The Voynich Ninja

(12-12-2022, 04:10 PM)nablator Wrote: You are not allowed to view links. Register or Login to view.If you had a consistent way of converting Voynichese to Greek it would help a lot in this discussion, even if it is a long list of either do this or that (several possibilities).

At the moment I don't claim to be able to read the whole text, but I always explain what value I give to the glyphs, as in my previous message.

(12-12-2022, 04:10 PM)nablator Wrote: You are not allowed to view links. Register or Login to view.I have tried several times to figure out how you do it from the words on your blog and here and failed.

What word did you fail to read?

(12-12-2022, 05:30 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.The points others have raised about entropy and so forth are relevant, but let's try a different approach.
According to You are not allowed to view links. Register or Login to view., the twelve most common words in a large sample corpus of Greek texts -- together with their token counts in it -- are:
καὶ ["and"] 4129066
δὲ ["but"] 1501550 
τὸ ["the"] 1414996
τοῦ ["the"] 1140938
τῶν ["the"] 1051317
τὴν ["the"] 993011
τῆς ["the"] 849596
ὁ ["the"] 831492
ἐν ["in"] 795289
γὰρ ["because"] 687117
τὸν ["the"] 679309
τὰ ["the"] 627063

Unfortunately or fortunately for us, we cannot ignore the fact that the entropy of our text is lower than that of a "normal" text.  I am not a mathematician, I just understand that the degree of disorder of our text is lower, and that it would be unwise for us to compare without understanding what we are comparing with.
For example, the twelve most frequent Greek words, which you quote, in my opinion, provide very little information and in a very ordered text, such as a collection of recipes, for example, would perhaps be in the minority.
Besides, δ' can probably be transcribed as q' and in that case the word frequencies would have to be redone

(12-12-2022, 05:30 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.Meanwhile, we could also compare the results we get by applying particular forms of analysis to known negative cases -- that is, cases in which we know that a script does not match a given language.  So, for instance, we might test the hypothesis that modern Slovenian is a system for writing modern Turkish.  What happens if we look at the vocabulary of Slovenian and see if we can find matching Turkish words?
Slovenian [biti] = Turkish [biti], "his louse"
...
Slovenian [ne] = Turkish [ne], "what"
... and so forth

I don't quite understand your example, are you trying to show that languages belonging to different families can have homophones?

(12-12-2022, 05:30 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.This seems to be working pretty well.  So should we conclude that we're probably on the right track?  
If not, why not?

If not, why not? 
That was my question, I hope you will help me to answer it?

(12-12-2022, 05:49 PM)Ruby Novacna Wrote: You are not allowed to view links. Register or Login to view.What word did you fail to read?

Lots. I don't remember.

Maybe the problem is that we have to guess the rules or read hundreds of posts (if they are all written in your blog posts, I don't know). Some rules may apply only in some contexts. Some rules may take precedence over some other rules. How are we to know if they are not all written down in a single place?

For example two rules found in a random post: You are not allowed to view links. Register or Login to view.
As a "next step" (?) replace:
r/ρ with s/ς
and 
η with αι

So it seems that y not only maps to η or ε at the beginning or ος at the end, but also because of this next step it can be αι.
chedykar = γιτηναρ = γιταινας

That's from a 5 minutes browsing. Any more possibilities for EVA-y that I missed?

From You are not allowed to view links. Register or Login to view.
403. sheolkeedy (1) – skianu89 – σκιαινιδος,
423. soain (1) – soain – σοαιν
430. sokal (1) – sonai – ζωναι,
471. yteey (28) – 9bu9 – ηβαιος = βαιος – small, little, few, short
472. yteos (3) – 9beos – ηβαιος = βαιος – small, little, few, short

So:
y can be αι
ol can be αι
ai can be αι
al can be αι
ee can be αι
eo can be αι
any more that I missed... ?

P.S.
You are not allowed to view links. Register or Login to view.
ydaraishy – ηταιρηκως
a can be αι too?

