The Voynich Ninja

(29-03-2019, 11:08 PM)davidjackson Wrote: You are not allowed to view links. Register or Login to view.OK, let's see if I can get this straight.

Your theory is that Voynichese is a transliteration of Judaeo-Greek, which I understand to be known more formally as Yevanic.
Yevanic is a dialect of Greek written using the Hebrew alphabet.

So far, so good. Yevanic is mutually intelligible with Greek; and the conquered Moors of Spain did the same thing with Arabic (the literary tradition of the aljamiadas). No doubt there are some differences, it isn't straight Greek, but I understand that essentially a Greek and a Yevanic can chat perfectly happily together.

Two questions thus occur to me. Maybe I'm just not understanding the system properly.

• Why the additional translation into the Greek alphabet? Essentially, the word is the same in Yevanic and Greek, just that two different scripts are used, unless we are looking at a dialect argot or Hebrew loan word. Are you saying that the user was a Greek writer who transliterated his text twice? Otherwise, we seem to have added an arbitrary layer of transliteration that isn't needed. Voynichese -> Yevanic -> Greek --> English can logically be distilled into Voynich -> Yevanic --> English.
  If the underlying text is Greek, then we don't need Yevanic. If it is Yevanic, we don't need the Greek.
  
  
• The whole point of Yevanic was because the script was sacred to the users. They used the sacred Hebrew characters to write down their (new) mother tongue, ie Greek. Transliterating Hebrew into Voynichese defeats the purpose of this. What's more, I would assume that an orthodox Yevanic (Jew) wouldn't do this - they would go from Greek and keep the sacred script intact. Of course, this is subjective and we can't prove this. It is a consideration however - it seems more likely that the Greek variant word would have been used rather than the Yevanic. It also seems strange that they would not include the diacritics, and would invert the script direction (left to right as per Latin, rather than right to left as per Hebrew / Yevanic).
  

.

David, these are good questions, and the issues are important to discuss and consider and understand, and I appreciate them.

Yevanic or Judaeo-Greek is unfortunately extremely poorly attested. We know the language has existed as a colloquial spoken vernacular for millennia, but it was rarely written down at all, because those who spoke it, mainly did their reading and writing in Hebrew. Medieval vernacular Yevanic is essentially unattested in written form (unless the Voynich MS is written in it!). The oldest extant Yevanic text is a 1547 polyglot Bible in Hebrew, Ladino (Judaeo-Spanish), and Yevanic. We also have medieval Greek translations of the Bible written in the Hebrew script, but they are not in Yevanic per se. Still they provide information about how the Greek language was written in Hebrew letters.

Thus the only way to proceed is to make a series of plausible hypotheses about how medieval vernacular Yevanic would have been likely to have been written, if and when it was written. To do this we have to make use of our much more extensive knowledge of medieval Greek, although here too medieval vernacular Greek is not nearly as well attested as other registers, and also not nearly as well attested as vernacular Greek in other periods both more ancient and more modern. Strikingly, I have read that some scholars have used even more modern information about Yevanic, as a source of information for sound changes in medieval vernacular Greek, because Yevanic did not attempt to preserve classical Greek spellings in the same way that most medieval Byzantine Greek writers did!

In short, we "need the Greek" because our sources of information about it are many orders of magnitude more abundant than our sources of information about Yevanic. Then we must hypothesize about what the medieval Yevanic forms would have been likely to have looked like. I note that Ventris & Chadwick essentially did the same thing for Mycenaean Greek in their successful decipherment of Linear B. They certainly didn't have any other attested examples of Mycenaean Greek to compare it to!

David, I will reply to your second point later, as soon as I have the time to do so. Thank you again for your very important questions, about issues that we will need to understand much better if we are to make significant progress in this direction.

Geoffrey

David, while reading up on Yevanic I had much the same questions as the ones you put forward. Yevanic is essentially Greek, and they wrote in Hebrew script precisely because they needed their writings to be in this script.

Now in defense of the principle behind Geoffrey's theory, one could say that Yevanic shows how certain communities were accustomed to writing non-Semitic languages in Semitic script. As you say, there are other examples of this as well.

This might provide some cultural touchstone for Voynichese being a hypothetical language x written in abjad form.

Now I'll be honest, I find it difficult to assess how much freedom your system allows you in interpretation and whether or not it's too much. We've seen systems before which allowed the user to turn any string of Voynichese into virtually any translation.

