The Voynich Ninja

Full Version: geoffreycaveney's Judaeo-Greek theory
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Quote: Geoffrey: Now some people might argue, "Why should Geoffrey be trying to decipher and read and interpret the Voynich MS at all? If he thinks it is written in Greek in the late medieval period in a script based on or inspired by the Hebrew script of Yevanic or Judaeo-Greek, why not leave it to the scholars and specialists and experts of those languages and dialects and scripts to study and research and investigate?"

I don't think anything positive comes out of arguing that. You are entitled to study anything you want. If you don't have the skillset, you can learn it. It might take time, but it can be learned. Some people have been working on the VMS for over 20 years. A lot of skills can be acquired in 20 years.

Also, many important discoveries have been made by amateurs, or by professionals in fields outside their chosen profession. It's the integrity and rationality of the method and the quality of the outcome that counts.
Geoffrey, thank you for your interesting long post. I think we understand each other and I look forwards to reading more of your observations in the future. 
As with all research, the important thing is to observe and develop. One cannot simply present a few words and base a house of cards on it to delude oneself into thinking one has the solution. There are many people with specialised areas of interest on this forum who are happy to chip in and contribute to any meaningful discussion, and your theory is certainly worth further investigation. 
In your notebook on the first page there appear to be labels with niqqud; possibly this indicates the use of both vowel indication systems. Niqqud for certain short ambiguous text.
If you live in NYC, possibly there might be a local rabbi who could help with transcribing the text? It would be fascinating to see what a trained Hebrew speaker makes of it.
(03-04-2019, 06:20 AM)davidjackson Wrote: You are not allowed to view links. Register or Login to view....

If you live in NYC, possibly there might be a local rabbi who could help with transcribing the text? It would be fascinating to see what a trained Hebrew speaker makes of it.

Investigating some other issue ( Ha-Melacha ha-Ketzara ),the Jewish Theological Seminary in NY helped me a lot. Maybe that would be a possibility.

The Jewish Theological Seminary
3080 Broadway
New York, NY 10027
(212) 678-8000
Hi Geoffrey,
if I understand correctly, you are saying that it is not necessary to by an academic scholar to study the Voynich manuscript. I agree, of course: this is an internet forum where amateurs discuss their ideas. Between the two extremes of delegating all research to academical scholars and totally ignoring the work of others there is the middle way of studying what scholars and other amateurs have done and try to build upon it.

This is not the first time you compare your research with Michael Ventris' work, but there are major differences.

A first difference (that I previously pointed out) is Ventris' careful quantitative analysis of the unknown language he was studying. This is something I don't see in your research: not only you do not add to the objective observations that others have produced about Voynichese, but your theory does not try to explain any of the available evidence. For instance: the low entropy, the fact that EVA:p and f mostly occur in the first lines of paragraphs, Grove words, Currier A and B, the frequent reduplication and quasi-reduplication of words...

A second difference is that Ventris built on the work of others, while you ignore both scholarly Yavanic research and everything others have pointed out about the VMS. 
In 1940 (You are not allowed to view links. Register or Login to view., published when he was 18) Ventris was already familiar with the work of several academic researchers, such as Robert S. Conway, Arthur E. Cowley and (obviously) Arthur Evans. He wrote:

Michael Ventris Wrote:The enormous obstacles that stand in the way of decipherment, and the general lack of data, have led, on the part of more serious students, to considerable prudence and reserve: but equally, to a great variety of bizarre and ill-founded conjectures from the more amateur.
[...]
I do not propose to offer here any broad "translations" of the inscriptions. All I want to do, within this short space, is briefly review the evidence and see what lines of approach it suggests.
[...]
The real and valuable work on the Minoan inscriptions has so far consisted almost entirely of research into the method rather than the meaning. A great deal of labor has gone into classifying the signs and tracing their varying forms throughout the successive systems, in sorting out the records according to certain recurrent ideographs, and in the detection of that large element consisting of proper names. All this we owe to the untiring energy of Sir Arthur Evans. No further progress would be possible without it.

For more than ten years, Ventris built on the evidence collected and analyzed by Evans and later Kober and Bennett. Since 1941, he corresponded with Sir John Myres (who was publishing the Linear B tablets of Knossos). You write that "at a certain stage ... Ventris had to obtain the critical assistance and close cooperation of John Chadwick, a professional scholar and specialist in archaic Greek." You are not allowed to view links. Register or Login to view. says that:

Quote:Ventris had been invited to give a talk on the BBC Third Programme about Myres’s publication of the Knossos tablets. He took the opportunity to announce his decipherment and it was broadcast to the world on 1 July 1952. The talk was heard by John Chadwick, a newly-appointed lecturer in the Faculty of Classics at Cambridge, who, after getting a copy of Ventris’ material from Myres, was the first to write to him with congratulations, offering his help “if there is anything a mere philologist can do”.

