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
I must begin this post with a cautionary note that reading any individual line of the ms text must be treated as tentative and speculative, at our present stage of knowledge of the ms text (i.e., very little).

But this particular line came together so unexpectedly well, that I would like to share it with those who may find it interesting. (Others may well consider all such tentative speculation pointless and unproductive at this stage.)

I was actually investigating the distribution of [p] and [d] in the first lines of paragraphs (see the thread "glyph [d] as a substitute for [p] and [f]"). To the extent that I had any particular language in mind when I began to look at the 1st line at the top of You are not allowed to view links. Register or Login to view. 1, it was a Romance language such as Old Occitan or Middle French (see the thread "Old Occitan troubadour cryptograms...").

But with my provisional guesses about possible and logical phoneme values of certain characters and series of characters, I read the third word in this 1st line of You are not allowed to view links. Register or Login to view. 1, [epairody], as "ipeirous". I wasn't trying to force this reading; I didn't even have the correct language in mind as I transcribed it. The one letter here which will be different from most other proposed transcriptions will be [d] as "u", which I have discussed in the "glyph [d]" thread.

As I looked up Romance languages, I found instead that I had stumbled upon the Greek word for "continents".

***I wish to point out that VViews' blog post on the red text on You are not allowed to view links. Register or Login to view. 2 makes the insightful point that in many medieval texts, red text highlights and indicates a different language than the rest of the text. Although the opening 4-line paragraph at the top of You are not allowed to view links. Register or Login to view. 1 is not written in red, it is very prominent text on the adjacent and related page. It is possible that certain parts of this section may be written in Greek, while much of the rest of the ms may be written in a Romance language. This would surely wreak havoc with all of the statistical calculations performed on the entirety of the ms text.***

At this point I went back to my Greek grammar and phonology notes, and took a closer look at the rest of the line. Again, with a set of provisional guesses about possible and logical phoneme values of certain characters and series, I transcribed the entire line as follows:

[teeodaiin shey epairody osaiin yteeoey shey epaiin oaiin]

"geiopan tis ipeirous otan skiiois tis epan oan"

I note that in the first word, [d] occurs after an initial gallows character [t]. In a comment in the "glyph [d]" thread, I had proposed the hypothesis that the scribe may have substituted [d] for [p] and [f] wherever a gallows character had already occurred in the same word. My reading in this line here would be an example of such a substitution, as I read [d] as "p" rather than "u"/"v".

With these two critically significant readings of [d] in [teeodaiin] and [epairody], the rest of the line of text comes together as Greek strikingly smoothly, with only a couple minor and natural adjustments of closely related vowels. Such variation is only to be expected since we are dealing with the late Byzantine period of Greek, which will not be exactly the same as either Ancient Greek, Koine Greek, or Modern Greek.

"geio" = Earth
"pan" = all, the whole
"tis ipeirous" = the continents
"otan" = when
"skiiois tis" : read as "skiais t(a)is" = in the shadows (This is the Ancient Greek dative plural form, or an archaic expression in medieval Greek, and it could express the locative sense of "in")
"epan" : read as "eipan" = they said
"oan" : read as "oun" = thus, then, as
"eipan oun" = as they said

Thus the whole line may be read with the following meaning:

"When the whole Earth and all the continents are in the shadows, as they said"

=====

Detail notes on the provisional system of phoneme values for characters here:

[t] can be /k/ or /g/
[s] and [sh] are /t/  This is very logical for Greek since the common single-character word [s] could be the definite article forms "to", "tou", ta", abbreviated as "t' ".
[y] is /s/
[r] is /r/
[-iin] is /-n/  Thus [osaiin] is [o+s+a+iin] = /o+t+a+n/, whereas the [i] in [epairody] is not part of [iin], so it is part of the diphthong [ai] = /ei/.
[p] is /p/

[e] can be /e/ or /i/  This is not so surprising for medieval Greek, since a famous sound change from Ancient Greek to Modern Greek called "iotacism" made a large number of vowels that used to be pronounced /e/ or /ei/, all come to be pronounced as /i/.
[ai] is /ei/ 
[a] is /a/
[o] is /o/

Note: my interpretation of [t] as alternately /g/ or /k/ is significant. It indicates that the Voynich script may not distinguish between voiceless/voiced pairs such as k/g, t/d, p/b, s/z. This feature, along with the substitution of [d] for [p] and [f] in many environments, would have a dramatic effect in lowering the entropy, conditional entropy, and character pair distribution plots when performing statistical analysis on the text. 

