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
(09-04-2019, 04:30 PM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view.P.S.: Koen, I do hope that you will be so kind as to read and comment on the set of six plant root and leaf category 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. that I interpreted in my post the other day, as you had asked me to attempt to do with my method. I do understand that it may take some time for you to do so. Thank you.

Right, I missed this among other posts. 
Hmm well, your results are somewhat weird, but that may not be your fault, maybe the manuscript is weird. 
It's just not as illuminating as I would have hoped it would be. I still remain with the impression that your method allows you to reach a lot of results from any given string of Voynichese

Like, if two people used your system on any given paragraph from quire 20, would they reach similar results, or widely different subjects?
(09-04-2019, 05:49 PM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view.
(09-04-2019, 10:25 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.If the source text of the Voynich MS *really* was written in an unpointed Hebrew version of Greek, then there are two possibilities:
1) the person converting it to Voynichese did understand which vowel should be where
2) this person did not understand.

Now the Voynichese text (according to this theory) has vowels, in the right places, but not the right ones.
This does not fit with either option.

It depends how you define "the right ones". As I have already pointed out previously, in Judaeo-Greek the Hebrew letters Aleph and Ayin have identical phonetic values:

[ there was a figure here that did not survive my copying ]

Thus, if Voynich [a] represented Aleph, and Voynich [o] represented Ayin, the author could have used either of them, interchangeably, in any place where either of them could be written in Judaeo-Greek, in free variation, without having any effect on the pronunciation of the underlying text.

But this is not at all what you are doing. I just went to the first example of a translation that I could find:

Quote:first two lines of You are not allowed to view links. Register or Login to view. 1 in the Voynich ms text:

[t]eeodaiin  shey  epairody  osaiin  yteeoey  shey  epaiin  oaiin
[font=Arial, Helvetica, sans-serif]daiir  okeody  qoekeeg  sar  oeteody  oteey  keey  key  keeodal[/font]

my Judaeo-Greek interpretation of this text:

[]ei[A]pan  tis  ipeirous  otan  skiiAis  tis  ,  epan  oAn
par'  Atous  &Atees  tAr(a)  oikous  o(u)k-eis(i)  tees  ,  tes  []ei[A]pan

Now "normalizing" this Judaeo-Greek text into a more standard Greek form:

eipan  tis  ipeirous  otan  skiais  tis  ,  eipan  oun
para  autous  &  autes  tora  oikous  ouk  eisi  tes  ,  tes  eipan

So starting with the Greek text, we have lots and lots of vowels.
In the Judaeo-Greek version, they all survive somehow. That does not fit with an unpointed Hebrew text.
Just some examples (to keep it simple):

tis should have become ts.
para should have become pr.
tora should have become tr.
tes should have become ts.
[ etc ]

eipan could have become Apn (with A the Aleph). There are three of them in this short text.
But two end up as: eiApan
and the third as: epan

and finally in Voynichese they are all different:
[t]eeodaiin
epaiin
keeodal
(09-04-2019, 07:36 PM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.
(09-04-2019, 05:49 PM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view.
(09-04-2019, 10:25 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.If the source text of the Voynich MS *really* was written in an unpointed Hebrew version of Greek, then there are two possibilities:
1) the person converting it to Voynichese did understand which vowel should be where
2) this person did not understand.

Now the Voynichese text (according to this theory) has vowels, in the right places, but not the right ones.
This does not fit with either option.

It depends how you define "the right ones". As I have already pointed out previously, in Judaeo-Greek the Hebrew letters Aleph and Ayin have identical phonetic values:

[ there was a figure here that did not survive my copying ]

Thus, if Voynich [a] represented Aleph, and Voynich [o] represented Ayin, the author could have used either of them, interchangeably, in any place where either of them could be written in Judaeo-Greek, in free variation, without having any effect on the pronunciation of the underlying text.

But this is not at all what you are doing. I just went to the first example of a translation that I could find:

Quote:first two lines of You are not allowed to view links. Register or Login to view. 1 in the Voynich ms text:

[t]eeodaiin  shey  epairody  osaiin  yteeoey  shey  epaiin  oaiin
[font=Arial, Helvetica, sans-serif]daiir  okeody  qoekeeg  sar  oeteody  oteey  keey  key  keeodal[/font]

my Judaeo-Greek interpretation of this text:

[]ei[A]pan  tis  ipeirous  otan  skiiAis  tis  ,  epan  oAn
par'  Atous  &Atees  tAr(a)  oikous  o(u)k-eis(i)  tees  ,  tes  []ei[A]pan