As amateurs, we can still do our best to work as "serious students": creating "bizarre and ill-founded conjectures" is not the only option. If we are serious and persistent enough, we might one day have a chance to collaborate with academic researchers. There actually are Voynichese amateurs that have been as successful as that: it is not an impossible goal.
(03-04-2019, 06:20 AM)davidjackson Wrote: You are not allowed to view links. Register or Login to view.Geoffrey, thank you for your interesting long post. I think we understand each other and I look forwards to reading more of your observations in the future. 
As with all research, the important thing is to observe and develop. One cannot simply present a few words and base a house of cards on it to delude oneself into thinking one has the solution. There are many people with specialised areas of interest on this forum who are happy to chip in and contribute to any meaningful discussion, and your theory is certainly worth further investigation. 
In your notebook on the first page there appear to be labels with niqqud; possibly this indicates the use of both vowel indication systems. Niqqud for certain short ambiguous text.
If you live in NYC, possibly there might be a local rabbi who could help with transcribing the text? It would be fascinating to see what a trained Hebrew speaker makes of it.

Thank you David, I agree and I also look forward to having continued productive discussions with you and with many other contributors on this forum.

Geoffrey
(03-04-2019, 11:13 AM)MarcoP Wrote: You are not allowed to view links. Register or Login to view.Hi Geoffrey,
if I understand correctly, you are saying that it is not necessary to by an academic scholar to study the Voynich manuscript. I agree, of course: this is an internet forum where amateurs discuss their ideas. Between the two extremes of delegating all research to academical scholars and totally ignoring the work of others there is the middle way of studying what scholars and other amateurs have done and try to build upon it.

This is not the first time you compare your research with Michael Ventris' work, but there are major differences.

A first difference (that I previously pointed out) is Ventris' careful quantitative analysis of the unknown language he was studying. This is something I don't see in your research: not only you do not add to the objective observations that others have produced about Voynichese, but your theory does not try to explain any of the available evidence. For instance: the low entropy, the fact that EVA:p and f mostly occur in the first lines of paragraphs, Grove words, Currier A and B, the frequent reduplication and quasi-reduplication of words...

A second difference is that Ventris built on the work of others, while you ignore both scholarly Yavanic research and everything others have pointed out about the VMS. 
In 1940 (You are not allowed to view links. Register or Login to view., published when he was 18) Ventris was already familiar with the work of several academic researchers, such as Robert S. Conway, Arthur E. Cowley and (obviously) Arthur Evans. He wrote:

Michael Ventris Wrote:The enormous obstacles that stand in the way of decipherment, and the general lack of data, have led, on the part of more serious students, to considerable prudence and reserve: but equally, to a great variety of bizarre and ill-founded conjectures from the more amateur.
[...]
I do not propose to offer here any broad "translations" of the inscriptions. All I want to do, within this short space, is briefly review the evidence and see what lines of approach it suggests.
[...]
The real and valuable work on the Minoan inscriptions has so far consisted almost entirely of research into the method rather than the meaning. A great deal of labor has gone into classifying the signs and tracing their varying forms throughout the successive systems, in sorting out the records according to certain recurrent ideographs, and in the detection of that large element consisting of proper names. All this we owe to the untiring energy of Sir Arthur Evans. No further progress would be possible without it.

For more than ten years, Ventris built on the evidence collected and analyzed by Evans and later Kober and Bennett. Since 1941, he corresponded with Sir John Myres (who was publishing the Linear B tablets of Knossos). You write that "at a certain stage ... Ventris had to obtain the critical assistance and close cooperation of John Chadwick, a professional scholar and specialist in archaic Greek." You are not allowed to view links. Register or Login to view. says that:

Quote:Ventris had been invited to give a talk on the BBC Third Programme about Myres’s publication of the Knossos tablets. He took the opportunity to announce his decipherment and it was broadcast to the world on 1 July 1952. The talk was heard by John Chadwick, a newly-appointed lecturer in the Faculty of Classics at Cambridge, who, after getting a copy of Ventris’ material from Myres, was the first to write to him with congratulations, offering his help “if there is anything a mere philologist can do”.

As amateurs, we can still do our best to work as "serious students": creating "bizarre and ill-founded conjectures" is not the only option. If we are serious and persistent enough, we might one day have a chance to collaborate with academic researchers. There actually are Voynichese amateurs that have been as successful as that: it is not an impossible goal.

Marco, thank you for the comments about and quotes by Ventris. Everything Ventris said and did is absolutely useful for us in our analysis and research of the Voynich MS.

I do believe that I am building upon the work of the many other Voynich researchers who have performed detailed statistical analysis on the MS text for many decades. I have never tried to ignore it; I have read it and studied it for years. I read and studied just about every article and every piece of statistical analysis and information on Rene's voynich.nu website, and many other sources of information, beginning several years ago. It's not like I just began learning about the Voynich MS when I joined this forum a month ago.

Before I started posting anything about my quasi-Judaeo-Greek theory or any specific readings and interpretations of passages of text, I posted a detailed statistical analysis of the phoneme distribution of modern Greek in the following post toward the end (3rd page) of my "Bigram = phoneme theory (language agnostic)" thread:

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

To summarize the main conclusion of this detailed statistical analysis, "[Modern Greek] uses a small number of consonants very frequently, much more so than other European languages." The point was precisely to research a plausible statistical connection between consonant phoneme frequency and distribution, and the low entropy of the Voynich MS text.

For the record, I researched the "Bigram = phoneme" hypothesis in detail, but I rejected it in the end when it did not fit the additional data I collected. With bigrams I had hypothesized a possible connection to Hebrew at one stage. I analyzed frequent Voynich words, prefixes, and suffixes, and compared them with common Hebrew words, prefixes, and suffixes. But I rejected this hypothesis when I requested and received feedback from a Spanish scholar of medieval Hebrew, who explained that certain substitutions and alternations of guttural letters (aleph, ayin, he, chet, etc.) were not plausible in these most common words, because in the medieval period anyone writing Hebrew at all had to be familiar with standard literary Hebrew. The substitutions and alternations, he explained, would be more typical of a native speaker writing his language more or less phonetically rather than in accordance with a literary standard, but this type of speaker of Hebrew did not exist in the medieval period. So I rejected this hypothesis.