(12-12-2022, 06:58 PM)Ruby Novacna Wrote: You are not allowed to view links. Register or Login to view.I don't quite understand your example, are you trying to show that languages belonging to different families can have homophones?

No -- I was trying to apply the decipherment technique you've been using to another set of data in which I know that the script doesn't match the language.

You've been able to match individual Voynichese words to individual Greek words, and I've been able to match individual Slovenian words to individual Turkish words.

What differences do you see between my strategy for matching Slovenian to Turkish and your strategy for matching Voynichese to Greek?

(12-12-2022, 06:58 PM)Ruby Novacna Wrote: You are not allowed to view links. Register or Login to view.For example, the twelve most frequent Greek words, which you quote, in my opinion, provide very little information and in a very ordered text, such as a collection of recipes, for example, would perhaps be in the minority.

Perhaps, if the whole text consists of stuff like "add three teaspoons oregano, one tablespoon coriander, then whisk vigorously until well mixed"; but even then, wouldn't the list of most common Currier B words ([chedy], [Shedy], [daiin], etc.) still provide an enormously important clue?  Maybe even more of a clue than it otherwise would?

(12-12-2022, 07:41 PM)nablator Wrote: You are not allowed to view links. Register or Login to view.Maybe the problem is that we have to guess the rules or read hundreds of posts (if they are all written in your blog posts, I don't know). Some rules may apply only in some contexts. Some rules may take precedence over some other rules. How are we to know if they are not all written down in a single place?

Indeed, to understand my approach you should read my messages, in which I share all my ideas.
I didn't start by establishing the rules of reading, except the rule of trying to save consonants.
I'm comparing the words in the manuscript with Greek words in the dictionary, learning a little grammar in the process, and I hope to be able to generalise at the end.
You can imagine, though, that glyphs transcribed as EVA "y" will necessarily mean several things, placed between words, at the beginning, in the middle and at the end of words?

About replacing ro final with sigma final, I think it works pretty well, it could be the rhotacism of the B language dialect, as in lakonian/tsakonian, for example?

(12-12-2022, 08:26 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.What differences do you see between my strategy for matching Slovenian to Turkish and your strategy for matching Voynichese to Greek?

The number of words

(12-12-2022, 08:26 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.wouldn't the list of most common Currier B words ([chedy], [Shedy], [daiin], etc.) still provide an enormously important clue?  Maybe even more of a clue than it otherwise would?

It is the words chedy and shedy that prove that it is not Greek?

(13-12-2022, 12:09 AM)Ruby Novacna Wrote: You are not allowed to view links. Register or Login to view.

(12-12-2022, 08:26 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.What differences do you see between my strategy for matching Slovenian to Turkish and your strategy for matching Voynichese to Greek?

The number of words

Fair enough.  But there's a list You are not allowed to view links. Register or Login to view. of what are supposedly the 2000 most common Slovenian words.  I just chose the top ten for my earlier example, but I suspect it wouldn't be difficult to find hundreds more "matches" with Turkish words, especially if I allowed more flexibility in what counts as a "match": for instance, if I speculated that Slovenian [b] could represent not just [b] in Turkish, but also [p] or [m].  The eleventh Slovenian word in the list would have been [njegovo]; that's a tricky one, but perhaps I'd try matching [nj] to Turkish [ş], and [g] to Turkish [h], and then start examining Turkish words that begin with [şeh-], such as [şehiri], "his city," and see where that took me.
 
As you've probably guessed, I didn't choose Slovenian and Turkish arbitrarily: those are two other languages in which people have identified hundreds (I believe) of individual "matches" with Voynichese words.  So:

Q. What are examples of languages for which hundreds of individual word matches with Voynichese can be identified, given a transliteration scheme that's flexible (i.e., you couldn't just substitute characters using a strict one-to-one mapping and reliably get a valid word) but is still consistent enough to satisfy whoever's doing the research?