Now "normalizing" this Judaeo-Greek text into a more standard Greek form:

eipan  tis  ipeirous  otan  skiais  tis  ,  eipan  oun
para  autous  &  autes  tora  oikous  ouk  eisi  tes  ,  tes  eipan

So starting with the Greek text, we have lots and lots of vowels.
In the Judaeo-Greek version, they all survive somehow. That does not fit with an unpointed Hebrew text.
Just some examples (to keep it simple):

tis should have become ts.
para should have become pr.
tora should have become tr.
tes should have become ts.
[ etc ]

eipan could have become Apn (with A the Aleph). There are three of them in this short text.
But two end up as: eiApan
and the third as: epan

and finally in Voynichese they are all different:
[t]eeodaiin
epaiin
keeodal

I am glad that Rene has brought up this example of these lines and these words, because my explanation of them in response will be very instructive and helpful for Chen Zhe, Koen, and any other interested researchers to follow my method more clearly.

First of all, one technical correction: Rene is quoting an early version of my reading and interpretation of these lines, which is fair, since I posted it, but I do want to clarify that I have amended one significant detail in a later version (also posted): I now read and interpret the last word of the second line (this is all from the top of page You are not allowed to view links. Register or Login to view. 1, as Rene notes), [keeodal], as "t-eipan", with the prefixed "t-" representing a prefixed pre-verbal clitic object pronoun, which is a distinctive characteristic of Greek as part of the Balkan sprachbund. "t-" simply represents the word "to-", meaning "it", and the vowel is elided before the initial vowel of "eipan". The whole word "t-eipan" means "they said it".

Now about the main issues that Rene raises:

My hypothesis is that the author did make some effort to indicate many of the Greek vowels in some way in the MS text. Not every vowel, and not in a 100% entirely consistent way throughout the text, for better or worse. But some effort to indicate many of the vowels in some way. And yes, in the system that was used for these vowel representations, I do indeed observe certain connections with the Hebrew script and letters. But the connections are complex. It will take some time and effort to explain the system, and it will take some time and effort to comprehend the system.

The basic general principles that I observe in the author's system can be summarized in the following way:

1) Where front vowels, especially /i/, are indicated, the author uses any of the Voynich characters [ch], [e], or (i) (not part of [(i)in]) to do so. [Once again, please note that I cannot write (i) in square brackets because that is a symbol for formatting the entire following text of this post in italic type.]

2) Where back vowels, and sometimes short /e/ without a following y-glide, are indicated, the author uses either of the Voynich characters [a] or [o] to do so.

3) Where rounded vowels are indicated, the author uses the Voynich character [d] to do so.

Unfortunately, this system creates some potential overlap in the representation of "e", which can fall under either category (1) or (2), and in the representation of "o", which can fall under either category (2) or (3). It also creates some ambiguity in the representation of Greek upsilon, because it was pronounced as /i/ by itself, but still as a /w/ glide as part of the diphthongs "au" and "ou". Thus it could end up in either category (1) or (3), and due to this inherent confusion in its phonetic value, the author may sometimes have used a Voynich character for one category, when the actual pronunciation of the letter in the given context would appear to indicate the other category.

This is not a perfect system, and it is not an unambiguous system. But it is a system nonetheless, and it is the system that I have observed in my reading and interpretation of the MS text.

Vowel principle (1) is at least consistent with the "iotacism" tendency of post-classical Greek for many vowels and diphthongs all to become pronounced identically as /i/. It is also consistent with the existence of the Hebrew letter yod, which represents the glide /y/. I identify Voynich [e] as the purest equivalent of iota/yod, but [ch] is a very frequent alternate representation, which I consider to be often related to the Greek "rough breathing" pronunciation of an /h/ sound before many initial vowels.

Vowel principle (3) is consistent with the Hebrew letter vav, which represents the glide /w/. Elsewhere in this thread I have explained at some length and in some detail how both Hebrew vav and Voynich [d] can indicate the vowel sound /u/ or the consonant sound /v/, and how I believe the author thus came to use [d] to represent Greek beta as well, since Greek beta and Hebrew bet can both also be pronounced /v/. Since voiced and voiceless consonants are not distinguished, Voynich [d] thus came to represent /p/ or /f/ ("ph") as well.

Vowel principle (2) covers the Hebrew letters aleph and ayin that I described in the previous post. In general, they can show up where various different Greek vowels do, but they are definitely not preferred for /i/ or a /y/ glide, for which Hebrew has the letter yod, nor are they preferred for /u/ or a /w/ glide, for which Hebrew has the letter vav. For anything else, either aleph or ayin is "fair game", so to speak, and in this MS text, the Voynich characters [a] or [o] are likewise "fair game" as possible representations of any other vowel, if the author felt it necessary to represent a particular vowel at all.