Your system also allows for a lot of freedom, but I'm not sure if it's in the "still possible" zone. That's why I'd like to see how it handles labels. My reasoning is that the reader of the VM must have been able to understand the labels relatively unambiguously, otherwise there is no point in labelling things.

(30-03-2019, 12:27 AM)Koen G Wrote: You are not allowed to view links. Register or Login to view.David, while reading up on Yevanic I had much the same questions as the ones you put forward. Yevanic is essentially Greek, and they wrote in Hebrew script precisely because they needed their writings to be in this script.

Now in defense of the principle behind Geoffrey's theory, one could say that Yevanic shows how certain communities were accustomed to writing non-Semitic languages in Semitic script. As you say, there are other examples of this as well.

This might provide some cultural touchstone for Voynichese being a hypothetical language x written in abjad form.

Now I'll be honest, I find it difficult to assess how much freedom your system allows you in interpretation and whether or not it's too much. We've seen systems before which allowed the user to turn any string of Voynichese into virtually any translation.

Your system also allows for a lot of freedom, but I'm not sure if it's in the "still possible" zone. That's why I'd like to see how it handles labels. My reasoning is that the reader of the VM must have been able to understand the labels relatively unambiguously, otherwise there is no point in labelling things.

Koen, thank you for this very clear explanation of your evaluation of my theory at this stage. I appreciate your clarity very much.

I also now have a better understanding of your interest in the labels, such as the ones on You are not allowed to view links. Register or Login to view. , You are not allowed to view links. Register or Login to view. , You are not allowed to view links. Register or Login to view. , You are not allowed to view links. Register or Login to view. , etc. I do see the logic of your reasoning for this.

However, we have to keep in mind that in order for there to be a point in labeling things, the only necessity is that *a late medieval Judaeo-Greek physician*, or some similar person or people, must have been able to understand the labels relatively unambiguously. But we do not know how much distinctive knowledge and cultural context such a person or people would have had at that time, in that place, and with that cultural background. In other words, maybe such people in that time and place were able to understand the labels, but that does not mean that anyone today has the knowledge and background to understand the labels.

Of course, one could argue the same thing about the entire Voynich MS. But I still believe there is a better chance of figuring out the meaning of lines and paragraphs of text with grammar, such as some of the lines that I have posted in this thread, than there is of figuring out the meanings of individual one-word or two-word labels with no grammatical context to assist us.

As a comparison, there still remain undeciphered labels and other items in Linear B. Ventris & Chadwick deciphered the script, and identified the language as Mycenaean Greek. But certain items and words and labels did not have enough context clues for anyone to identify them, even knowing which language it was and even having deciphered the script. For example, the Linear B syllabary script literal reading of "rope" is ko-no. Believe it or not, this represents Greek skhoinos!! I have not proposed any such extreme divergence between representation and actual word in my Voynich theory, as the divergence that we see in this example. In this case, it must have been a picture of a rope that allowed them to make the identification anyway. But in other cases, the context is not so clear, and when the phonetic connection is as obscure as "ko-no" = "skhoinos", you can imagine that it may not be possible to decipher the meaning of the labels without sufficient context.

I fear that may be the case with some of the plant, herb, and root labels in the Voynich MS. However, as I wrote before, we are very lucky that the language is Greek at all, because we have Galen and Dioscorides and Theophrastus as well as many Byzantine and other medieval Greek herbals, etc., to assist us and to guide us. So perhaps an expert in the subject matter could learn the system in the letter correspondence table, and use their knowledge of medieval Greek herbals to figure out what many of the labels were intended to mean. I am not an expert in medieval Greek herbals, etc., so I may not be able to successfully accomplish this task myself.

I have already proposed potential identifications for a number of labels in one particularly significant set of labels, the 12 red labels on You are not allowed to view links. Register or Login to view. 2. The red label words are admittedly hard to read and hard to interpret. But I have made potential identifications of about 7 out of the 12 of them as Greek body part names, roughly corresponding to the body parts in the familiar medieval "Zodiac Man", in the proper order. The information is all there in the thread "Greek body parts, head to toe in order, in red labels on You are not allowed to view links. Register or Login to view. 2".

I also identify one famous label, the word next to the group of seven stars on You are not allowed to view links. Register or Login to view. 3, as [doaro] = Πουλ[ι]α, a modern Greek name for the Pleiades, keeping in mind that in terms of the history of the Greek language, "modern" often means "after 1100 CE", or "during or after Late Antiquity", or "any time after 5th century (BCE) Athens". I think that's a fairly clear label identification, given the limitations of the script. The iota is missing, but in this script that would require the horizontally extended character [ch], since the sequence [-re-] is very rare but [-rch-] is less rare. Perhaps the author felt that writing out [doarcho] would not be aesthetically ideal in this location on this page, and thus omitted one letter and simply wrote [doaro]. Who knows how many such decisions the author made in abbreviating words and labels? I doubt the missing iota would have caused a problem for a Judaeo-Greek reader in understanding what this label is supposed to mean.

