| Welcome, Guest |
You have to register before you can post on our site.
|
|
|
| Character entropy of Voynichese |
|
Posted by: Anton - 23-01-2016, 02:08 PM - Forum: Analysis of the text
- Replies (95)
|
 |
Comparatively low character entropy of Voynichese has traditionally been used as an argument against the natural language hypothesis. To make it clear, it is not the single argument against the natural language, but it seems to me that it is not the strongest one.
Why? Because I wonder how we can be sure of our calculation of Voynichese character entropy if we don't know the real Voynichese alphabet? We only work with transcriptions (such as EVA), which maybe (and I'm sure they are) not that adequate.
Take an English text and substitute all instances of "d" with "cl" (which visually is much the same, but not linguistically). In other words, exclude the letter "d" from the English alphabet and imagine that letters "c" and "l" now do all the work. I think that character entropy of English will then change (namely, decrease), will it not?
So if we decrease the level of decomposition in our transcriptions of Voynichese, entropy is likely to rise.
|
|
|
| Quire 13 |
|
Posted by: VViews - 22-01-2016, 06:51 PM - Forum: Imagery
- Replies (33)
|
 |
Hello everyone,
A few months ago I made a wordpress to share my views of Quire 13.
It's just one page, but here's a link in case anyone is interested:
You are not allowed to view links. Register or Login to view.
As I explain on that page, it may be that Glen Claston or Nick Pelling had come to some of the same conclusions before, but I never read any explanation of what their exact reordering was or the exact reasons they gave for it beyond the connected tubs on two of the folios.
I don't have the courage to trawl through old VMS list archives in search of their exchanges on the subject, if they're even on there. If someone has a link to them I'd love to read them.
If my observations have been made by others before, I'll be happy to edit and give credit as necessary, or even delete the whole thing if its entirely redundant.
I'm not really sure whether to post this here or in some other section. I put it here because my views are based on illustrations. Admins, feel free to move it, or tell me how to do it.
|
|
|
|
| Decomposition of the "gallows" characters |
|
Posted by: Anton - 20-01-2016, 03:05 PM - Forum: Analysis of the text
- Replies (50)
|
|
This is an interesting discussion in itself, so I decided to open a separate thread.
This thought occurred to me right after reading Cham's You are not allowed to view links. Register or Login to view. about the "Curve-Line system" which, among other things, makes accent on how some Voynich characters can be decomposed into more elementary characters - the basic e or i and one of the possible "tail modifiers" - an idea generally expressed earlier by You are not allowed to view links. Register or Login to view..
It appears that the gallows can well be decomposed in a similar way (please excuse awkward graphics, I'm not a professional web designer):
gallows_decomposed.jpg (Size: 39.24 KB / Downloads: 479)
All four plain gallows result from combining either EVA "q" or the vertical line with one of the two tail modifiers (marked as t1 and t2). Note that these tail modifiers, when combined with elementary e or i, mostly yield valid Voynichese characters:
- e + t1 = d
- e + t2 = g
- i + t1 = z (?)
- i + t2 = m
The only point of question here is the third expression. "z" is a valid character that is encountered, though rarely, in the VMS, but in contrast to i which is normally at a 45 deg. angle, z is perpendicular to the baseline. So it is not clear whether z is really meant to be a combination of i and t1.
A question may arise why p and t gallows, if indeed containing q as component, do lift over the text line and not extend below the baseline (as the normal q would). There is a ready answer - this is to make the "coverage" behaviour (explained You are not allowed to view links. Register or Login to view.) possible!
One also may argue that a standalone vertical is not met with in the VMS. Maybe it is not (I'm not sure), but I remember seeing vertical as a component in some rare characters.
|
|
|
|
| The Voynich Ninja goes public |
|
Posted by: Anton - 19-01-2016, 05:25 PM - Forum: News
- Replies (3)
|
|
So, pox leber, the day has come when the long-awaited Voynich Forum goes public.
Actually, the forum has already been publicly available for some time (as the current participants do know), and we have had some very interesting
duscussions and, of course, the "killer application" - the Voynich Blogosphere Reader (thx David!).
But until today we did not make any major announcements or the like. Today we feel that most things required for comfortable discussion (such as forum structure, policies and features) are in place, so we are glad to invite all people interested in the VMS to participate.
If you run a Voynich blog or a website, or simply participate in any Voynich-related community, it would be great if you could make a small announcement about this forum there.
The idea of this forum was not an idle one, but one that aimed to address the actual and often-expressed need of a dedicated Voynich-discussion site. So we very much hope that it may fill the gap. We tried to plan the forum structure so that it not only provides for discussions, but is task-oriented.
We would like to thank all those who have been sharing their feedback. The forum will continue to be polished from the features and probably design perspective, so please excuse any potential deficiencies that may still be in there.