You may not like these 3 vowel representation principles of the system that I observe. You may think they are too ambiguous. Be that as it may, they are the vowel representation principles that I have observed in my reading and interpretation of the MS text.

And following these 3 principles, I can indeed explain the instances that Rene cites:

"tis" remains "tis", not "ts", because Hebrew has a perfectly good letter yod for the /y/ glide which corresponds to this vowel /i/. And indeed it shows up in this word of the MS text as the corresponding Voynich character [e].

"para" becomes "par", not "pr", because the author did want to make some effort to indicate some vowels in some way. The author did not choose to employ nikkud dots for this purpose, but as an alternative in some places in the text, such as in the middle of this word, the author used the equivalent of Hebrew aleph, the Voynich character [a], to represent the Greek vowel sound /a/. Neither Hebrew yod nor vav work for this vowel, so aleph or ayin were the options the author had, and in this case he or she chose aleph, that is, Voynich [a].

"tora" becomes "tar", not "tr", for the same reason as "para" becomes "par", not "pr". I find that the author did not frequently represent Greek /o/ with Voynich [d], although it may occur occasionally. More often the author includes Greek /o/ in "vowel category (2)" above, rather than vowel category (3). So for better or worse, I find in this text that Greek /a/ and /o/ were usually not distinguished from each other.

"tes" remains "tes", not "ts", because the author felt it important to indicate this vowel in some way in this word. In this particular case the author chose to emphasize that it is a front vowel, not /a/ nor /o/ for example, and so he or she here treated this /e/ as "vowel category (1)" rather than category (2). Thus it is represented here with the equivalent of Hebrew yod, that is, Voynich [e].

Now about the three "eipan"s in these two lines:

Hebrew has the perfectly good letter yod for the /y/ glide, which is perfectly natural to use to represent the Greek vowel sound /i/. Now the Greek spelling remains "ei", but this entire diphthong is also simply pronounced /i/ due to the "iotacism" sound change of post-classical Greek. In these lines, in two cases the author wrote the vowel as the double letter Voynich [ee], but in one case as the single letter Voynich [e]. I do not see much difference in these two representations; some medieval spelling variations were simply arbitrary.

Likewise it is perfectly natural for the author to wish to indicate the Greek vowel /a/ in "eipan" in some way. In all three cases here the author chose Voynich [a] to do so, in line with the choice in the word "para" above.

I explained the [k] in [keeodal] above, and I note here that I consider the [t] in [teeodaiin] as a pilcrow, not a letter.

I have explained elsewhere how Voynich [d] and [p] may both represent Greek "p". I have explained Voynich [d] in more detail above.

I have explained elsewhere how Voynich final [(i)in] is an alternate representation of the same sounds as Voynich [l].

The most significant difficulty in explaining the vowel representations in these forms is the presence of Voynich [o] between [ee] and [d] in the first and third cases [(t)eeodaiin] and [keeodal]. The only thing I can say about this at the present stage is that I note the well-known (to Voynich researchers) distinction between "Currier A dialect", which almost never writes the combination [ed], and "Currier B dialect", which very frequently writes the combination [ed]. But Currier A dialect does use the combination [eod]. So these forms here appear to reflect the "Currier A dialect" style, avoiding the combination [ed].

Geoffrey
(09-04-2019, 07:16 PM)Koen G Wrote: You are not allowed to view links. Register or Login to view.
(09-04-2019, 04:30 PM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view.P.S.: Koen, I do hope that you will be so kind as to read and comment on the set of six plant root and leaf category 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. that I interpreted in my post the other day, as you had asked me to attempt to do with my method. I do understand that it may take some time for you to do so. Thank you.

Right, I missed this among other posts. 
Hmm well, your results are somewhat weird, but that may not be your fault, maybe the manuscript is weird. 
It's just not as illuminating as I would have hoped it would be. I still remain with the impression that your method allows you to reach a lot of results from any given string of Voynichese

Like, if two people used your system on any given paragraph from quire 20, would they reach similar results, or widely different subjects?

Here is a good test to answer your question: Have a different person who also knows enough Greek grammar (ancient and modern) tackle the first two lines of the final paragraph on page You are not allowed to view links. Register or Login to view. (the final paragraph is only the final five lines of text). See if they can use my same system to come up with any alternate reading and interpretation, significantly different from my reading and interpretation, with internally coherent semantics as plausible as my interpretation (albeit different), and above all with Greek phonology, morphology, syntax, and all grammatical characteristics as good as my interpretation.

