The Voynich Ninja

I am working on a cipher at the moment and it has been very illuminating in lots of different ways. It got me thinking about the questions of corrections as I discussed in a different thread. I was just watching a video of Claire Bowern talking about the Voynich and she mentioned the subject of the difficulty of distinguishing between certain characters and deciding if in a given instance we have 1 or 2 different characters. I obviously was aware of this subject, but the video reminded me of it.

I have found the same problem in the cipher from the same time that I am deciphering. Sometimes it is not clear if two characters are the same or different. In fact this seems to be a common problem. So the Voynich is far from unique in this regard. I am not sure if there is a general method or approach applicable to ciphers of that time to discern the true underlying characters and avoid unnecessary duplication and incorrect conflation. I do think there is value in the exercise of trying to decipher old ciphers as it makes one consider simple common issues like these as well as more complex parallels.

One thing which may be harder in ciphers other than the Voynich is identifying spaces. I get the impression that spaces are fairly well defined in the Voynich. In some other ciphers of the time that may not be the case which adds a difficulty in deciphering them, so maybe we should be grateful at least for that when it comes to Voynichese.

Of course, this assumes that Voynichese spaces correspond to real spaces.

(28-12-2022, 12:42 AM)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.One thing which may be harder in ciphers other than the Voynich is identifying spaces. 

Hi, Mark:

I wanted to confirm my experience with 15th C ciphers.  Even if it is just a "simple substitution cipher" the only thing I see consistently is inconsistency.

What I mean by that is: (1) the cipher shapes are not drawn consistently (thus making homophonic keys where maybe it is or maybe it isn't?), (2) word spaces are inconsistent -- particularly at line breaks, where you'd think that might be a "stronger" space but instead what I see (although it is dependent on the scribe) the cipher writer just keeps going until space runs out and then finishes the rest of the word on the next line (no difference if it is Latin or venacular plain text -- this is seen in all cases); and (3) the often recognized Medieval spelling and grammar inconsistencies of the time for both Latin and venacular not surprisingly continue in the ciphers' ultimate plain text.  It is particularly discouraging when you are trying to decipher maybe a few words or a short text and you can see in the plaintext that the same non-ciphered word is spelled different ways in the same line.  Makes things really hard.

Thus, I think that attacking even a "simple cipher" with a modern mindset is just impossible because there was just no expectation of communicating in writing in a consistent manner.  I really don't think this attitude changed until after the printing press had been around for a while and even then it took some time (probably 100s of years) before some consistency could be counted on.  And this doesn't even take into account all the possible abbreviations and variations that can be introduced with that extra layer (again, both in Latin and in vernacular).

So people can sneer at how the ciphers (other than the Voynich, if it is a cipher) are "simple" but in reality actually figuring out what was encoded is a multiple level puzzle that in the community (and even on this board, dare I say it) just doesn't get enough respect.  Until you've tried a few yourself --- then I want to hear how "simple" you think these things really are.  I found it gave me a very different view into the deciphering process and recommend it very highly to anyone who is working with the text.   There are plenty that have never been figured out . . .

Okay, off my soap box but I do recommend maybe making a New Year's Resolution to try a few and prove to yourself how really shaky the understanding of how ciphers work was in this time period.  And a very lossy cipher (like the Voynich probably is, if that is what is going on) just doesn't shock me at all because I haven't found any indication that even the most "careful" cipher writers displayed any understanding of how sense could be so easily lost by how they actually did ciphers.  This is particularly true with non-diplomatic ones -- which is where I have focused my efforts.

But I'd be interested in your view, Mark, on how it plays out in diplomatic correspondence.

Thanks for reading my rant,

Michelle

(28-12-2022, 02:28 AM)MichelleL11 Wrote: You are not allowed to view links. Register or Login to view....you can see in the plaintext that the same non-ciphered word is spelled different ways in the same line

So we can expect to find some words in the VM written in different ways, apart from abbreviations and ligatures?

(28-12-2022, 12:22 PM)Ruby Novacna Wrote: You are not allowed to view links. Register or Login to view.

(28-12-2022, 02:28 AM)MichelleL11 Wrote: You are not allowed to view links. Register or Login to view....you can see in the plaintext that the same non-ciphered word is spelled different ways in the same line

So we can expect to find some words in the VM written in different ways, apart from abbreviations and ligatures?

Yes, it is well documented how spelling and grammar were very fluid at the time.  This is perhaps best studied in Latin, which had been essentially “reconstituted” from what classic texts were available and then “adapted” through cumulative interpretations and regional idiosyncrasies that were perpetuated from teacher to student.  

Some attempts to quantify these issues in medieval corpus have been made, see for example, Timo Korkiakangas, You are not allowed to view links. Register or Login to view., which l current only have access to the abstract.