A big thank you to all members of the initial focus group (you know who you are).
|
|
|
|
| The Impossibility of Double Gallows |
|
Posted by: Emma May Smith - 16-01-2016, 10:09 PM - Forum: Analysis of the text
- Replies (79)
|
 |
I've been playing around with some bigram tables lately, and though I haven't yet seen anything shockingly new, there is something reassuringly old which keeps pressing me to explain. Namely, that no two gallows characters can appear next to one another in the text. It likely seems as though such an obvious fact needs no explanation, but I feel it does.
By gallows characters I mean [k, t, f, p] and their bench versions [ckh, cth, cfh, cph]. Despite the 8x8 possible combinations for such characters, they basically do not exist in pairs. There are, as far as I can find, six exceptions in the whole text. Not six combinations, but six, individual, once occurring combinations of any two gallows characters.
Why is this important and why does it need explaining? Firstly, it is important because the way characters fit together should belie the linguistic facts underlying the text (assuming the text is linguistic, naturally). Next, because the statistics for gallows combinations are so stark, there being next to none.
Most combinations of characters appear a few times even if they do not normally go together. So [ak], a combination which should not ideally exist, gets about 40 hits; [oq] gets 20; [lm] gets about 10; and [en] 15. At these levels the stats are basically noise. They are likely no more than writing and reading errors, or missplit words. Even if they encode something genuine it cannot be a main part of the underlying language.
Yet, even with the possibility of errors, two gallows characters do not occur next to one another. Why? Well, here's my guess and what it means.
1) The gallows characters are distinct in essential form from all other characters. Many characters begin with a small round or straight stroke (such as what Cham's stroke theory is based on) and can be easily confused: [ei] for [a], [ch] for [ee], [r] for [s], among others. But gallows characters all begin with a long straight stroke above the line which only they use. Although the writer may have mistakenly written one gallows character when he meant another—and a reader likewise—they can only ever be confused for each other and never a non-gallows character.
2) Although gallows often come at the beginning of words they almost never come at the end. Even when a space between two words is ambiguous, the joining of two neighbouring words will not bring together two gallows. The reader cannot misread their way to double gallows.
3) The gallows all take the same place within the structure of a word. One loop or two, one leg or two, bench or no bench, no variation in their shape causes them to take a different place or makes it possible for them to occur together.
4) The structure of words is strict and variations simply don't occur all that often. This is a point made by researchers a long time ago but bears repeating. Characters fall into classes according to their distribution and role within words. They don't move about and do different things (I can only think of one possible exception to this). This is something fundamental to what they represent.
5) All the gallows must thus share some feature which puts them into the same class and makes them work in similar ways.
6) Their similar role and their similar appearance suggests that whomever invented the Voynich script did so with a clear understanding of not only how the underlying language worked, but also how languages in general work. The gallows itself as a character category is also a linguistic category.
7) Further, it is most likely a phonological category, which would explain why two sounds cannot appear together and why they must appear in certain places within words. Given the constraints on possible 'sound sets' within any language, and the number of distinct characters, the gallows as a feature can only represent a handful of phonological features.
|
|
|
|
| Currier A and B |
|
Posted by: Anton - 16-01-2016, 02:02 PM - Forum: Analysis of the text
- Replies (58)
|
|
The original Currier's explanation can be found You are not allowed to view links. Register or Login to view..
"Currier A" and "Currier B" "languages" have been distinguished from each other by statistically significant peculiarities in formation of sequences.
Of course, it was made clear from the beginning that Currier A and B are not languages in the common sense of the word, the term "language" is used only formally for brevity. We do not know what is the real underlay mechanism or phenomenon that leads to the differences observed.
The problem is that while A and B are different, they are not totally different. Namely, there are vords that occur both in Currier A and B. In the course of working upon my latest blog post, I made some initial screening on the "Voynich stars" occurrences in the botanical folios with respect to Currier A and B. This was not included in the paper and the examination was far from comprehensive, but off-hand I did not see any strong correlation between star occurrences and Currier language. E.g., "dayside" (f68r1) and "nightside" (f68r2) stars on general may occur both in Currier A and Currier B. A given star may occur both in Currier A and B. E.g. the most frequent star of "otol" occurs mainly in Currier A, but has two occurrences in Currier B (these stats are limited to plant folios only). "otor", if I am not mistaken, happens roughly equally between Currier A and B.
The question is whether there has been any investigation of vords with respect to Currier languages. Which vords do occur only in Currier A or only in Currier B? Which vords occur both in A and B? Do all labels with significant frequency count occur both in A and B? The latter question, as you understand, is aimed at establishing whether labels are Currier-invariant.
|
|
|
|
|