I caution any takers of this challenge in advance: In general, it takes me about 8-10 hours of concentrated effort to read and interpret one line of Voynich MS text with my system. And I am the one who worked out the system, so naturally at this stage right now I know it better than anyone else does. So one can expect that it will take another person at least 20 hours to complete this task, and possibly 40-50 hours or more, of concentrated effort.

(In case anyone is curious: at this rate, working 8-10 hours every day, 5 days a week, I estimate that it would take me about 20 years of concentrated effort to read and interpret the entire Voynich MS text using my method at the pace that I am able to do so currently.)

Since I just provided a summary of the "3 vowel representation principles" (or the 3 (overlapping) vowel categories) of my system in my latest reply to Rene, and I have frequently in this thread described the 6 phonological categories of consonant phonemes distinguished by distinct sets of Voynich characters in the MS text according to my system, it may be useful here to summarize all 9 categories together, in brief. Basically, this will serve as an organized summary description of the same information contained in my letter correspondence table. Here are the 9 categories:

1) front vowel characters: [ch], [e], or (i) can represent /i/ or sometimes /e/
2) low/mid vowel characters: [a] or [o] can represent /a/, often /o/, or sometimes /e/
3) rounded vowel character: [d] can represent /u/ or occasionally /o/ [Note: [d] can also be used as a consonant; see below]
[Caution: since Greek upsilon is spelled as a "u" but pronounced as an "i" in medieval Greek when used by itself, it may be represented by either a "front vowel character" or a "rounded vowel character"]
4) liquid consonant character: [r] can represent /r/ or /l/
5) nasal consonant characters: [l], final [(i)in], or [sh] with a closed loop on top, can represent /m/ or /n/
6) sibilant consonant characters: [y], [m], or [g] can represent /s/ or /z/ [Note: at the end of a word, such a character may often represent the entire suffix "vowel+/s/"]
7) labial consonant characters: [d], [p], or [f] can represent /p/, /f/ ("ph"), /b/, or /v/
8) dental consonant characters: [k], [s], or [sh] with an open loop on top, can represent /t/, /th/, or /d/ (pronounced in post-classical Greek as "dh" or voiced "th", as in English "the")
9) guttural consonant characters: [t] or [q] can represent /k/, /x/ ("kh" or "ch"), or /g/ 
[Note: post-classical Greek gamma is a special case since it can be pronounced as a /y/ glide, in which case it may fall under vowel category (1) above]

Geoffrey
(09-04-2019, 11:02 AM)ChenZheChina Wrote: You are not allowed to view links. Register or Login to view.(continued)

1. The omission rules are not stable.

You have used Linear B as an example where Greek could be written with a hugely different script. However, according to You are not allowed to view links. Register or Login to view., the omission rules seem quite stable and constant for me. For example:

Omission Rules:

  1. Final -l, -r, -n, -m, -s and some diphtongs cannot be expressed in Linear B, so they are simply omitted.
  2. Consonant clusters are sometimes expressed by inserting an adjacent vowel.
  3. However, initial s in consonant clusters is usually omitted.
  4. Linear B has only k/q, so k/kh/g are not distinguished.
  5. Linear B has only p, so p/ph/b are not distinguished. Except a possible phu is used sometimes.
  6. Linear B has d/t, so d/t are distinguished, but t/th are not.
  7. Linear B has only r, so l/r are not distinguished.
  8. Linear B has j/w, so they are properly used.
And let us see some example of words:

a-pi-qo-ro = amphiquoloi (1) (5) (7)
a-te-mi-ti-jo = artemitios (1) (6) (8)
a-to-ro-qo = anthrōquos (1) (6) (7)
ko-no = skhoinos (1) (3) (4)
ko-no-so = knōsos[/i] (1) (2) (4)
ku-ru-so = khrusos (1) (4) (7)
ku-wa-no = kuanos (1) (4) (8)
o-da-tu-we-ta = odatwenta (1) (6) (8)
phu-te-re = phutēres (1) (5) (6) (7)
pu-te = phutē (5) (6)
qo-u-ko-ro = guoukoloi (1) (4) (7)
ti-ri-po = tripos (1) (2) (5) (6) (7)
tu-ri-so = tulisos (1) (6) (7)

They all fit the simple rules above quite well.

As we know, without relatively fixed rules, it would be hard to interpret from defected script back into original script.