One way to test this, would be to take regular texts in Greek, Latin, Italian, French, or what have you, and change all the voiced phoneme letters to voiceless phoneme letters, g>k, d>t, b>p, z>s, as well as p>v and f>v to account for the hypothesis about [d] replacing [p] and [f]. (Also, write "u" and "v" the same, as in Classical Latin.) Then do the entropy and character pair distribution analysis on these adjusted texts, and see how their statistics compare with the Voynich ms text.

=====

As I emphasized at the start of this post, of course this is only one line and as such it must be considered as tentative and speculative. Nevertheless it arose naturally from a logical investigation of the phonological system as a whole, not from an attempt to force one particular interpretation in one particular language on the ms text. As I said above, I thought I might be looking at a medieval Romance language, when I stumbled upon "ipeirous" instead, and this piece of data forced me to go back and consider a Greek reading again.

As a provisional approach to relate characters to phonemes, this is still at a very tentative stage. As an interesting speculation about the possible content of this 4-line introduction to the astrology / astronomy section of the ms, I thought it was worth sharing.

Geoffrey Caveney
I've managed to do this in Greek, get full sentences out of it.

But I've also managed to do it in Portuguese and Spanish (in the big-plants section) and, to a lesser extent, Latin. I was even able to do it, to some extent, with the old Baltic languages. Also, to a lesser extent with old Persian but it's a language I don't know, so it's harder to recognize how it might work and I couldn't work out full phrases as well as in the Romance languages. I haven't had much success with Old or Middle English.

But when I try to generalize the system (even in a loose way) to full paragraphs, or even to many of the labels, it falls apart except for little "islands" of text here and there.
One problem comes to my mind immediately: "pan" is a prefix, it cannot go behind the root word - that's ungrammatical nonsense. 

According to wiktionary, it's only a suffix in Nahuatl, which is funny in its own way.
The most common words in the Voynich text are:

daiin
ol
Chedy
aiin
Shedy
chol
or
ar
Chey
dar
qokeey
qokeedy
Shey
qokedy
dy

I would be grateful if you could give your readings of these words and their meanings.
(14-03-2019, 11:03 PM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view....


Detail notes on the provisional system of phoneme values for characters here:

[t] can be /k/ or /g/
[s] and [sh] are /t/  This is very logical for Greek since the common single-character word [s] could be the definite article forms "to", "tou", ta", abbreviated as "t' ".
[y] is /s/

...

y is virtually always at the beginnings or ends of tokens (or standalone).

If you assign phonemes to Voynich glyphs, you would have to account for why a letter as common as "s" (if represented by y) is found only at the beginnings and ends. In other words, you have to resolve the issue of spaces between tokens and even if the spaces are taken with a grain of salt, the positionality of y is still highly significant (and not distributed within words as they are in Greek).
(15-03-2019, 01:31 AM)Koen G Wrote: You are not allowed to view links. Register or Login to view.One problem comes to my mind immediately: "pan" is a prefix, it cannot go behind the root word - that's ungrammatical nonsense. 

According to wiktionary, it's only a suffix in Nahuatl, which is funny in its own way.

A couple possible explanations:

1) The word break in the ms text could be in the wrong place, either accidentally or deliberately, and the phrase could be "geio (read "gaia") pantes ipeirous" = "Earth (and) all the continents". The case and gender of "pantes" and "ipeirous" do not exactly match, but we have to remember the early 15th century was very late Byzantine Greek, it would have been different from Ancient Greek, different from Koine Greek, and different from Modern Greek. Or the 15th century author could have made a slight grammatical mistake in the case and gender of the Greek.

2) The word "geio-pan" could be word play in Greek, along the lines of the popular French word play game "Verlan", in which the syllables of words are inverted.

Actually, if there is systematic rearrangement of syllables in the Voynich ms text, that could explain a lot of things.
Quote: Geoffrey: 1) The word break in the ms text could be in the wrong place, either accidentally or deliberately,

I agree that wrong word breaks should be on the list of possible explanations, but I don't think there are any accidental word breaks in the VMS (if there are, the incidence is very low).

I've created four transcriptions of the entire manuscript and have noticed that the position of glyphs within tokens is very consistent throughout the manuscript. If it deviated, it would be noticeable.
(15-03-2019, 01:45 AM)-JKP- Wrote: You are not allowed to view links. Register or Login to view.
(14-03-2019, 11:03 PM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view....


Detail notes on the provisional system of phoneme values for characters here:

[t] can be /k/ or /g/
[s] and [sh] are /t/  This is very logical for Greek since the common single-character word [s] could be the definite article forms "to", "tou", ta", abbreviated as "t' ".
[y] is /s/

...

y is virtually always at the beginnings or ends of tokens (or standalone).