You will recall that very shortly before I developed my quasi-Judaeo-Greek theory, I was also researching such languages as Old Occitan, including certain "cryptographic" and wordplay poetry from the late medieval period, precisely in order to research a possible connection with and explanation for the low entropy, unusual bigram distribution patterns, and other statistical analysis of the Voynich MS text. I can hardly be accused of ignoring or disregarding or downplaying the significance of such analysis.

In fact, I was researching the Voynich MS text to test the Old Occitan / Middle French / late medieval Romance language hypothesis at the time that I stumbled upon my Greek theory in its current form. I had already observed the Greek body part words in the red labels of the Zodiac circle on You are not allowed to view links. Register or Login to view. 2, but I did not post anything about this at the time since I considered that it was not yet compelling in the absence of any other evidence from the text. (I have since posted about it.) At the time I was considering a polyglot or multiple language theory of the MS, in which certain significant text such as these red labels might be in Greek, but other parts of the text might be in a Romance language. The main essential features of my letter correspondence table, by the way, were first hypothesized by me on a language-agnostic basis, based on general linguistic principles of phonological inventories and on the low entropy and bigram distribution statistical analysis of the Voynich MS text. I first considered the idea of Voynich [d] as alternating between /u/~/v/ and /b/~/p/ not while testing my Greek theory, but while testing Occitan and related Romance languages.

I was in fact attempting to read the first line of You are not allowed to view links. Register or Login to view. 1 as Occitan or French, as part of my hypothesis testing, when the word [epairody] arose as "ipeirous", forcing my attention back to Greek once again. It was only after reading and interpreting this word and this line, in addition to the red label Greek body parts that I had not posted about yet, that I first posted about my theory on this forum, in the very first post of this long (and productive) discussion thread.

The whole concept of the ambiguation of related groups of phonemes in my letter correspondence table is precisely designed to explain all of the statistical analytical properties of the Voynich MS text that Rene and so many other researchers have studied for decades. Again and again, the statistical analysis of researchers finds that the Voynich text most closely resembles Hawaiian. My letter correspondence table takes Greek and quasi-Judaeo-Greek, and transforms it via the encryption process that I have described in detail, into a final form that possesses many of the statistical properties of Hawaiian.

I further explored these concepts in another thread, "exercise: 9-phoneme Greek text generation":

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

The point is that Greek may be recognizable and readable, albeit with difficulty, even when its phoneme and letter inventory has been condensed into 9 phonemes and 11 letters. This is very much akin to the process that I propose was employed to transform Greek or quasi-Judaeo-Greek into the Voynich script. It gives Greek many of the statistical properties of Hawaiian.

Marco, you performed a similar exercise with English, an effort which I appreciate very much. However, I do note that I believe Greek is a much more natural candidate language to remain recognizable and readable (albeit with difficulty) after this process than English is. Modern Greek has 15 consonant phonemes with distinct letters (/b/, /d/, /g/ are rare and must be represented with the unusual letter combinations "mp", "nt", "gk") and 5 vowel phonemes. Modern American English has 24 consonant phonemes and 15 vowel phonemes. Other dialects of English have even more vowel phonemes. It is reasonable and plausible to condense the Greek phoneme inventory into a Hawaiian-style system and still hope to be able to work out what it means. It is much less reasonable and plausible to do so with English.

About EVA [p] and [f] mostly occurring in the first lines of paragraphs, I believe Rene has acknowledged that my theory is one of the few that even attempts to account for this phenomenon, although he still does not accept my theory. In fact, the original motivation for me to hypothesize and discover the unusual alternation of [d] as both /u/~/v/ and /b/~/p/ arose precisely from the attempt to explain the problem of EVA [p] and [f]. In short, [p] and [f] are the "pure" forms of the labial consonant series /p/, /ph/, /b/~/v/, but the author only bothered to write these ornate characters, sometimes, in the first lines of paragraphs, and simply used EVA [d] as a substitute almost always everywhere else, and sometimes in the first lines too. The logical reason for the substitution is that EVA [d], the core value of which is /u/ or Hebrew vav, can alternately represent /v/ as Hebrew vav does; this then led to the extension to represent Greek beta or Hebrew bet, which are also typically pronounced /v/ (since the phoneme /b/ itself is rare in Greek since the sound change of beta et al., which had already happened in Late Antiquity). Since the entire system I propose merges voiced and voiceless consonants, the extension of EVA [d] to represent Greek beta / Hebrew bet also entailed its extension to represent Greek pi, phi / Hebrew pe as well.

Does this count as "my theory tries to explain the available evidence about the fact that EVA [p] and [f] mostly occur in the first lines of paragraphs"?

I will be more than happy to discuss and consider all of the other phenomena you cite as examples of more such available evidence about the quantitative analysis of the Voynich MS text. I do not claim to have explained every single statistical phenomenon observed in the text as of yet. But I believe it is fair to say, based on the evidence that I have provided in this post, that I have made a substantial effort to study, consider, and account for the quantitative statistical analysis of the Voynich MS text, in the course of the development of my theory.