For example, last month, I heard a Japanese saying a new word I’ve never heard before: sokui. To figure out what it might be, I have done these things in my mind.

  1. From context (The Sokui of the Emperor is going to be in May) I know it’s a verb-noun or noun, so adjacents are not possible.
  2. The k is a short consonant, not long consonant, so this is probably a Sino-Japanese word or a native Japanese word, but almost impossible to be an English word.
  3. If it is a native Japanese word, there is no way I could figure it out by guessing. Go on with Sino-Japanese word.
  4. In Sino-Japanese word, each kanji corresponds to 1 or 2 morae, so so-ku-i must be at least two kanji.
  5. I searched in my mind and found no Kanji could be read as kui, so soku-i is the most possible candidate. So-ku-i is also possible, but most Sino-Japanese words are 2-kanji words. We’ll come back later if soku-i fails.
  6. Japanese i usually corresponds to wei in Mandarin Chinese, with 位 (seat, place) being very common. Because Chinese wei was transliterated into Middle Japanese wi, and then Middle Japanese wi became Modern Japanese i.
  7. At this step, a Chinese word 即位 jíwèi (enthrone, enthronement) came into my mind. This word fits the context well.
  8. I verified the Japanese reading of 即 and it is certainly soku.
  9. Resolved the new word sokui as enthronement.

As you could see, there are many rules that helped me during the deduction.

However, in your You are not allowed to view links. Register or Login to view., you have to add arbitrary suffixes:

Quote:interpretation:
" ouk-ouden  an  [e]meis  kathar'[a]  ethelom[en]  autes  ti'-[o]poies  ekheis  ou  ethel'[ei]  oks[e]os "

" ουκ-ουδεν  αν  [η]μεις  καθαρ'[α]  εθελομ[εν]  αυτες  τι'-[ο]ποιες  εχεις  ου  εθελ'[ει]  οξ[ε]ως "

I’m not a professional in Greek, so I cannot be certain if these suffixes could be omitted in this way but still recognized by readers. I would place a question mark here.

(This is long, so I’ll write in separate posts)

Chen Zhe: Once again, thank you for your detailed commentary on my theory. I am grateful and appreciative for the careful attention that you have given to my ideas.

I explained the logic behind the possibility of omitting "[e]" in "[e]meis" in a You are not allowed to view links. Register or Login to view. in this thread.

The word "kathar'[a]" is an adverb: "purely, neatly, plainly, clearly". Even if one were to interpret this word as an adjectival form somehow, the overall meaning of the sentence would not change much: "...we purely [don't] want/wish for..." vs. "...we pure ones [don't] want/wish for...." In this word, with this root, in this place in the sentence, the suffix simply can't affect the meaning in any very significant way. 

Moreover, in this case in Greek pronunciation we have elision of the final vowel before the initial vowel of the following word. In pronunciation, it is actually more natural to drop this vowel than to pronounce it. Modern example to illustrate the point: The second line of the Greek folk song "Pes mou pos ginete" begins with the words "αφου εχεις..." ("aphou ekheis..."). You can read the lyrics You are not allowed to view links. Register or Login to view.. Now listen to Greek singer Haris Alexiou sing the line at 0:25 to 0:30 of You are not allowed to view links. Register or Login to view.. You will hear that she sings "af-ekhis...." She does not pronounce the vowel of the final diphthong of the first word at all, even though it has the stress accent of the first word on it.

By far the most significant omission is that of "ethelom[en]". But I note first of all that I found this same abbreviation of this same suffix, the 1st person plural verb ending, three more times in the following line alone. The MS text is at least consistent in this respect. Second, the meaning of the abbreviated suffix is not in doubt, because the subject "we" was explicitly stated in the word "[e]meis". In short, I see no way that the words " 'meis ethelom' " could possibly be unclear to any Greek speaker and reader. I note further that in modern Greek, the "-n" of this suffix has already been dropped completely.

In fact, if in the author's late medieval dialect the "-n" of the suffix had already been dropped, then the further elision of the final vowel before the initial vowel of the following word would also be natural and normal in Greek.

The only other arbitrary suffix in this line is in "ethel'[ei]". First of all, once again, and most importantly, this is above all elision of a final vowel before the initial vowel of the following word. Again, this is completely natural and normal in Greek.

In this case, it is the second part of the verb phrase "ekheis...ethel'[ei]", meaning "you-have...wanted/wished for". In fact, in this post-classical Greek grammatical construction, the syntax is rather similar to modern English. The omitted suffix "[ei]" is rather similar to the "ed" at the end of "wanted" or "wished". So this omission creates a Greek phrase rather akin to an English phrase like "you have...want" or "you have...wish". The meaning would not be unclear at all. But more importantly, the elision of the final vowel before a following initial vowel is the natural and normal pronunciation in Greek.

