Hello everybody once again.
I have some ideas concerning VM and was advised to start separate thread. I'll try to describe them briefly.
First of all the map from Voynich to Latin (* stands for any vowel sequence):
o - v, y, b, u
e - c, t, g
y - com-, con-, -is, -us
d - l
ch -
e*r
l - m, n
k - d
r - r, -rum
n - (none)
t - l*l
[font=Eva]q[/font] - q, k
[font=Eva]sh[/font] - pr, fr
[font=Eva]s[/font] - s, x
p[font=Arial] [/font][font=Arial]- p, f[/font]
m - ris, rus
cth - l*l*r
[font=Eva][font=Eva]ckh[/font] - d*r[/font]
f[font=Arial] [/font][font=Arial]- s[/font]
[font=Eva][font=Eva]cfh[/font] [/font]- f*r
g[font=Arial] [/font][font=Arial]- lis, lus[/font]
x[font=Arial] [/font][font=Arial]- z[/font]
ai-family are vowels
Second hypothesis: words with gallow in front are capitalized, others are not. So,
a) if a word starts with gallow, but shouldn't be capitalized,
o is written in front (that's why texts near images often start with
o)
b) if a word starts with non-gallow, but should be capitalized, gallow is written in front.
Separate words undoubtedley can be translated with this key. For example, famous daiin appears to be form of Latin ille.
More specific details can be found in attachment.
Cheers
Omitting vowels would lead to the increase of the character enthropy, while this is the opposite to what is observed with Voynichese.
I read your work and wish to make four comments:
1) I like that you identify [i] sequences as potential vowel extensions. I've had a similar thought myself and am glad to hear others come to the same conclusion.
2) I would like to see a long piece of text in translation. I understand that your Latin may be below average, but I want to see if the words you propose cluster around a topic and have some kind of coherency.
3) Can you explain why you attribute certain values to certain characters? What leads you to believe that [k] is a /d/ sound, for example?
4) How do you deal with the fact that because [p, f] mostly occur in the first line of paragraphs, words containing /p, s, f/ must also occur mostly there?
Anton, if I recall formula for entropy correctly, when we decrease number of possibilities, entropy also decreases. In any case in we also have an effect of one Voynich symbols substitutes several latin letters.
Emma May Smith, thank you for reading the work
2) Short words are the problem. For others I can select more or less suitable meaning
3)
o is the most obvious one for me. v and b are undoubtedly related, u and y go with them. I guess (don't laugh) th
at [font=Eva]o comes from Greek omega and [/font]is mostly pronounced [w]. So
oror from last page may be something like [warwarum]. c is connected to g from one side, and to t from another. f-ph-p is yet another chain. If it's really Latin and
daiin is most frequent word then
d is obviously l.
r,
l appear among most frequent endings. Difficult to imagine
r to be other letter than r, so
l is obviously m/n. And yes, m and n go together in all abbreviations I've seen. For
k I had a choice between d and r and even dr. When d translations are more adequate. What I really like in this key is that it is absolutely natural and allow to find translation for most of words.
4) I have no theory about that. Is it more strange than "no gallow before
q" fact?
-JKP-, thanks for the information.
(16-09-2016, 07:12 PM)farmerjohn Wrote: You are not allowed to view links. Register or Login to view.3) o is the most obvious one for me. v and b are undoubtedly related, u and y go with them. I guess (don't laugh) that [font=Eva]o comes from Greek omega and [/font]is mostly pronounced [w]. So oror from last page may be something like [warwarum]. c is connected to g from one side, and to t from another. f-ph-p is yet another chain. If it's really Latin and daiin is most frequent word then d is obviously l. r, l appear among most frequent endings. Difficult to imagine r to be other letter than r, so l is obviously m/n. And yes, m and n go together in all abbreviations I've seen. For k I had a choice between d and r and even dr. When d translations are more adequate. What I really like in this key is that it is absolutely natural and allow to find translation for most of words.
I still don't know how you reached the value for each character.
It seems as though some values were proposed based on the shape of the Voynich character and its similarity to existing scripts: so [o] and [r] are based on what they look like in Greek and Latin respectively.
Otherwise they were based on the output of the text once the value was applied: [daiin] should be 'illi' because the language is Latin and you were seeking a frequent word.
Can you tell me how you discovered the text was in Latin?
(16-09-2016, 07:27 PM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view. (16-09-2016, 07:12 PM)farmerjohn Wrote: You are not allowed to view links. Register or Login to view.3) o is the most obvious one for me. v and b are undoubtedly related, u and y go with them. I guess (don't laugh) that [font=Eva]o comes from Greek omega and [/font]is mostly pronounced [w]. So oror from last page may be something like [warwarum]. c is connected to g from one side, and to t from another. f-ph-p is yet another chain. If it's really Latin and daiin is most frequent word then d is obviously l. r, l appear among most frequent endings. Difficult to imagine r to be other letter than r, so l is obviously m/n. And yes, m and n go together in all abbreviations I've seen. For k I had a choice between d and r and even dr. When d translations are more adequate. What I really like in this key is that it is absolutely natural and allow to find translation for most of words.