Geoffrey
geoffreycaveney Wrote:The whole concept of the ambiguation of related groups of phonemes in my letter correspondence table is precisely designed to explain all of the statistical analytical properties of the Voynich MS text that Rene and so many other researchers have studied for decades.

Stating that you are explaining all of the statistical properties of the manuscript is quite bold. Sadly, this generic statement is not accompanied by specific and verifiable measures: quite a contradiction. 
I think the evidence we have suggests that You are not allowed to view links. Register or Login to view. has been designed to give you enough freedom to project Voynichese into Greek words. I see few traces of any awareness of the properties of Voynichese.

I don't claim to cover "all of the statistical analytical properties of the Voynich MS text" but I will only list some broad categories. Your table does not say anything useful about any of them.

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

Your table actually seems to cause a slight increase in conditional entropy: character sequences are more unpredictable. In Greek (as in basically all languages) vowels and consonants tend to alternate. Your process removes some vowels and has characters that can function as both vowels and consonants: the result is more unpredictability i.e. higher conditional entropy. Even if your table slightly reduces the size of the alphabet, the ambiguation process results in only a small decrease in character entropy. In other words, the transformed text has entropy values close to those of Greek, with a conditional entropy that is marginally higher than the original Greek and much higher than Voynichese.

For instance, the Iliad fragment you considered You are not allowed to view links. Register or Login to view. has
character entropy: 4.10
conditional entropy: 2.45

A Voynich fragment of the same length:
character entropy: 3.77
conditional entropy: 1.68 (much lower than Greek)

If you transform the Iliad fragment according to your table you get something like this (one of an almost infinite number of possible outcomes):

Sisir eor kC li dCye koa kZey rtrdy
ZitsZi qiray prS sSkoeo kiray k
i liS q okC SaSN Zrol paril aldelatolae
roZ SeS lai SayZay Zr triy pkCkd yZae
e s tS eqs CtSC dCrCl y pasrCka iaCaN
dreZo Soi qry oyZro di kardN Z Sde edS
ayas eka ta l dto siry kSsaCa qitoiC

(where C stands for "ch", N for 'iin', S and Z for open/closed "sh")

Entropy values are now:
character entropy: 3.95
conditional entropy: 2.54 (just higher than Greek, much higher than the VMS)

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

The positional constraints of Voynichese are not only unexplained, they are not even modelled by your table. E.g. the fact that q- typically occurs at the beginning of words. Voynichese word structure is ignored, but for -iin having to occur word finally and -y corresponding to -ς.

Of course, other less obvious features are also totally ignored. E.g. more than 80% of occurrences of q- appear in the form qo-Gallows: why?

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

This whole phenomenon is not explained (nor modelled) in its entirety.

You are not allowed to view links. Register or Login to view. has shown evidence that refutes the only side of this issue you have discussed (d as a substitute for p/f). BTW, it's interesting that Nablator kindly checked things for you and you didn't even thank him and just ignore his work: so much for "building upon the work of other Voynich researchers." By now I know that my observations here will be hand-waved away with some text-wall devoid of all objective, quantitative data. I should really refrain from wasting my time on this...

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

A good and simple test for a theory (proposed by You are not allowed to view links. Register or Login to view. and recently mentioned by Farmerjohn) is to check that the most frequent Voynichese words match the most frequent target language (i.e. Greek) words. The You are not allowed to view links. Register or Login to view. you have provided suggests that this is not the case. Actually, you interpret the most frequent words as suffixes / prefixes. Since your system does not (yet) allow  to glue together words that appear as distinct in the manuscript, this also seems like a total failure.


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

What?

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

Uh?


Quote:I further explored these concepts in another thread, "exercise: 9-phoneme Greek text generation":

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

The point is that Greek may be recognizable and readable, albeit with difficulty, even when its phoneme and letter inventory has been condensed into 9 phonemes and 11 letters. This is very much akin to the process that I propose was employed to transform Greek or quasi-Judaeo-Greek into the Voynich script.

No, that is not "very much akin" to the process you propose. That is a small sub-step in your process. Each step in the process throws away some of the information in the original Greek:
  • compression of the alphabet (each Greek character mapped into a distinct character)
  • ambiguation of the alphabet (distinct target characters are replaced by overlapping character sets, with vowels and consonants sometimes mapped into the same character)
  • joining function words with other words
  • dropping of vowels
  • dropping of suffixes
  • dropping of consecutive identical characters

What you say about the first step also applies to the others: each single step makes the text more difficult to read. The combination of all steps makes the text hopelessly ambiguous.
Marco, I appreciate all of your observations. I see them not as refutations of my theory, but as good questions that need to be further investigated and explained. I do not deny that much further investigation and explanation is certainly necessary. If I ever suggested or claimed otherwise, then that was a mistake and a misstatement, and I apologize if I gave anyone that impression.

My previous statement about "all of the statistical analytical properties" was clearly a misstatement on my part. The word "all" was of course inaccurate. I apologize for this misstatement. I did address a number of such statistical properties in my previous post, so I should have written a phrase along those lines in my summary description.

I very much appreciate nablator's research on the frequency of Eva [d] in lines without [p] or [f] vs. lines with [p] or [f]. I certainly did not mean to ignore it. I must have temporarily set the issue aside in the course of other research, and I inadvertently failed to get back to it to follow up. I apologize for this oversight as well.