A. Greek, Turkish, and Slovenian.

Q. Given that information, if all we know is that someone using a flexible transliteration scheme has identified hundreds of individual word matches between some language and Voynichese, what's the maximum probability of this discovery corresponding to a correct language identification? 

A. ~33%, if one of the three language identifications cited above is correct.

To be clear, this doesn't disprove your Greek hypothesis (or the Slovenian or Turkish ones).  But it suggests that finding even hundreds of word "matches" has a high likelihood of being a false positive.

This is relevant to your question, I think, because if there are some "clues" in favor of Greek and other "clues" against Greek, I assume we'd want to weigh the reliability of the different kinds of clue involved.

(13-12-2022, 12:09 AM)Ruby Novacna Wrote: You are not allowed to view links. Register or Login to view.

(12-12-2022, 08:26 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.wouldn't the list of most common Currier B words ([chedy], [Shedy], [daiin], etc.) still provide an enormously important clue?  Maybe even more of a clue than it otherwise would?

It is the words chedy and shedy that prove that it is not Greek?

"Prove" is a strong word -- you only asked about "clues."

If you were tentatively to identify the most common Voynichese words with Greek words for "the," "and," and "but," that could be somewhat persuasive.

If you were tentatively to identify them with Greek words for, say, "stir" and "mix" and "boil," that could still be thought-provoking.

If they don't seem to match any Greek words at all, or if you tentatively identify them with words for, say, "gimlet," "bog," and "dolphin," then perhaps there's a problem.

(13-12-2022, 03:44 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.But it suggests that finding even hundreds of word "matches" has a high likelihood of being a false positive.

Since there is a talk about Slovenian language, let me give you my opinion. First of all, do not trust Google and do not compare present day Slovenian with 15th century Slovenian. 
Because of high inflection and major spelling changes the Slovenian language is hard to translate with Google and hard to computer analysed.
Matching 100 words letter -to- letter, using any known or unknown VM transliteration or translation alphabet is not as easy as one would think, because many words are not found in dictionaries or recognized by Google.
Besides, the even the 100 exactly spelled words are not proof that the language is Slovenian, Turkish, Greek or any other. 
To look for matching words, the similarity of spelling is not always the best way. The changes have to be supported by rules that applied in the 15th century.
Finding approximate matches is a lot easier. If Slovenian language contains about 600 Sanskrit words, and Greek and Turkish contains them, the chances are that all three languages contain 100 similar words with similar meaning. After all, Greek was spoken in Anatolia, where Slavic language was a second language, and Greek was spoken in parts of Slovenia after the Roman Empire was divided. Greek and Latin was taught at schools up to the Second World War.
I believe the grammar is much better indicator of language, but this can also be problematic, because most Slavic languages have similar grammar, such as suffixes, prefixes, inflections, even vocabulary.  The distinction can only be made by the process of elimination of as many things that do not belong to a particular language that matches the VM. For example, the absence of G in the VM (noted also by dr. King), is not a proof that the language is Czeck.
The comparison cannot be made if different researchers use different transcription alphabet. Although EVA has a potential of being transcription alphabet, some improvements are needed.

(13-12-2022, 03:44 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.If you were tentatively to identify the most common Voynichese words with Greek words for "the," "and," and "but," that could be somewhat persuasive.
If you were tentatively to identify them with Greek words for, say, "stir" and "mix" and "boil," that could still be thought-provoking.
If they don't seem to match any Greek words at all, or if you tentatively identify them with words for, say, "gimlet," "bog," and "dolphin," then perhaps there's a problem.

In which case do you classify my proposals for identifying the most frequent words?

(13-12-2022, 09:54 PM)cvetkakocj@rogers.com Wrote: You are not allowed to view links. Register or Login to view.I believe the grammar is much better indicator of language

This is certainly true. All of us, whatever language we speak, know many Greek words, learned at school in maths, physics, chemistry etc; without necessarily realising it. I am aware that we need more elements to determine the language of our text