I still don't know how you reached the value for each character.
It seems as though some values were proposed based on the shape of the Voynich character and its similarity to existing scripts: so [o] and [r] are based on what they look like in Greek and Latin respectively.
Otherwise they were based on the output of the text once the value was applied: [daiin] should be 'illi' because the language is Latin and you were seeking a frequent word.
Can you tell me how you discovered the text was in Latin?
That was a long and painful way
To be serious, one day I asked myself what if
q is really q (both occur always* in the beginning of the word)? Then
o, which always* follows
q, is u(v) and language is probably Latin or English. As frequenly noted by
-JKP- there are a lot of abbreviations from medieval Latin and the author certainly knew Latin very well. So should be Latin. Ok. Great. But
o is u/v? Ha-ha.
Then another day I found word which would have acceptable meaning, if
o is u/v. That was striking for me. Later I wrote computer program that uses Whitaker's great dictionary and finds proper translations for Voynich word. And the key converged to its present state.
Of course amateur with no proper knowledge has a little chance, but when your key produces "petresillum" for plant from You are not allowed to view links.
Register or
Login to view. and "cunila" for plant from You are not allowed to view links.
Register or
Login to view. it gives you tiny hope
*almost
Quote:Anton, if I recall formula for entropy correctly, when we decrease number of possibilities, entropy also decreases.
Character entropy is mean information per character. When letters are omitted, ceteris paribus mean information per letter increases.
(16-09-2016, 07:57 PM)farmerjohn Wrote: You are not allowed to view links. Register or Login to view. (16-09-2016, 07:27 PM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view. (16-09-2016, 07:12 PM)farmerjohn Wrote: You are not allowed to view links. Register or Login to view.3) o is the most obvious one for me. v and b are undoubtedly related, u and y go with them. I guess (don't laugh) that [font=Eva]o comes from Greek omega and [/font]is mostly pronounced [w]. So oror from last page may be something like [warwarum]. c is connected to g from one side, and to t from another. f-ph-p is yet another chain. If it's really Latin and daiin is most frequent word then d is obviously l. r, l appear among most frequent endings. Difficult to imagine r to be other letter than r, so l is obviously m/n. And yes, m and n go together in all abbreviations I've seen. For k I had a choice between d and r and even dr. When d translations are more adequate. What I really like in this key is that it is absolutely natural and allow to find translation for most of words.
I still don't know how you reached the value for each character.
It seems as though some values were proposed based on the shape of the Voynich character and its similarity to existing scripts: so [o] and [r] are based on what they look like in Greek and Latin respectively.
Otherwise they were based on the output of the text once the value was applied: [daiin] should be 'illi' because the language is Latin and you were seeking a frequent word.
Can you tell me how you discovered the text was in Latin?
That was a long and painful way
To be serious, one day I asked myself what if q is really q (both occur always* in the beginning of the word)? Then o, which always* follows q, is u(v) and language is probably Latin or English. As frequenly noted by -JKP- there are a lot of abbreviations from medieval Latin and the author certainly knew Latin very well. So should be Latin. Ok. Great. But o is u/v? Ha-ha.
Then another day I found word which would have acceptable meaning, if o is u/v. That was striking for me. Later I wrote computer program that uses Whitaker's great dictionary and finds proper translations for Voynich word. And the key converged to its present state.
Of course amateur with no proper knowledge has a little chance, but when your key produces "petresillum" for plant from You are not allowed to view links. Register or Login to view. and "cunila" for plant from You are not allowed to view links. Register or Login to view. it gives you tiny hope
*almost
What do you get for the star labels on f68r3 ?
PS this is a very nice set of conjectures!
(16-09-2016, 07:57 PM)farmerjohn Wrote: You are not allowed to view links. Register or Login to view. (16-09-2016, 07:27 PM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view....
Can you tell me how you discovered the text was in Latin?
That was a long and painful way
To be serious, one day I asked myself what if q is really q (both occur always* in the beginning of the word)? Then o, which always* follows q, is u(v) and language is probably Latin or English. As frequenly noted by -JKP- there are a lot of abbreviations from medieval Latin and the author certainly knew Latin very well. So should be Latin. Ok. Great. But o is u/v? Ha-ha.
I assume you mean this is the logic you followed to get where you are now, but I want to make it clear for those who might not be regulars and haven't followed the history of my posts, that I was not the one who stated it "should be Latin" but rather
your reasoning that brought you to the place where you feel it should be Latin.
Where I stand on this is that many of the shapes are definitely Latin, without a doubt.
What language underlies the text could be many things (natural language, constructed language, carefully structured gibbish (but not random text)) but I have never asserted it to specifically be Latin. I am not discounting the possibility, but it's not my main line of research and I wouldn't want someone to misread your statement as originating from me.
Now, to get back to your idea that it should be Latin based on the preponderance of Latin characters, take a good look at Yulia May's analysis of the text. It is along the same line as you are investigating. You can find it on this thread:
You are not allowed to view links.
Register or
Login to view.