Geoffrey
(10-04-2019, 01:03 AM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view.(In case anyone is curious: at this rate, working 8-10 hours every day, 5 days a week, I estimate that it would take me about 20 years of concentrated effort to read and interpret the entire Voynich MS text using my method at the pace that I am able to do so currently.

If you're working on the "right" solution, it will speed up a lot. Because there must be a system and you will discover it.

If you're working on the "wrong" solution, you will stall. You need to keep adding more and more freedom.
In a recent mail I just saw yet another, new, possible meaning of the Voynich character d , where it could represent Greek 'ai'.

Until now, you have a minimal amount of plain text, and the relation of this plain text with the Voynich text is extremely thin.

For an example of a 'translation' with fewer degrees of freedom than yours, with a longer plain text, that seems to be somewhat on topic, see the paper "The cannabis page of the Voynich Manuscript".
Link: You are not allowed to view links. Register or Login to view.

For me, both are equally wrong, but if you think that you are right and he is wrong, you have to be able to point to a clear reason why his is wrong and you are not.
(09-04-2019, 09:36 AM)Koen G Wrote: You are not allowed to view links. Register or Login to view.Zhe: all things individually might be possible, but I have some reservations.

1) When you turn English into abjad spelling (I'd really prefer to call it "dropping vowels" since abjads are something very specific), you notice that there are certain things you cannot do. For example, it's often very hard to drop initial vowels. The word "initial" would become "ntl", which is hard to revert properly. However, in his decipherment Geoffrey does take such liberties with adding vowels.
2) In isolation every aspect of Geoffrey's method works to some extent. But what if you combine them, especially consonant flexibility and vowel dropping? 

I'm not completely certain yet, but for now my impression is that it would be a one-way cipher. Flexibility is so great that the initial meaning is not recoverable.

Hi, Koen.

First, I have to say that I am sorry because I had to leave in a hurry yesterday.

I also have concern about initial-vowel-dropping, but I think it could be split into two smaller questions.

1.1 Dropping initial vowels from short, common words.

I think it is mostly fine to drop initial vowels from shortest and most common pronouns and prepositions. For example, I think it could be probably fine if we write “am” as “m”, “and” as “nd” and “if” as “f”. However, this must be strictly limited to a few words, or we might not be able to tell “are” from “or”, or tell “is” from “as”. A possible solution is to write “are” as “r” but “or” still as “or”, and “is” as “s” but “as” still as “as”.

In the example above, Geoffrey added “e” before “meis” to form “emeis”. This is where I feel unsafe. From his explanation, I understand that “emeis” is 1st person pronoun, which, I guess, might be common enough. However, I think it is not a short one, and it might be possible to interpret “meis” into other words.

For now, Geoffrey did not add initial vowels to too many words, or maybe I am missing some other examples in other posts. As far as I know, I think it is still too early to say the arbitrary is too much.

1.2 Dropping initial vowels from long words.

That is almost impossible, for languages where initial vowels count a lot for meaning.

René gave a simple solution: to add a mark for there-is-a-vowel-here, so that initial becomes 'ntl. Yes, it is still hard to tell it from similar words like intel, but it is caused by English using ti digraph to render sh-sound. There would be no ambiguity if we have separate letters reading '-N-SH-L.

2 Combining Consonant Flexibility and Vowel Dropping.

You have a good point here. Small approximations might accumulate into large errors.

You are not allowed to view links. Register or Login to view. blurred the boundary between voiced and voiceless consonants, so that any reader could have very high fliexibility. At the same time, it has also dropped diphthongs that it could not express into monophthongs, and consonant clusters into single consonants. In sum, it combines consonant flexibility and some kind of vowel dropping, but still be able to record real Greek. I think it would be fine to have some extent of flexibility.

However, as the You are not allowed to view links. Register or Login to view., I would like to see the losing rule, or defecting rule, is simple and relatively fixed. That’s why I have placed a question mark about his adding of suffixes. I saw Geoffrey have had an answer, but I don’t have time to read it for now.

Now I have to go. I’ll come back and post my other concerns when I have time.
(10-04-2019, 08:03 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.
(10-04-2019, 01:03 AM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view.(In case anyone is curious: at this rate, working 8-10 hours every day, 5 days a week, I estimate that it would take me about 20 years of concentrated effort to read and interpret the entire Voynich MS text using my method at the pace that I am able to do so currently.