Again, I view all of these questions Marco raises as important issues that need to be further investigated and explained. I believe that further investigation may well lead to further insights that are capable of explaining these issues, within the broad context of my hypothesis as I have presented it. I get the feeling that some researchers here may believe that the issues Marco raises mean that my theory should be rejected or abandoned. I on the other hand view the post as a very useful summary of difficult issues, which I can use as a basis and a guide for further research and investigation to identify possible explanations for these issues. Such research and investigation may take months, not a day or two.

I cannot reply to all of Marco's observations right now in this comment. But I can give a few brief comments and observations, while recognizing that much further research is clearly necessary.

1) First, I want to briefly address nablator's research on the frequency of [d] in lines with and without [p] or [f]. This is not intended as a full or complete answer or explanation, just as a brief initial comment on the issue. nablator showed that [d] is not more frequent in lines without [p] or [f] than it is in lines with [p] or [f]. I understand that this analysis prima facie fails to confirm my hypothesis of [d] as a substitute for [p] and [f]. I will have to consider other or additional explanations for what is going on, in order to explain and make sense of this statistical data. For example, in general in the MS text, I have read and also observed myself that similar word forms tend to occur in more or less close proximity to each other; that is, one often finds variants of a word form with a letter or two altered or with other small changes, within the surrounding lines above or below the word form. I would describe this as a form of "character alliteration" in the Voynich MS text. We may not know for certain which actual letters or sounds each character represents, but whatever letters or sounds they are, they tend to occur in a repeated manner within a paragraph or passage or block of lines of text. I suggest that this may represent a frequent usage of alliteration and assonance on the part of the author in the underlying language of the text. If this is the case, then it would mean that lines with [p] or [f] may contain alliteration or assonance of the sounds that these characters represent. My theory proposes that these sounds are Greek /p/, /f/ ("ph"), /v/ ("b"); others may have other ideas. In this case, the use of alliteration or assonance would mean that the combined frequency of [p], [f], and [d] as a substitute for the same letters and sounds, would naturally be greater in lines containing this alliteration / assonance of labial sounds, and the combined frequency would be smaller in lines without such alliteration or assonance of labials. This would explain why [d], even as a substitute for [p] and [f], is still no less frequent in lines with [p] or [f] than it is in lines without [p] or [f]. 

Now the follow-up question would be, how then does one explain the particular greater combined frequency of [p], [f], and [d] in the first lines of paragraphs, as opposed to their smaller combined frequency in other lines of the text? All I can suggest for now is the following idea: Perhaps the author chose to use alliteration or assonance of labial sounds more often in the first lines of paragraphs in particular. One may then ask, is there any particular plausible reason one can think of to explain why the author would do this? I can suggest this idea: The elaborate and ornate characters [p] and [f] simply look better aesthetically in first lines, with more room to draw their tall shapes above the line, so the author preferred to employ alliteration and assonance of their type of sounds in those lines. I do acknowledge and accept Emma's point in the other thread that the author was capable of adjusting the size and height of glyphs, and so could have used these characters and these sounds just as well and just as often in other lines if he or she had wished to do so. Still, for aesthetic reasons the author may have decided that these particular characters [p] and [f] look best when there is more room to draw them to their full height, and thus he or she was inspired to employ alliteration and assonance of their labial sounds more often in these first lines.

Still, just because such labial sounds would thus be less frequent in other lines, does not mean of course that the author would be able to avoid these sounds entirely elsewhere. For that purpose of representing these sounds in other lines, I propose that the author there usually chose to use [d] as a substitute for [p] and [f].

2) Concerning other issues of word structure and the line as a functional unit, I believe that my table at least provides the basis for a possible plausible explanation for many of these issues. Not only can [d] be used as a substitute for [p] and [f], but also likewise [s] is essentially a substitute for [k], and [q] is a substitute for [t]. (Also, although I have not added it to the table yet, I suspect that [g] and [m] are essentially most often substitutes for [y].) The possibility of all of these substitutions of alternate characters, without affecting the underlying language of the text at all, may possibly explain the existence of some of the more unusual features of Voynich word structure and line structure. Basically, having all of these optional alternate characters at his or her disposal as substitutions wherever and whenever desired, the author would have been free to create many unusual character distribution patterns, and again it would not have affected or changed the underlying language of the text at all.

Consider for example the virtual impossibility of double gallows (non-occurrence of any two characters of the set [k, t, p, f, ckh, cth, cph, cfh] next to each other, with only 6 isolated exceptions in the entire MS text), which Emma raised and discussed in a long-running thread on this forum. My theory and table offer a simple explanation: the author was easily able to avoid double gallows by always using an alternate character as a substitute for one of them instead. Now in Greek, many of these combinations will not often naturally occur anyway, but one such type of combination will occur indeed: "-pt-" and "-kt-", as in "hepta" and "okto". The author did not have to use double gallows for such combinations, because he or she could use [ps] and [ts] instead (usually in the form of [psh] and [tsh]). The combination [ks] (usually as [ksh]) for "-tt-" is also possible. Alternately, the author could write [dk] to represent "-pt-" or "-bd-" for example, which in fact I propose occurs in the word [chedkaly] that I identified as "hebdom[e]s" ("sevenths" or I suggested "sabbaths").

