+- The Voynich Ninja (https://www.voynich.ninja)
+-- Forum: Voynich Research (https://www.voynich.ninja/forum-27.html)
+--- Forum: Analysis of the text (https://www.voynich.ninja/forum-41.html)
+--- Thread: [split] An interview with Stephen Bax - 1 to 1 substitution discussion (/thread-2100.html)
RE: An interview with Stephen Bax - Emma May Smith - 17-09-2017
(17-09-2017, 12:25 PM)-JKP- Wrote: You are not allowed to view links. Register or Login to view.How many natural languages have restrictions as severe as these as to a character's position in a word? I've done a very extensive search of languages trying to find something that is even half as restrictive and have not, so far.
Lots.
English, for example, only allows /ng/ in the syllable coda and it must come immediately after the vowel. Also, /h/ is only allowed in the syllable onset and only in combination with a semivowel.
And English is nowhere near the most restrictive language on Earth. Indeed, it is relatively liberal.
RE: [split] An interview with Stephen Bax - 1 to 1 substitution discussion - davidjackson - 17-09-2017
I'd point out that you're assuming this is a natural script.
That implies that the script has "evolved" alongside the underlying language to cover all eventualities.
Let us assume for a moment it's not, as Bax has postulated - it could be a script devised to encode an existing language for a specific purpose. In which case, the phonomes would be limited to those the creators assumed existed.
A real life example - 16th century aljamised Spanish (Castillian written using the Arabic script by the conquered Moors of early christian Spain).
The Spanish was turned into a partial abjad - the first writers didn't cover vowels, but then discovered you couldn't reinterpret the Spanish without some vital ones. So they continued to drop most vowels except for important ones, which were indicated by modified consonants or diacritics. O and U shared the same diacritic.
You ended up with 28 Arabic characters (plus four vowel diacritics) encoding all the sounds of Spanish. But, and this is interesting, many of the sounds were encoded using different letters depending upon their positional function in spoken Spanish.
Which is exactly what Bax is arguing. I'm not supporting his theory here, simply pointing out a real life example.
Here's the list, pinched off a You are not allowed to view links. Register or Login to view.page:
[spoiler]
Letra árabe
Transcripción
Consonantes
ا
Ø, h
ب
b, v, p
ت
t
ث
t
ج
ch, j, g
ح
h
خ
ch
د
d
ذ
d
ر
r
ز
z
س
s, ç
ش
x, s
ص
s
ض
d
ط
t
ظ
d
ع
h, Ø
غ
g
ف
f
ق
q, qu, c, k
ك
q, qu, c, k
ل
l
م
m
ن
n
ه
h
و
gu, gü, hu
ي
y
Vocales
َ
a
ا َ
e
ِ
i
ُ
u, o
Digram "ch" (which was important enough to warrant its own letter in the old Spanish alphabet) is encoded twice, once as a stand-alone sound and once grouped together with /ch/,/j/,/g/. This, I assume, is simply because of its function in spoken Spanish - it can be pronounced three ways, depending on how it's forming the word.
We see the same effect with other sounds such as /d/, /s/ which are assigned to different letters.
Taking it the other way, we see the sounds /gu/, /gü/, /hu/ all lumped together to one Arabic letter.
This doesn't prove anything with Voynichese, of course. But it is an example of how people developed new scripts for existing languages.
RE: [split] An interview with Stephen Bax - 1 to 1 substitution discussion - Davidsch - 17-09-2017
JKP
Regarding the 1 to 1 substitution discussion, when I started 3 years back I wanted to make sure that it wasn't a 1-1.
The only way to show that scientifically is compare it with all language groups known in Europe. It took me 2 years but that's finished.
All possible phonetically, text alterations, backwards, shuffled, no vowels, no doubles, partial substitutes, abbreviations etc. has been covered by my research.
Although, if you insist and want to study & compare phonetic transliterations, it is a good idea to start with the alphabet in that language. Then study some on the grammar and the inflections, and then research and study the pronunciations in medieval times, invent your own way in transcribing that consistently, write some software for comparing that, and finally start with comparing your results with the Voynich text.
It is not important how I worked but there's just no match, so I don't bother to lay down my specifications and results, the methods etc in a paper or a-like, because I'm not academical in this area and really, in the end we only care about the results. This year I studied basically (medieval) ciphers and mystical (wrongly valued) science. Now I believe the Voynich text is a cipher. There are some things on the text that point at that direction, but I see no motivation for spending time in defending or discussing my own research results here, until convinced otherwise. It is very possible that finally the text is a mix of code+cipher+substitution of abbreviation, such as is usual around the time 1400-1600.
(writing this small posting took me 35 minutes)
RE: [split] An interview with Stephen Bax - 1 to 1 substitution discussion - farmerjohn - 17-09-2017
(17-09-2017, 01:39 PM)Davidsch Wrote: You are not allowed to view links. Register or Login to view.Although, if you insist and want to study & compare phonetic transliterations, it is a good idea to start with the alphabet in that language. Then study some on the grammar and the inflections, and then research and study the pronunciations in medieval times, invent your own way in transcribing that consistently, write some software for comparing that, and finally start with comparing your results with the Voynich text.That's the only meaningful strategy, I think.
