alphabets, but I have not seen anything of the kind done for the Latin Scribal Abbreviations.

  152. Mark Knowles on January 20, 2018 at 7:11 pm said:

    Rene & JKP:

    Obviously someone could say that the author made up all the characters from his/her own imagination without any source of inspiration. This is possible, but the question is what is the likelihood of any explanation being true?

    What is the source of best fit?

    Again I think this can be amenable to more rigourous analysis by considering frequencies and probabilities. Question like how common a certain abbreviation was and how common a diplomatic cipher character was as well as how common a Voynich character is, are worth asking.

  153. Mark Knowles on January 20, 2018 at 7:34 pm said:


    So to clarify, do you think that the Voynich characters were derived from Latin abbreviations whilst not retaining the meaning of the abbreviations and subsequently used as arbitrary characters in a cipher?
    Or do you think it is not a cipher?

  154. Mark,

    I am not trying to stop you in any way, but it is important to be aware all the time of the difference between knowing and thinking/suspecting.

  155. Charlotte Auer on January 20, 2018 at 9:27 pm said:


    I really don’t want to offend you, but I’m afraid that your obvious lack of paleographic experience will lead you into the wrong direction. There is no such alternative as “either the Latin Scribal Abbreviations or the diplomatic cipher alphabets” because this would exclude most characters which are simple “plain old” Latin characters, as Rene mentioned them above. Even the so-called Gallows have a clear origin and purpose. An explanation of the origin of the Voynich alphabet is not needed at all because every experienced paleographer can recognize its origin at a first glance. The secret of the writing system of the VM lies in the use of the characters and not in their origin. Plain and simple.

    Sorry for writing this, but I hope it will help you to see the undeniable paelographical facts a little clearer now.

    Good luck for your Visconti research!

  156. Mark Knowles on January 20, 2018 at 9:41 pm said:


    I understand, but I feel in your approach you do not view things in terms of probabilities, which can make this terminology rather vague.

    What do “knowing”, “thinking” or “suspecting” mean in a scientific context? I am not quite sure.

    When you say X “may” be true, “may” covers all probabilities from it is true to it is not true. So the statement is virtually a tautology i.e. without wishing to cause offense, almost not worth saying at all.

    Whilst this is not a pure science I think we must try to keep our eye or aim on how we can resolve these questions pertaining to the Voynich rigourously. It may not always be possible, but rigour should be our goal in trying to determine which theories are more or less probable. We need to analyse arguments and the assumptions on which they are based and I think by doing these things progress can be made.

    I feel your “it could be this or it could be that who can say” approach is almost a barrier to progress as it tacitly implies that these question are like great philosophical/existential mysteries which can never be resolved.

    I am not trying to be hard on you, but I think we all need to positively try to think of ways we can start to resolve these many questions rather than shrugging our shoulders.

Leave a Reply

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

Post navigation