Koen, I will definitely keep in mind your points and your interest in the capability of this theory and system to identify and read the labels in the Voynich MS. Thank you for your very valuable feedback. I appreciate it very much.

Geoffrey

Quote:Geoffrey: Voynich [d] is a critical letter. I understand JKP's concern about the status of this character as either a vowel or a consonant, but in fact this also arises naturally from Greek and Hebrew phonology....

It was actually less about that (which is why I used the example of the letter "y" in English, which can be both consonant and vowel) than it was about the degrees of freedom increasing rather substantially when transliterating the VMS characters.

If

1. the vowels can be substituted to a significant extent AND
   
2. you have a character that can function as both a consonant and a vowel AND
   
3. you add vowels to the end as needed...
   

you end up with a system in which there is so much latitude, that it's difficult to assess whether the tokens are being translated or created to suit the language at hand (whether it were Judaeo-Greek or another language wouldn't matter). I can get a lot of words and phrases out of VMS tokens using only number 1.

(30-03-2019, 04:00 AM)-JKP- Wrote: You are not allowed to view links. Register or Login to view.

Quote:Geoffrey: Voynich [d] is a critical letter. I understand JKP's concern about the status of this character as either a vowel or a consonant, but in fact this also arises naturally from Greek and Hebrew phonology....

It was actually less about that (which is why I used the example of the letter "y" in English, which can be both consonant and vowel) than it was about the degrees of freedom increasing rather substantially when transliterating the VMS characters.

If

1. the vowels can be substituted to a significant extent AND
   
2. you have a character that can function as both a consonant and a vowel AND
   
3. you add vowels to the end as needed...
   

you end up with a system in which there is so much latitude, that it's difficult to assess whether the tokens are being translated or created to suit the language at hand (whether it were Judaeo-Greek or another language wouldn't matter). I can get a lot of words and phrases out of VMS tokens using only number 1.

I understand the concern. It is a valid concern with any ambiguous writing system. But your 3 points would also be valid to challenge the legitimacy of the Judaeo-Greek script itself! All 3 criteria apply to Judaeo-Greek written in the Hebrew script, as well as to many other such abjad scripts. And yet people have been able to read and write and understand such scripts for many millennia - as long as the readers are native speakers of the underlying language. Of course for those of us who are not native speakers of Judaeo-Greek, it is much more difficult than it would have been for the intended audience of the Voynich MS.

Here is a way to challenge and test the degrees of freedom and latitude in my system: Someone who knows Greek can use my letter correspondence system and attempt to create their own alternate interpretations of the lines I interpret as I have presented them here. If someone else can produce their own equally valid, equally grammatical Greek sentences, with equally plausible semantic content of any nature, using the same system, but with substantially different Greek readings and interpretations, then that would indeed be a legitimate challenge to the validity of the system that I propose. However, if such alternate readings and interpretations are less compelling, if the Greek is less grammatical, if the semantic content is less plausible, or if they have to stretch the latitude and the degrees of freedom to a far greater extent than my readings and interpretations do - then that would suggest that my system does not create too much ambiguity or too much latitude or too many degrees of freedom after all.

Geoffrey

---

(29-03-2019, 11:08 PM)davidjackson Wrote: You are not allowed to view links. Register or Login to view.OK, let's see if I can get this straight.

Your theory is that Voynichese is a transliteration of Judaeo-Greek, which I understand to be known more formally as Yevanic.
Yevanic is a dialect of Greek written using the Hebrew alphabet.

So far, so good. Yevanic is mutually intelligible with Greek; and the conquered Moors of Spain did the same thing with Arabic (the literary tradition of the aljamiadas). No doubt there are some differences, it isn't straight Greek, but I understand that essentially a Greek and a Yevanic can chat perfectly happily together.

Two questions thus occur to me. Maybe I'm just not understanding the system properly.

• Why the additional translation into the Greek alphabet? Essentially, the word is the same in Yevanic and Greek, just that two different scripts are used, unless we are looking at a dialect argot or Hebrew loan word. Are you saying that the user was a Greek writer who transliterated his text twice? Otherwise, we seem to have added an arbitrary layer of transliteration that isn't needed. Voynichese -> Yevanic -> Greek --> English can logically be distilled into Voynich -> Yevanic --> English.
  If the underlying text is Greek, then we don't need Yevanic. If it is Yevanic, we don't need the Greek.
  