If you're working on the "right" solution, it will speed up a lot. Because there must be a system and you will discover it.

If you're working on the "wrong" solution, you will stall. You need to keep adding more and more freedom.
In a recent mail I just saw yet another, new, possible meaning of the Voynich character d , where it could represent Greek 'ai'.

I believe you mean to say Greek "ia" here, and I believe you are referring to the Greek word "diaktos", meaning "carried through pipes". This is a rare word, and its meaning is closely related to the Latin root of "ductus", and I believe the author was writing "duct'-" in this part of this word, which comes out as Voynich [sdqs-].

There will always be irregular individual cases like this in any MS text. In some MS texts it could be much worse. Imagine if someone were trying to decipher an English text and had to explain the spelling and phonology of a word such as "women"! When the person reading and interpreting the text first tried to explain how the standard form of the word is written with the letter "o" but pronounced as /i/, it would be difficult to explain and to understand, would it not? The irregularities in my readings and interpretations are quite minor by comparison.

(10-04-2019, 08:03 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.Until now, you have a minimal amount of plain text, and the relation of this plain text with the Voynich text is extremely thin.

For an example of a 'translation' with fewer degrees of freedom than yours, with a longer plain text, that seems to be somewhat on topic, see the paper "The cannabis page of the Voynich Manuscript".
Link: You are not allowed to view links. Register or Login to view.

For me, both are equally wrong, but if you think that you are right and he is wrong, you have to be able to point to a clear reason why his is wrong and you are not.

I am glad Rene brought up this example, because it just so happens that I have researched the work of this researcher, J. Michael Herrmann, who claims that the Voynich MS text is written in Book Pahlavi. Now first of all, Book Pahlavi is a fascinatingly ambiguous script, and it is well worth studying its details for anyone who wishes to comprehend how an extremely ambiguous script could possibly work in practice. Studying it, you will find it hard to believe that anyone could ever read it, but yet it was the main script for Zoroastrian texts in Middle Persian for over a millennium! My system for Judaeo-Greek in Voynich characters is a model of logical clarity by comparison with Book Pahlavi, which is both ambiguous and a total mess in terms of groups of sounds represented by single letters, multiple letters, ligatures, etc., etc.

However, Herrmann's proposed theory of the Voynich MS as Book Pahlavi has all kinds of fundamental problems with it. I wrote up a summary of some of the problems in my notes a while ago, based on another paper of his, "The Voynich Manuscript is Written in Natural Language: The Pahlavi Hypothesis".
Link: You are not allowed to view links. Register or Login to view.

Here are my notes on the fundamental problems with Herrmann's hypothesis and paper:

The Voynich-Pahlavi character table is fundamentally flawed in many ways, as I can tell very quickly based on all of my experience studying this ms. His main character for Pahlavi "k" is Voynich [f] and for Pahlavi "p" is Voynich [p]. I do not see any explanation for how this can be reconciled with the restriction of these characters to first lines of paragraphs. His main character for Pahlavi "t" is Voynich [l], which is strange since "t" should be frequent in initial position while [l] is much more frequent in final position. His main character for Pahlavi "n" is Voynich [d], which is strange for the opposite reason. Neither makes sense. To compensate for these and other problems, he includes two different variants of Pahlavi scripts to make the letters even more ambiguous than Book Pahlavi is. So Voynich [t], which is Pahlavi "m", can also be a Pahlavi "t", and Voynich [ch], which is Pahlavi "š", can also be a Pahlavi "t" as well, among other things. Even Book Pahlavi isn't that bad by itself.

Yes, there are ambiguous values in my table, but they are phonologically logical and consistent categories such as labial, dental, guttural, liquid, nasal, sibilant. Even for my most complicated character, Voynich [d], there is a logical connection from /u/ to /v/ to /b/. There is nothing in my table remotely close to Herrmann's illogical groups of letters, such as Voynich [t] representing Pahlavi "m" and "t".

And then when he lists all the Zodiac signs and the Voynich and Pahlavi words, the correspondences don't even match with his own problematic character table. He includes photos of the Voynich words, but he doesn't list his own Voynich transcriptions for them that he gives in his character table. Probably that's because they don't match up to the Pahlavi words based on his table. His Voynich [t] as Pahlavi "m"/"t" character also gets matched with a Pahlavi "k" in the Zodiac chart, but even he didn't include that option in his character table.