If you assign phonemes to Voynich glyphs, you would have to account for why a letter as common as "s" (if represented by y) is found only at the beginnings and ends. In other words, you have to resolve the issue of spaces between tokens and even if the spaces are taken with a grain of salt, the positionality of y is still highly significant (and not distributed within words as they are in Greek).

There are several possible explanations:

1) Given the general tendency toward uniformity in Voynich ms text "word" length, it is plausible that many syllables are written as separate words. If so, many more "s's" will occur at the beginnings or ends of "words" than if the syllables were all written together as words in the standard forms that we are familiar with.

2) Some scripts such as Arabic and Syriac have certain letters that cannot "connect" to a following or preceding letter. Perhaps the Voynich scribe adopted this feature in his script, and the character [y] is "unconnected" to its right, which appears to us as a word break, whether it is actually a word break or not.

3) Elsewhere we have discussed the idea that [y] looks like the medieval Latin ms abbreviation symbol for "-us", etc. Perhaps the Voynich scribe adopted it for a broader variety of letter sequences, such that [y] could also represent the final sequence "-si" in Greek "etsi" for example, or also such that [y] could represent the initial sequence "eis-" in Greek "eiste" (or "es-" in Old French "estont", etc.). Basically, this would mean that the scribe did not write any vowel occurring before "s" at the beginning of a word, or after "s" at the end of a word. Such a thing sounds strange to us, but it would not seem all that unusual in the highly abbreviated world of medieval ms writing before the invention of movable type and the printing press.

4) It is possible that another character, such as [l], also represents "s". I do not prefer this solution, since it means one less character available to represent another phoneme. (I would prefer to have [l] represent "n", for example, so that "n" could sometimes occur in other positions than word-final, as it must when represented as [n].) But the character [l] does not occur in the particular line of text that I transcribe in this post, so with this hypothesis it remains an open questions what the phonemic value of [l] may be.

I'm sure there are other possible explanations, but those are the ones that occur to me off the top of my head.
(15-03-2019, 02:37 AM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view.
(15-03-2019, 01:45 AM)-JKP- Wrote: You are not allowed to view links. Register or Login to view.
(14-03-2019, 11:03 PM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view....


Detail notes on the provisional system of phoneme values for characters here:

[t] can be /k/ or /g/
[s] and [sh] are /t/  This is very logical for Greek since the common single-character word [s] could be the definite article forms "to", "tou", ta", abbreviated as "t' ".
[y] is /s/

...

y is virtually always at the beginnings or ends of tokens (or standalone).

If you assign phonemes to Voynich glyphs, you would have to account for why a letter as common as "s" (if represented by y) is found only at the beginnings and ends. In other words, you have to resolve the issue of spaces between tokens and even if the spaces are taken with a grain of salt, the positionality of y is still highly significant (and not distributed within words as they are in Greek).

There are several possible explanations:

1) Given the general tendency toward uniformity in Voynich ms text "word" length, it is plausible that many syllables are written as separate words. If so, many more "s's" will occur at the beginnings or ends of "words" than if the syllables were all written together as words in the standard forms that we are familiar with.

I'm open to the possibility that Voynichese is text that has been broken into syllables (or in which the spaces are not necessarily literal). I've suggested the idea of syllables a few times (the suggestion didn't get a very good reception but I still consider it a possibility). If we break these into syllables

Καλησπέρα   [b]Κα λη σπέ ρα
Ευχαριστώ    [b]Ευ χα ρισ τώ
Ασπρο πάτο  [b]Ασ προ πά το

the "s" appears at the beginning in the first example, and at the end, in the second and third examples, but note the great variation of letters that precede and follow the "s" if you do this with a list of words. This doesn't happen in Voynichese.

Quote:3) Elsewhere we have discussed the idea that [y] looks like the medieval Latin ms abbreviation symbol for "-us", etc. Perhaps the Voynich scribe adopted it for a broader variety of letter sequences, such that [y] could also represent the final sequence "-si" in Greek "etsi" for example, or also such that [y] could represent the initial sequence "eis-" in Greek "eiste" (or "es-" in Old French "estont", etc.). Basically, this would mean that the scribe did not write any vowel occurring before "s" at the beginning of a word, or after "s" at the end of a word. Such a thing sounds strange to us, but it would not seem all that unusual in the highly abbreviated world of medieval ms writing before the invention of movable type and the printing press.

I've written quite a few blogs about this, so obviously I don't disagree that y might stand for a sequence of letters (might be an abbreviation) and that the expansion might differ depending on whether it's at the beginning or the end (as in Latin).