  
• The whole point of Yevanic was because the script was sacred to the users. They used the sacred Hebrew characters to write down their (new) mother tongue, ie Greek. Transliterating Hebrew into Voynichese defeats the purpose of this. What's more, I would assume that an orthodox Yevanic (Jew) wouldn't do this - they would go from Greek and keep the sacred script intact. Of course, this is subjective and we can't prove this. It is a consideration however - it seems more likely that the Greek variant word would have been used rather than the Yevanic. It also seems strange that they would not include the diacritics, and would invert the script direction (left to right as per Latin, rather than right to left as per Hebrew / Yevanic).
  

.

This is the second part of my reply to David.

I understand the point about the sacredness of the Hebrew script. I understand the corresponding point about the likelihood of using Greek variant words rather than Yevanic. Indeed, this is quite possibly true. So far, all I can say is that the systemic structure of the script seems to be based on Yevanic written in the Hebrew script (i.e., essentially an adapted abjad), and that I find a strikingly significant sentence containing Jewish cultural content, in Greek, in a particularly significant place in the MS, the very final paragraph. Beyond those two observations, I cannot yet say whether the underlying language of the text is more Greek or more Yevanic. I do observe that certain grammatical forms and phrases appear to be more ancient or classical Greek, such as the dative case, but written to reflect more medieval Greek pronunciation. 

Diacritics are mainly necessary for people who are not native speakers of the language. The diacritics were added to the Hebrew Bible in the standard Masoretic Text because by that time period (7th-10th c. CE), Hebrew was no longer a living spoken vernacular language, and thus diacritics that had not been necessary in earlier times had become necessary by then. Native Arabic speakers and readers do not need diacritics to read Arabic. In the parallel text Hebrew and Judaeo-Spanish book from which I posted a page earlier in this thread, it is striking that the Hebrew side has vowel diacritic dots but the Judaeo-Spanish side does not. It was written for native Judaeo-Spanish speakers who didn't need the diacritics for their native spoken language, but did need them for Hebrew.

There are a significant number of elements that appear to be signs of a deliberate intention to make the script hard to read for the uninitiated, even those who knew Greek and Yevanic. Ambiguating the consonants, then creating alternate characters for each type of those ambiguated consonants, strikes me as a clear sign of an intention to make the script effectively a cipher that only those who knew the encryption and decryption system would be able to figure out, even among those who knew Greek and Yevanic, which knowledge would of course also be necessary to read and understand the script. The reversal of the script direction could have been another step to disguise the Yevanic / Hebrew / Jewish nature and origin of the script and text.

I can think of logical reasons for the author to feel the need to create such a cipher.

The Voynich MS dates to the period of the Ottoman conquest of the Byzantine Empire. It is true that Constantinople itself did not fall until 1453, but many parts of the Byzantine Empire were conquered by the Ottomans well before that date, by the late 14th and early 15th centuries, including substantial areas of Greece itself. Perhaps the author of the Voynich MS felt the need to hide the contents of it during this period, and even to prevent outsiders from being able to identify the script or language or cultural origin of the MS at all. That would be a justification for many of the features and elements of the encryption system that I have described here, including the decision not to write in their sacred script, which could be identified as such even by others who didn't know or read the language. Of course there are many possible reasons for an author to want to conceal the contents of a text, not just from Ottoman conquerors in one particular time and place.

An alternative explanation for the different, non-sacred, script might be found in the Karaite Jewish community. Karaites are Jewish but they do not accept the Talmud as an authoritative interpretation of the Torah and the Tanakh, which distinguishes them from Rabbinic Judaism. There was a substantial community of Constantinopolitan Karaites in the Byzantine Empire in the medieval period; a small community still exists today. They also spoke Yevanic or Judaeo-Greek, but in the case of their community the language was called Karaeo-Greek or Karaitika. The Karaites were significant enough among medieval Greek Jews in particular, that some well-known leading Romaniote Greek Jewish scholars (Romaniotes were and are Rabbinic) wrote about the need to reunite the two Greek Jewish communities. The Constanopolitan Karaites were also known for producing leading scholars and writers.

I bring all this up, because although the Karaites also wrote Judaeo-Greek in the Hebrew script, they have been known historically to have less attachment to the sacredness of the Hebrew script in and of itself. For example, consider the information in the following paper:

You are not allowed to view links. Register or Login to view.