But l’d like to emphasize that these issues are just an extra layer on top of the fundamental problem of the lack of grammatical “word patterns” in the Voynich text that preclude it, in my belief, from being a 1:1 substitution in any likely language.  If it is a cipher, some other text manipulations have occurred - likely more than one.  This is not my idea by any shape or form but something that becomes a necessity as you work with the text under the cipher theory.

(28-12-2022, 02:28 AM)MichelleL11 Wrote: You are not allowed to view links. Register or Login to view.

(28-12-2022, 12:42 AM)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.One thing which may be harder in ciphers other than the Voynich is identifying spaces. 

Even if it is just a "simple substitution cipher" the only thing I see consistently is inconsistency.

What I mean by that is: (1) the cipher shapes are not drawn consistently (thus making homophonic keys where maybe it is or maybe it isn't?), (2) word spaces are inconsistent -- particularly at line breaks, where you'd think that might be a "stronger" space but instead what I see (although it is dependent on the scribe) the cipher writer just keeps going until space runs out and then finishes the rest of the word on the next line (no difference if it is Latin or venacular plain text -- this is seen in all cases); and (3) the often recognized Medieval spelling and grammar inconsistencies of the time for both Latin and venacular not surprisingly continue in the ciphers' ultimate plain text.  It is particularly discouraging when you are trying to decipher maybe a few words or a short text and you can see in the plaintext that the same non-ciphered word is spelled different ways in the same line.  Makes things really hard.

And this doesn't even take into account all the possible abbreviations and variations that can be introduced with that extra layer (again, both in Latin and in vernacular).

So people can sneer at how the ciphers (other than the Voynich, if it is a cipher) are "simple" but in reality actually figuring out what was encoded is a multiple level puzzle that in the community (and even on this board, dare I say it) just doesn't get enough respect.  Until you've tried a few yourself --- then I want to hear how "simple" you think these things really are.  I found it gave me a very different view into the deciphering process and recommend it very highly to anyone who is working with the text.

There are plenty that have never been figured out . . .

This is particularly true with non-diplomatic ones -- which is where I have focused my efforts.

But I'd be interested in your view, Mark, on how it plays out in diplomatic correspondence.

Hi Michelle

I will try to give what you have written more thought and I will write a more detailed response. But so far I would say that I absolutely agree with most of what you have written, as is very often the case.

I think ranting from time to time is a good thing. As anyone knows I do plenty of it myself.

The most important point that you make I think is simply when you say: "There are plenty that have never been figured out." That is absolutely true. I would further add that there are plenty of ciphers from the time of the Voynich manuscript that are unknown to the Voynich community. In short, there is plenty more ciphers to be found. But they will only be found if people look for them.

Apart from a very small number of other individuals namely: Nick Pelling and David Scheers most Voynich researchers knowledge of ciphers from the time of the Voynich manuscript is at most having glanced at "Die Anfange.." by Meister and possibly the Tranchedino Cipher Ledger and possibly Bischoff's article and maybe the Steno Letter. Every so often someone will dredge up a cipher from the very end of the 15th century or the 16th century as if it constitutes a significant discovery whereas for me it isn't very different from them locating a 19th century cipher in its relevance. I created a poll, with the help of Koen, on the forum some time ago to get an idea of the percentage of people who think the Voynich is written in cipher and to my surprise a significant majority did and yet paradoxically there seems to be so little interest in ciphers from the time of the Voynich manuscript; to be fair most of what remains to be found is in archives and not online.(How about that for a rant?)

If I could mobilise people a little to do some searching that would push things forward.

Thanks again

Mark

(28-12-2022, 03:54 PM)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.paradoxically there seems to be so little interest in ciphers from the time of the Voynich manuscript

Correct me if I'm wrong, but so far I am not aware of any cipher that illuminates what could be going on with the VM. I suspect there would be more interest in this if there was a closer connection between the VM and the known ciphers of the time. Right now it feels like looking for a needle in a haystack where you don't even know if there is any needle at all. 

Combing through archives is typically the kind of undertaking that has to be a passion project for the researcher to stay motivated. I think this explains why many people seem not interested: it is the search for ciphers itself that is more rewarding than the relevance for Voynich research so far. It is more like a separate hobby that requires a lot of time and dedication. Of course, studying ciphers in archives might be exactly what will solve the VM one day, but so far the results are discouraging (granted, this could be said about any approach).

(28-12-2022, 07:51 PM)Koen G Wrote: You are not allowed to view links. Register or Login to view.Correct me if I'm wrong, but so far I am not aware of any cipher that illuminates what could be going on with the VM. I suspect there would be more interest in this if there was a closer connection between the VM and the known ciphers of the time. Right now it feels like looking for a needle in a haystack where you don't even know if there is any needle at all. 