On the other hand, in Greek such combinations with "k" rather than "t" as the second letter in the cluster will be much, much rarer. This is why we don't have words such as "hepcagon" or "otcagon" in our languages today. And here we find a possible explanation for the positional restriction of [q] in the Voynich MS text, which issue Marco raises in his post. Unlike [s]/[sh], the author didn't need to use the alternate character [q] in this cluster-final position, because unlike Greek "t", Greek "k" rarely occurs there.

As for the almost exclusive restriction of [q] to word-initial position, and the high frequency of [qok-] (by far the most frequent [q]-sequence) and [qot-] (by far the second most frequent one), I can also offer plausible explanations in line with my theory. It would be plausible that the author preferred to use the most prominent gallows character for the first consonant of the root of a word. So for example when the ubiquitous simple Greek conjunction "kai" ("and") is prefixed to a following word, the author chose to use [qo-], representing "ka'-", rather than [to-], so that the first consonant of the following root word would remain the most visually prominent character in the word, as the first and usually the only gallows character in the word. This way the little "ka'-" conjunction prefix would not "steal the thunder" of the much more important and significant first consonant of the main root word: the pride of place as the prominent first gallows character of the word was reserved for that first root consonant, and the smaller, simpler, less prominent [q] was used for the "k" in "ka'-" instead.

However, when the first consonant of the root word was not a gallows character anyway, this aesthetic issue was not relevant, so there the author did not necessarily need to use [q] as a less prominent alternate character for [t] in the prefixed conjunction "ka'-" ("kai"). Thus, the primary purpose and usage of [q] throughout the MS text is mainly to represent the "k" in "ka'-" before "gallows-initial" root words, in order to avoid the use of the "competing" prominent gallows character [t] as "k" in such positions. This explains the frequency of the occurrence of [qo-Gallows] that Marco raises the question about.

3) About word frequency, and the correspondence of the most frequent Voynichese words with the most frequent Greek words, of course I am highly focused on this issue. This is and always has been one of my main concerns when I even begin to investigate any possible hypothesis about the underlying language of the Voynich MS text. I do not recall that I have ever stated or implied that my theory does not allow for the representation of some prefixes and especially suffixes as separate words that appear as distinct in the manuscript. For example, let us go back and consider the final line of the four-line poem at the top of You are not allowed to view links. Register or Login to view. 1, which lines were in fact the very first passage of text that I posted on this forum and in this thread with a quasi-Judaeo-Greek reading and interpretation. If we go back and review my reading and interpretation of that final line, we will see that I read and interpreted the two separate, distinct Voynichese words "[soar cheey]" as the one single Greek word δολιας ("dolias"), the genitive singular feminine form of the adjective δολιος, meaning "deceitful" or "treacherous". To review all the steps, I read [soar cheey] as "doAl--iies", which I interpreted as "dol-ies", which I consider a plausible possible pronunciation of δολιας. (In fact, if you listen carefully to a modern Greek pronunciation of this word, you will hear that the "a" is not pronounced as a European "a" sound, but rather much closer to an American English "short a" sound as in "cat". This latter sound is in fact phonetically closer to short "e" than it is to a pure "a".) Thus, it is fully in line with my theory, and with my readings and interpretations in practice to date, to interpret some of the most frequent Voynichese words as Greek suffixes / prefixes.

Entropy is of course also a vital issue to consider, discuss, and address. It is such a big issue that I will have to return to it and treat it separately in another post. There I will also include a discussion of the relevance of the "9-phoneme (11-letter) Greek text generation".

In the meantime, I am also going to continue to respond to requests such as Koen's, who asked about how my system handles labels such as those 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. , and You are not allowed to view links. Register or Login to view. (mainly labels of roots and leaves of plants). I have been working on such readings and interpretations in response to Koen's request, and I hope to be able to post something about this topic relatively soon.

Geoffrey
(04-04-2019, 10:01 AM)MarcoP Wrote: You are not allowed to view links. Register or Login to view.* You are not allowed to view links. Register or Login to view.

Your table actually seems to cause a slight increase in conditional entropy: character sequences are more unpredictable. In Greek (as in basically all languages) vowels and consonants tend to alternate. Your process removes some vowels and has characters that can function as both vowels and consonants: the result is more unpredictability i.e. higher conditional entropy. Even if your table slightly reduces the size of the alphabet, the ambiguation process results in only a small decrease in character entropy. In other words, the transformed text has entropy values close to those of Greek, with a conditional entropy that is marginally higher than the original Greek and much higher than Voynichese.

For instance, the Iliad fragment you considered You are not allowed to view links. Register or Login to view. has
character entropy: 4.10
conditional entropy: 2.45

A Voynich fragment of the same length:
character entropy: 3.77
conditional entropy: 1.68 (much lower than Greek)

If you transform the Iliad fragment according to your table you get something like this (one of an almost infinite number of possible outcomes):

Sisir eor kC li dCye koa kZey rtrdy
ZitsZi qiray prS sSkoeo kiray k
i liS q okC SaSN Zrol paril aldelatolae
roZ SeS lai SayZay Zr triy pkCkd yZae
e s tS eqs CtSC dCrCl y pasrCka iaCaN
dreZo Soi qry oyZro di kardN Z Sde edS
ayas eka ta l dto siry kSsaCa qitoiC

(where C stands for "ch", N for 'iin', S and Z for open/closed "sh")

Entropy values are now:
character entropy: 3.95
conditional entropy: 2.54 (just higher than Greek, much higher than the VMS)

This is the second part of my reply to Marco's post. Here I want to focus on this issue of the entropy values of the text, character entropy and conditional entropy.