This paper notes for example that in the 10th and 11th centuries, Karaites in Palestine and Egypt even wrote manuscripts of the Hebrew Bible in Arabic transcription. A letter described in the paper shows that Karaite scribes made copies of Arabic language ms both in Judaeo-Arabic (that is, in the Hebrew script) and in the Arabic script. The paper cites a 10th century Karaite scholar who defended the writing of the Hebrew language in the Arabic script, making the argument that (quoting the paper) "written letters are merely the symbols of the language. The language would be the same whatever script it is written in." But he acknowledged disagreement on this matter among Jews and perhaps among Karaites themselves.

Of course none of this means that Karaites necessarily have a connection to the Voynich MS, but it does show that one significant and prominent medieval Judaeo-Greek community belonged to a tradition that perhaps had less religious attachment to the sacredness of the Hebrew script itself, even though they used it to read and write both Hebrew and Judaeo-Greek. Perhaps Karaites would have had fewer qualms about writing a text such as the Voynich MS in such a strange and different script. Again, it is just an idea, not an essential part of my theory.

Geoffrey

Quote:In the parallel text Hebrew and Judaeo-Spanish book from which I posted a page earlier in this thread, it is striking that the Hebrew side has vowel diacritic dots but the Judaeo-Spanish side does not.

Not exactly. Judaeo-spanish (more formally known as ladino) is written in a variant Hebrew script known as Rashí. This script is phonological in style, with common letters for common vocal sounds, but the vocals can be repeated (two symbols for a, u, o and others depending upon their word position). 
Diacritics aren't used because they aren't needed. It's not an abjad. Spanish would be impossible to write in an abjad.

Quote:Geoffrey: Here is a way to challenge and test the degrees of freedom and latitude in my system: Someone who knows Greek can use my letter correspondence system and attempt to create their own alternate interpretations of the lines I interpret as I have presented them here. If someone else can produce their own equally valid, equally grammatical Greek sentences, with equally plausible semantic content of any nature, using the same system, ...

It doesn't have to be Greek. If this system works with another language in essentially the same way (alternate vowels, one char both consonant or vowel with a rational linguistic reason, added vowels at the end as needed (or alternated endings as needed for the language)), it calls into question the system itself.

[quote="davidjackson" pid='25932' dateline='1553926697']

Quote:Spanish would be impossible to write in an abjad.

This seems like a crucial point. Indeed, abjad scripts are particularly suited for semitic languages, and writing Spanish as such would result in a highly flawed system.

But is it really different for Greek? I don't see why a partial abjad would work for Greek but not for Spanish.

(30-03-2019, 06:07 AM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view.

(30-03-2019, 04:00 AM)-JKP- Wrote: You are not allowed to view links. Register or Login to view.

Quote:Geoffrey: Voynich [d] is a critical letter. I understand JKP's concern about the status of this character as either a vowel or a consonant, but in fact this also arises naturally from Greek and Hebrew phonology....

It was actually less about that (which is why I used the example of the letter "y" in English, which can be both consonant and vowel) than it was about the degrees of freedom increasing rather substantially when transliterating the VMS characters.

If

1. the vowels can be substituted to a significant extent AND
   
2. you have a character that can function as both a consonant and a vowel AND
   
3. you add vowels to the end as needed...
   

you end up with a system in which there is so much latitude, that it's difficult to assess whether the tokens are being translated or created to suit the language at hand (whether it were Judaeo-Greek or another language wouldn't matter). I can get a lot of words and phrases out of VMS tokens using only number 1.

I understand the concern. It is a valid concern with any ambiguous writing system. But your 3 points would also be valid to challenge the legitimacy of the Judaeo-Greek script itself! All 3 criteria apply to Judaeo-Greek written in the Hebrew script,

The fact that your claims about Judaeo-Greek come without examples from actual Judaeo-Greek sources (carefully transcribed, grammatically analyzed and translated) suggests  that you have never read a Judaeo-Greek text. I find you claims unreliable.

Quote: But is it really different for Greek? I don't see why a partial abjad would work for Greek but not for Spanish.

Greek, of course, famously being the first ancient script to force the inclusion of vowels because the ambiguity of an abjad made it unworkable for their language.

That's not to say that the Voynich isn't an abjad. An often understated point when discussing this is that to have a pure abjad, you need very strict lexemes and word structure. Arabic, for example, mainly being constructed of three consonant roots which can then be expanded as needed.