Quote:4) It is possible that another character, such as [l], also represents "s". I do not prefer this solution, since it means one less character available to represent another phoneme. (I would prefer to have [l] represent "n", for example, so that "n" could sometimes occur in other positions than word-final, as it must when represented as [n].) But the character [l] does not occur in the particular line of text that I transcribe in this post, so with this hypothesis it remains an open questions what the phonemic value of [l] may be.

I'm not too thrilled about this explanation either. The VMS character set is already very constrained, with a small proportion of the glyphs representing the majority of the text. Compressing it further seems untenable.



I wish I had more time to respond to this, but I have to go. I have some important deadlines.
(15-03-2019, 01:38 AM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.The most common words in the Voynich text are:

daiin
ol
Chedy
aiin
Shedy
chol
or
ar
Chey
dar
qokeey
qokeedy
Shey
qokedy
dy

I would be grateful if you could give your readings of these words and their meanings.

Of course, as I noted to begin my original post, this conjecture must remain tentative and speculative at our present stage of very low knowledge of the text.

But sure, I'll give you some of my conjectures about possible readings for the most common words in your list:

[daiin] : "pan-" where the ms text represents Greek, "van" where it represents a Romance language ("-van" or "-ban" can also be the suffix of many Romance 3rd person plural verb forms, especially in the past or imperfect tenses)

[ol] : the suffix "-on" or "-os", depending on which hypothesis one chooses for [l]. Both are extremely common suffixes in Greek, and in some Romance languages as well. If we can read the vowel as "-un" or "-us", a very natural linguistic variation, that creates even more possible suffixes that this word could represent.

[chedy] : the suffix "-ius" or "-eus". [ch] may be the glide /j/, or the letter "h" which is silent in most of these languages. Actually, it could also be "-ibus" with my theories about [d] -- now it's starting to look like Latin after all!

[aiin] : "-an", a ubiquitous suffix in many languages. Also the Greek particle "an" meaning "if". If more word play is involved here, it could also represent the even more ubiquitous Greek subjunctive particle "na", the most common single word in Modern Greek.

[shedy] : I believe there are multiple forms of the Voynich character we transcribe as "[sh]". The most frequent form is the one I interpret as Greek "t" in the 1st line of You are not allowed to view links. Register or Login to view. 1 in the original post. This gives [shedy] = "-tius" or "-teus", or possibly "-tivus" or "-tibus". Again, these are starting to look more like Latin suffixes than I had expected. 

I also believe an alternate form of [sh] may represent the phoneme /m/, which I think must be represented in the text somehow. In these cases, [shedy] = "-mius", "-mivus", "-mibus", or "meus". Looks very much like Latin again.

[chol] : "-ion" or "-ios", again depending on the hypothesis for [l]. We sure are getting a lot of very common Latin and Greek suffixes in this list.

[or] : "-or". Also a very common Latin suffix.

[ar] : "-ar" or "al". I believe the only way to account for the underlying language(s)' entire phonemic inventory may be for the script to represent the liquids "r" and "l" both with the same character [r].

[chey] : "-ies". Another common suffix.

[dar] : "var-", "val-", and various others. Speaking of "various", I can write it in Voynichese as [dar chody].

[qokeey] : Well now these words depend on one's theory about [q]. I tend to think it must be an ampersand. Now about [k]. This is an open question in the hypothesis in this post, since [k] did not occur in the line I transcribed. But I tend to think it is part of the series of stops with [t] = /k/, /g/ and [p] = /p/, /b/, so I suspect that [k] = /t/, /d/, and [s] and [sh] are simply alternate ways to write /t/ or /d/. 
Given all of these suppositions, I propose that [qokeey] = "& autes", Greek for "and they", "and these", "and those" (feminine plural). In the bathing women section of the ms, the frequent repetition of these feminine plural pronouns would make a lot of grammatical sense. It seems natural enough to me to represent the Greek diphthong "au" simply as "o"=[o].

[qokeedy] : "& auteeus". Perhaps the vowels need to be slightly re-interpreted here to make the Greek ending grammatical. I still suspect that this is also another form of the same Greek pronoun as [qokeey].

[shey] : "tis" a form of the Greek definite article, as seen twice in my interpretation of the 1st line of You are not allowed to view links. Register or Login to view. 1. The letter sequence "-tis" could also occur in many other places in both Greek and Latin.

[qokedy] : See [qokeedy] above.

[dy] : "-us", perhaps the most frequent suffix in Latin. (I mentioned elsewhere the hypothesis that the Currier A scribe just used [-y] to express "-us", akin to the medieval Latin ms abbreviation, but the Currier B scribe chose to write the letter "u" separately, as [d], producing the common suffix [-dy] in the B text but not in the A text.)

This has been a most interesting exercise. Thank you for the challenging request that prompted me to work through each of these words/vords/suffixes systematically.

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