We must distinguish between two very different steps of the encryption process: 

1) First there is the compression of the alphabet, whereby each group of phonetically related letters is merged into a single letter each. 
Thus 
"t"/"th"/"d" all > [k]
"k"/"kh"/"g" all > [t]
"p"/"ph"/"b" all > [d]  (in the majority of the text where [p] and [f] are not used)
"s"/"z" both > [y]
"m"/"n" both > [l]
"l"/"r" both > [r]
(The vowels unfortunately overlap somewhat and thus are not quite this simple; I consider this a result of the role of the Hebrew alphabet, or a similar writing system, at some stage of the encryption process.)

This is actually the most complicating step and creates the most actual difficulties in reading and interpreting the text. When I refer to the "ambiguation" of the text, I am referring to this step of compressing the alphabet. That is because the resulting characters are ambiguous: [k] could represent "t" or "th" or "d", [r] could represent "r" or "l", etc. 

2) The following step, although it obscures the text on the surface, is actually simpler in principle, and much easier to account for when reading and interpreting the text, once one has figured out the encryption method correctly. It is the classic cryptographic method of "homophonic substitution", which by the way was already known to have been used in northern Italy at the beginning of the 15th century, and so is not anachronistic as applied to the Voynich MS. In this step, for each of the above single cipher characters, the author then created one or more alternate cipher characters that mean the same thing:

[k] > [k] or [s] or [sh] with open loop
[t] > [t] or [q]  ("g" pronounced as a "y" glide in Greek is a separate special case)
[d] > [d] or [p] or [f]  (here only considering the original consonants)
(Note: I believe in the actual composition of the MS that in fact [d] was a substitute for [p] and [f], not the other way around. But it makes no difference in the end result of the final MS text, as long as the author could still recognize which [d]'s were "consonantal" and only (optionally) replace some of them with [p] or [f] in first lines of paragraphs.)
[y] > [y] or [m] or [g]
[l] > [l] or [sh] with closed loop or final [(i)in]
(Again, I am only describing the original consonants in this summary so that it is easier for the reader to follow and understand the basic process.)

I describe this second step as "obfuscation" rather than "ambiguation". It is a classic cryptographic method designed to make letter frequencies harder to recognize. But once such a step is discovered and the encryption method understood, it can be deciphered readily and no longer presents a difficulty in reading and interpreting the text.

Now here is the key point to grasp in terms of the effect of these two distinct steps on the entropy of the text:

The second step, the "homophonic substitution" or "obfuscation" step of the encryption, can be carried out by the author in a very great variety of complicated ways. But the author doesn't have to use such a great variety of complicated ways to carry out this step. An author whose aim is to achieve maximum cryptographic complexity and obfuscation of the resulting cipher text, will choose a very complicated way to perform this step. I believe that Marco did just such a thing in his example of the application of this whole encryption system to the given passage from the Iliad

But the author is not and was not obligated to make the actual application of this step of the process so complicated and obscure. In fact, I conclude that the Voynich MS author actually chose a rather simple and for the most part regular method to carry out this step, with particular alternate characters occurring in particular parts of words and lines, and before or after particular neighboring characters.

The point is, the author's choice of how to carry out this homophonic substitution step, will have a very significant effect on the entropy values of the resulting text. Marco chose to perform this step in a way that created a resulting text with rather high entropy values. But one could just as well carry out this step in a more predictable, regular, systematic way, with the same substitutions being performed in the same parts of words and lines, and before or after the same neighboring characters. This different approach and application of the same encryption system would create a resulting text with much lower entropy values. And this in fact is what I conclude that the Voynich MS author actually did.

This explains why the encryption system that I propose, which could create a resulting text with rather high entropy values, as Marco demonstrates, still at the same time could also create a resulting text with much lower entropy values, such as the Voynich MS text itself. It all depends upon the free choice of the author with respect to how to carry out this homophonic substitution step of the encryption process.

Geoffrey
Koen has asked how my system works for labels such as those 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. , and You are not allowed to view links. Register or Login to view. , and Linda has asked about page You are not allowed to view links. Register or Login to view. .

This post is my reply to both requests.

For the labels, for now I will focus on You are not allowed to view links. Register or Login to view. and You are not allowed to view links. Register or Login to view. . I will address f89 as soon as I can.

My interpretation of each row of labels on You are not allowed to view links. Register or Login to view. and You are not allowed to view links. Register or Login to view. is that the leftmost label in each row is the "category" label for that group of plant roots or leaves, and the other labels in each row identify the individual plant roots or leaves. It seems to me that each row has one more label than plant part illustrations. Also, the leftmost label tends to be the longest and most distinct label word in each row. Perhaps the individual labels are abbreviated, and the category labels are considered more important and so are written out more fully.

In this post I will focus on the leftmost "category" labels in each row. I identify three of them on You are not allowed to view links. Register or Login to view. , and three more of them on You are not allowed to view links. Register or Login to view. . I believe it will be much more difficult to hope to identify abbreviated individual labels of possibly obscure late medieval Greek or Judaeo-Greek names of particular plant roots or leaves. It may require a professional historian of medieval Greek botany and medicine to accomplish that task. But I think we have a chance to identify the category labels, which should be less obscure words.

