14-03-2019, 11:03 PM
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
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