Finally, here is his effort to translate part of the 3rd paragraph of the 1st page of the ms. I present only the final English translation of all the words put together, exactly as he presents it in his paper:

"nature flower/fragrant herb pass by sprout blood family Spenta Armaiti capable sensitive guide separation/distribute me+worthy name/cause two/evil being beginning sprout (?) Spenta Armaiti (?)"

Please do not try to tell me that any of this in any way even remotely resembles my own theory and my own readings and interpretations of the text.

Geoffrey
(10-04-2019, 09:03 AM)ChenZheChina Wrote: You are not allowed to view links. Register or Login to view.
(09-04-2019, 09:36 AM)Koen G Wrote: You are not allowed to view links. Register or Login to view.Zhe: all things individually might be possible, but I have some reservations.

1) When you turn English into abjad spelling (I'd really prefer to call it "dropping vowels" since abjads are something very specific), you notice that there are certain things you cannot do. For example, it's often very hard to drop initial vowels. The word "initial" would become "ntl", which is hard to revert properly. However, in his decipherment Geoffrey does take such liberties with adding vowels.
2) In isolation every aspect of Geoffrey's method works to some extent. But what if you combine them, especially consonant flexibility and vowel dropping? 

I'm not completely certain yet, but for now my impression is that it would be a one-way cipher. Flexibility is so great that the initial meaning is not recoverable.

Hi, Koen.

First, I have to say that I am sorry because I had to leave in a hurry yesterday.

I also have concern about initial-vowel-dropping, but I think it could be split into two smaller questions.

1.1 Dropping initial vowels from short, common words.

I think it is mostly fine to drop initial vowels from shortest and most common pronouns and prepositions. For example, I think it could be probably fine if we write “am” as “m”, “and” as “nd” and “if” as “f”. However, this must be strictly limited to a few words, or we might not be able to tell “are” from “or”, or tell “is” from “as”. A possible solution is to write “are” as “r” but “or” still as “or”, and “is” as “s” but “as” still as “as”.

In the example above, Geoffrey added “e” before “meis” to form “emeis”. This is where I feel unsafe. From his explanation, I understand that “emeis” is 1st person pronoun, which, I guess, might be common enough. However, I think it is not a short one, and it might be possible to interpret “meis” into other words.

For now, Geoffrey did not add initial vowels to too many words, or maybe I am missing some other examples in other posts. As far as I know, I think it is still too early to say the arbitrary is too much.

1.2 Dropping initial vowels from long words.

That is almost impossible, for languages where initial vowels count a lot for meaning.

René gave a simple solution: to add a mark for there-is-a-vowel-here, so that initial becomes 'ntl. Yes, it is still hard to tell it from similar words like intel, but it is caused by English using ti digraph to render sh-sound. There would be no ambiguity if we have separate letters reading '-N-SH-L.

2 Combining Consonant Flexibility and Vowel Dropping.

You have a good point here. Small approximations might accumulate into large errors.

You are not allowed to view links. Register or Login to view. blurred the boundary between voiced and voiceless consonants, so that any reader could have very high fliexibility. At the same time, it has also dropped diphthongs that it could not express into monophthongs, and consonant clusters into single consonants. In sum, it combines consonant flexibility and some kind of vowel dropping, but still be able to record real Greek. I think it would be fine to have some extent of flexibility.

However, as the You are not allowed to view links. Register or Login to view., I would like to see the losing rule, or defecting rule, is simple and relatively fixed. That’s why I have placed a question mark about his adding of suffixes. I saw Geoffrey have had an answer, but I don’t have time to read it for now.

Now I have to go. I’ll come back and post my other concerns when I have time.

Chen Zhe: Thank you again for the detailed comments and observations. I appreciate the clarity of your comments very much.

I can summarize my reply about the deletion rule, or my adding of suffixes, in a very concise way: In very many cases, such deletion of suffixes can be explained completely naturally in Greek as elision of a final vowel before a following initial vowel. I cannot claim that this rule will explain all of the cases, but it will explain a significant number of the cases, and it does explain two of the cases in the line you cited, and it may explain the third case as well if the late medieval Greek dialect had already dropped the "-n" in the "-omen" suffix, which "-n" has indeed been dropped entirely in modern Greek.

So we may begin with this as our first deletion rule, and then we can examine the more limited set of other cases to identify other possible deletion rules.

Geoffrey
So far I am finding Geoffrey's theory to be quite plausible and at least he shows his work in some detail.

However, he has only been at it a few weeks and it seems quite bizarre that he already has to defend his thesis in detail against all comers - a task which he is managing to do very well.

This task will obviously slow down further development of the thesis , of course.
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23