I will not proceed strictly in order, but rather I group the category labels into logical sets to present together. I think the logic will be clear as I proceed.

I begin with the simplest and least interesting label words: The leftmost labels of the top row of You are not allowed to view links. Register or Login to view. and the middle row of You are not allowed to view links. Register or Login to view. .

leftmost top row label You are not allowed to view links. Register or Login to view. :
[otorchety]
leftmost middle row label You are not allowed to view links. Register or Login to view. :
[otoram]

my readings:
"o-khoriik[o]s"
"o-khorAs"

[Note: For practical purposes I have been treating the Voynich characters [m] and [g] as alternate characters for [y] for some time now. I have added them to my own provisional working letter correspondence table. I have certainly never interpreted [m] and [g] in any other way in any previous readings or interpretations.]

my interpretations:
ο χωρικ[ο]ς
ο χωρος

translations:
"the rustic/regional/village/country"
"the field/farm/estate ; the country/land/region"

Comments:
These are perfectly normal standard Greek words, and perfectly appropriate as labels for plant roots and leaves, although I admit that personally I find these particular labels rather pedestrian, boring, and not terribly informative: I would think such descriptions would apply to an awful lot of plant roots and leaves. Be that as it may, I see nothing wrong with either of these label words.
If one prefers a plural word interpretation for the adjective in the first category label, an alternate interpretation as "ο[ι] χωρικ[ε]ς" is possible.

Next I identify two more specific botanical labels:

leftmost middle row label You are not allowed to view links. Register or Login to view. :
[otaldy]
leftmost top row label You are not allowed to view links. Register or Login to view. :
[okolyd]

[Note: I suspect that [okolyd] may simply be a misspelling of the much more normal-looking Voynich word [okoldy].]

my readings:
"Okan[th?]us"
"Otomsu" (misspelling of "Otomus")

my interpretations:
ακανος / [font=Arial, Helvetica, sans-serif]ακανθος[/font]
[font=Arial, Helvetica, sans-serif][font=Arial, Helvetica, sans-serif]ατομος[/font][/font]

translations:
"thistle-head" / "bearsfoot" (cf. "acanthus")
"uncut"

Comments:
For the first label, another closely related Greek word is "ακανθα", meaning "any thorny or prickly plant". Likewise an acanthus is a prickly plant with toothed leaves. The toothed leaves at least are a good fit for the plant illustrations in this row, although such illustrations certainly occur elsewhere as well.
The second label is quite a good fit for its row: the roots in these illustrations are indeed uncut, whereas many other roots in the other illustrations on these pages are cut open.

Finally, the last two labels on these pages must be treated individually:

leftmost bottom row label You are not allowed to view links. Register or Login to view. :
[daramdal]

my reading:
"phalAs-phan[e]"

my interpretation:
φαλλοσ-φαν[η]

translation:
"phallus-like [ones]"

Comments:
Some people may find such an interpretation outrageous, but if one looks at the shapes of the roots in the illustrations in this row, it must be said that the label is at least not inaccurate, if one chooses to view things in this way.

leftmost bottom row label You are not allowed to view links. Register or Login to view. :
[ofysky-dal]

my reading:
"opsthds-phan[e]"

my interpretation:
οπ[ι]σθ[ι]δ[ιο]ς-φαν[η]

translation:
"hinder-part-like [ones]"

Comments:
The author does omit a lot of vowels in this word, but to be fair the author was short of space for such a long word in this particular place on the page. As a comparison, when I am making grammatical notes about a language I am studying, I use "ptcpl." as a standard abbreviation for the word "participle". Given the context, the all-consonant abbreviation does not present any barrier to comprehension. This word likewise is long enough that I consider that it would be the same: "οπ'σθ'δ'ς" seems to me like its meaning would be fairly clear to a Greek speaker and reader.
"οπισθιδιος" means "hinder, belonging to the hinder part".

=======

Now to You are not allowed to view links. Register or Login to view. , which Linda has asked me about.

For now I can only address the prominent letters running down the left side margin of the top half of the page:

4th line: [s]
6th line: [d]
8th line: [q]
11th line: [s]
14th line: [o]
17th line: [l]
21st line: [k]
24th line: [r]
29th line: [s]

Treating them as one word or phrase, we have:
[sdqsolkrs]

My reading:
"dukt-Ondrd"

My interpretation:
δυκτ[οι]/διακτ[οι]-ανδρ[ι]δ[ες]

Translation:
"women carried through pipes"

Comments:
The Greek adjective "διακτος" is found in Liddell & Scott's A Greek-English Lexicon, with the meaning "carried through pipes, of oils or unguents used at the bath". Citations include "τα κατ' ανδρα δ[ιακτα] IGRom.4.860 (Laodicea ad Lycum)." Such citations suggest late classical Greek influenced by Latin, and I note the Medieval Latin word "ductus", meaning "conveyance (of water); hence, a channel". Thus the first vowel "u" in this word may not be so surprising. I further remind readers that by the medieval period the Greek vowel letters upsilon and iota were pronounced identically.
The Greek word "ανδρις, ανδριδος" is also found in A Greek-English Lexicon, with the meaning "fem. of ανηρ" (the familiar root and word ανηρανδρος "man"), "woman", and the citation "Sm.Ge.2.23."

Geoffrey
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23