geoffreycaveney's Judaeo-Greek theory

[geoffreycaveney]

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:

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:

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