The reason why I'm drawn to trying to understand on a fundamental level what ciphers have been found and to look at any others that can be located is because of my experience of being in the "innovation business" for my entire career.  If there is one truth I've learned is that nothing innovative comes from out of the blue.  Any one who claims that or tries to present their work as being disconnected from others is either not telling the truth and/or not giving credit to previous ideas (often, sadly, on purpose).

Thus, it makes sense to me to understand how ciphers worked at the time and to really study the small amount of written evidence we have around and as immediately after the carbon dating as we can get on the topic if the theory that the Voynich was written in a cipher is to be fully investigated.

Of course, as you say, Koen, nothing connected to the Voynich text may survive -- but the last thing I want to do is overlook what might be there for the mere fact that I didn't recognize it for what it was.

I know that some on the board don't agree with this sentiment, but I have been convinced that the written word functioned fundamentally differently in the medieval period than currently (both as communication and socially through possession) and the greater the understanding of this mindset by the person(s) trying to interpret the Voynich text to the person(s) who created it, the more likely it could be understood.

Of course, I would not be running after this understanding if I didn't find it incredibly interesting in its own right.

Will this work on medieval ciphers solve the Voynich text?  Likely not.  But I am trying to increase the chances that I can contribute by learning these aspects that are interesting to me, that make sense to me to be part of what a solution would take, and collaborating with others that see value in this approach.

(28-12-2022, 07:51 PM)Koen G Wrote: You are not allowed to view links. Register or Login to view.

(28-12-2022, 03:54 PM)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.paradoxically there seems to be so little interest in ciphers from the time of the Voynich manuscript

Correct me if I'm wrong, but so far I am not aware of any cipher that illuminates what could be going on with the VM.

You are right that you are not aware of any cipher that illuminates what could be going on with the VM. But you haven't looked, so of course you aren't aware.

I increasingly think that the ciphers I have been studying could definitely have implications for the Voynich cipher. In a 1440 cipher key almost all substitutions are for Arabic numerals, which makes it a verbose cipher. However more interestingly a 1424 cipher has many substitutions for Roman Numerals; I noted when Claire Bowern did one of her presentations or wrote in one of her papers that she stated that Roman Numeral substitutions would produce many of the statistics found in the Voynich. So I think there is definitely some contemporary precedent with the Voynich manuscript for the use of verbose ciphers, which a Roman Numeral substitution cipher is. More interestingly unlike other ciphers that I have seen specific substring substitutions are being used.

By substring substitutions I mean substitutions for common word substrings. Using the English language as an example(the plaintext of the 1424 cipher is naturally latin):

"str" is a common start to some words in English. It is not an English word in and of itself, but a common part of a word.

"ing" is a common end to some words in English. It is not an English word in and of itself, but a common part of a word.

So consider:

"str"    ->     XIV
and
"ing"    ->     45
then
"string"   ->   XIV45

This doesn't preclude simple substitution as well, so for example:

"s"     ->     V
then
"sing"    ->     V45

So any word can be constructed by that method. Words in English and many languages are often constructed with standard prefixes and suffixes.

This is completely consistent with the 1424 cipher that I have mentioned and I would be very interested in anyone who can show me that that wouldn't produce the kind of statistics we see in the Voynich.(It is not necessary to use Roman or Arabic numerals )

I am keenly working on locating more ciphers from that time, but it is a lonely task. As is clear to some I am particularly interested in Milanese ciphers from the early 1430s.

For the 1424 cipher the 3 intercepted letters from which the cipher key can be derived can be found in:

Dieci di Balia 7
Archivio di Stato di Firenze

They are Milanese, but we're intercepted by the Florentines during the First War in Lombardy, which was fought between the Duchy of Milan and the Republic of Florence.

The late 19th century Florentine Archivist, Abbot Pietro Gabrielli, made a very good attempt, though not complete, at deciphering these letters. The cipher keys he produced can be found in his work "Crittographia Fiorentina" kept in the Archivio di Stato di Firenze. I am currently working on making some improvements to his decipherment. If I had a time machine I would love to travel back to visit Abbot Gabrielli and show him the Voynich manuscript and see what he made of it, of course he would not have seen it himself. He deciphered over 1300 historical ciphers in Italian archives.

Anybody interested in seeing the 1424 ciphers I have referred to can contact the Archivio di Stato di Firenze and request photoreproductions, which should be very inexpensive.(If you wish to make them public you will have to get permission from the archives and there may be a charge for them.)

I am working on a paper on early 15th century cryptography, expanding on the presentation that I gave in October, which will certainly go into more detail. I don't know when that will be published yet, although I am writing it in part in response to a request from a journal.

Yes, such a cipher might work: we know the VMS "word" lengths are binomial distributed, so that fits with them being numbers. A couple of properties would need to be reproduced: a) removing the first glyph/character/symbol from the word often makes another valid word (this works for your XIV45 example), and b) (more difficult) the frequent occurrences of word repetitions. For b) it's hard to explain those if there is a one to one mapping between the plaintext and the ciphered form, right?