(17-09-2017, 01:39 PM)Davidsch Wrote: You are not allowed to view links. Register or Login to view.All possible phonetically, text alterations, backwards, shuffled, no vowels, no doubles, partial substitutes, abbreviations etc. has been covered by my research.Why have you rejected medieval Latin as a candidate?
RE: [split] An interview with Stephen Bax - 1 to 1 substitution discussion - Anton - 17-09-2017
I find the very way this question is put counter-productive. What's the reason to try to enthusiastically persuade people in what VMS is not? This reminds me something of missionary activities. Let's make it clear - until we know for certain (which we don't), it's the matter of probability. Some consider it probable, others (me included) do not. There are known arguments, and they have been extensively discussed, so one cannot say that a more or less erudite Voynich researcher just does not know them. If these arguments are known to all "Voynich factions", then what for is this trying to convert? Why not just let those researchers who stick to the "cipher faction" pursue their own research, those who stick to the "natural language faction" - their own research, and so on? And then share the results. This spirit of freedom in research seems only natural to me. The more research directions, the better, because the VMS is then attacked from different avenues. The only reason for to limit this freedom in any way is to effectuate research which has been funded - but this is clearly not the case here. Noone of us does fund Voynich research of others.
RE: [split] An interview with Stephen Bax - 1 to 1 substitution discussion - Koen G - 17-09-2017
Well said, Anton. I for one am glad that qualified people are researching all kinds of avenues I find unlikely myself. Like that I don't have to and I can focus on the stuff I do prefer to research
RE: [split] An interview with Stephen Bax - 1 to 1 substitution discussion - -JKP- - 17-09-2017
So I retract my wording.
I was tired and surprised to be getting nonanswers to genuine questions and thus worded the question about 1-to1 substitution codes the way I did out of sheer frustration.
I am, however, interested in the answer to this question. I will post it as a poll instead.
RE: An interview with Stephen Bax - Torsten - 17-09-2017
(17-09-2017, 11:03 AM)Stephen.Bax Wrote: You are not allowed to view links. Register or Login to view.I love the words 'almost' and 'with rare exceptions' in these examples. That means immediately that these are NOT 'severe restrictions'! This for me is an example of overanalysing and ending up in a self-imposed dead end.
There are perfectly enough symbols in the Voynich script to accommodate two sounds in the region of /r/ , and a terminal symbol (EVA m) as a counterpart to the main symbol (EVA r). But we are so far from having a full sound/symbol correspondence list that at this stage the discussion is a waste of breath. It doesn't move us forward in any way, so personally I intend to waste no more time on it, and focus instead on continuing to try to decode the script.
Sorry, but you can't argue that you are interested in 100 % rules only. It is a fact that the order of letters in the VMS is highly predictable. For this reason the second order entropy for the VMS is lower then in other European languages.
For instance the [q]-glyph is in 99,5% the first glyph in a glyph group and is followed in 97 % by an [o]-glyph. The [d]-glyph is followed in 53 % by [y], in 31 % by [a] and only in 4 % by [o]. ... It is reasonable that the text of the VMS contains some errors and that the transcription of the VMS contains errors. Therefore it would makes sense to ask if a [d] followed by [o] is maybe an error.
Even if the order of letters in words is highly predictable the order of words in the text is not. The text of the VMS contains very few repeated word bigrams and trigrams. In a text using human language grammatical relations should exist between words and these relations should result in words used together multiple times. If you assume that the text of the VMS is representing natural language the lack of [font=Arial]repeated word bigrams and trigrams should at least surprise you. Also Curriers observation that the text in herbal A is statistical different from the text in herbal B should surprise you. Why the text of the VMS behaves so different from text containing natural language?[/font]
There are different thinkable explanations for the text of the VMS:
1) the text represents an unknown natural language
2) the text represents an invented language
3) the text is some kind of cipher text
4) the text is a meaningless pseudo text
Which explanation is the correct one? If you are just guessing you have a 25 % chance that your choice is the correct one. Therefore it would be much more convincing if you would explain your decision.
RE: [split] An interview with Stephen Bax - 1 to 1 substitution discussion - Anton - 17-09-2017
Quote:For this reason the second order entropy for the VMS is lower then in other European languages.
To be strict, to this one should get a reply that the 2nd order entropy for the VMS is not actually the 2nd order entropy
. In the first place - because we don't know the actual alphabet behind the observed script. I did not succeed, nonetheless, in "normalizing" h2 by way of decomposing or aggregating EVA symbols (I did not do plenty of checks, just a few most evidently suggestive). The other way, nonetheless, is the possible "expansion" of glyphs (say, such as o) into several underlay letters (say, Latin ab). This is something that is virtually impossible to machine-check due to the huge amount of possibilities. But that is something which moves us from the plain language to cipher.
RE: [split] An interview with Stephen Bax - 1 to 1 substitution discussion - davidjackson - 17-09-2017
Quote:For instance the [q]-glyph is in 99,5% the first glyph in a glyph group and is followed in 97 % by an [o]-glyph.
OK, the first part of the statement doesn't make sense because 99,5% of words in the corpus aren't [q] AFAIK. And if it is followed 97% of the times by [o] that's no different from the Romance language rule of Q[u].