The Voynich Ninja

(31-10-2018, 05:09 PM)VViews Wrote: You are not allowed to view links. Register or Login to view.In some of the examples nablator points out, the character looks like a z. In others, it looks more like an m.
For example, the 69v one and the 70v outer ring one don't feature the descending long tail that m has, so I was inclined to think they could indeed be z. Later, on f70v, the other z that nablator points out is (to me) clearly an m.

Thanks for the review, I had these in my files from a previous (quick) pass, not yet finalized, so I have to revise these. Rare glyphs are not a pressing issue anyway because they don't matter in statistical analysis.

Quote:Interestingly, it appears that double m is also rare but possible; voynichese.com (and therefore takahashi) records two occurrences of double m:
daimm on 3r (that one I agree is definitely a double m) and damamm (which actually looks to me more like dazazm ) on 67v1.

On 3r it is clearly daimg IMHO, therefore not repeated.

So JKP are you saying the slant is more important than the presence/absence of the descending tail?
Sometimes there is a tail but not much or no slant. Sometimes there is a slight slant but no tail. 
I'm now completely confused about m and z.

Edit: nablator I forgot about the g character, but I guess now we can add it into the mix of characters that barely differ from m.

(31-10-2018, 05:38 PM)VViews Wrote: You are not allowed to view links. Register or Login to view.So JKP are you saying the slant is more important than the presence/absence of the descending tail?
Sometimes there is a tail but not much or no slant. Sometimes there is a slight slant but no tail. 
I'm now completely confused about m and z.

Edit: nablator I forgot about the g character, but I guess now we can add it into the mix of characters that barely differ from m.

It depends. In some cases, the slant doesn't matter at all. But here's the problem. Let's say you were adding the loop (which in Latin most often stands for "is" suffix) and you were adding it to an "r" or you were adding it to an "i", it might be difficult to tell which one it is supposed to be... so... some scribes would lean back the "i" (or alternately the "r") in order to distinguish them. And... in some cases, they didn't care. They wrote them both the same or both ways.

The long tail (beneath the loop) does matter. It signals that it's an abbreviation (rather than a ligature). The length of the tails varied a lot, but most scribes lengthened the tail when it was an abbreviation.


So... if you see the shape without the descender it's more likely to be a ligature. If you see it with a tail, it's more likely to be an abbreviation. The slant, well, that's scribe dependent but is sometimes an important distinction (and sometimes not).

(31-10-2018, 10:45 AM)VViews Wrote: You are not allowed to view links. Register or Login to view.Hi DONJCH,
according to voynichese.com, there are only two occurrences of z.
The double loop version apparently occurs only three times (four if we count the somewhat dubious one noted by ChenZheChina on f31r).
I'd say that definitely puts them in the "rare" or unusual category, or possibly in the "error" category.

I was assuming these were the half gallows mentioned here - was that wrong?
You are not allowed to view links. Register or Login to view.

Also, I guess in any attempt at taxonomy, there will be people who are "splitters" and others who are "lumpers".

This is a very important observation, and it plays a key role in transcribing.
Once certain characters are lumped together, in a transcription file, they cannot be split anymore, except by hand, and by going back to the scans.
If they have been split 'too much',, this can be fixed by automatic scripts.

I agree with JKP that complex gallows can be their combinations. As I see it, I wrote in the message # 1, 5. You are not allowed to view links. Register or Login to view. But, an easy way is possible to write KK. This is the writing of the double loop on the right. I already gave an example on f17v, but I found another case in the B language on f113v.
[attachment=2471]

How big the distance ( chars ) between any two gallows is can be graphically represented. Overall, the curve drops evenly. For the entire VMS it looks like this:


You are not allowed to view links. Register or Login to view.

(31-10-2018, 05:09 PM)VViews Wrote: You are not allowed to view links. Register or Login to view.

(31-10-2018, 01:22 PM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.This depends on which transcription was used in voynichese.com .
The fact that it has its own 'basic Eva' code means that there must be at least 10 occurrences.
This is still not a lot of course.

Thanks ReneZ, I didn't know 10 was the minimum requirement for the basic EVA characters.

In fact, I decided to do a count, and found only 7 of them in my latest transcription file.
What I do remember for certain is that, at the end of our transcription exercise, the three most frequent 'unusual characters' (i.e. anything that was not a Currier or FSG character, or a ligature of them) were assigned basic Eva codes.
These were b , u and z .

No other character appeared 10 times or more.
So my information was not fully correct.

(01-11-2018, 11:06 AM)bi3mw Wrote: You are not allowed to view links. Register or Login to view.How big the distance ( chars ) between any two gallows is can be graphically represented. Overall, the curve drops evenly. For the entire VMS it looks like this:

Of course it drops. I made the exact same observation 4 - 5  years ago, not only for characters, but also for words, and for adjacent characters etc.
It drops because the spread is evenly per word, and it reaches its top occ. at the average length. For words the average is 6 chars. 

This thread is about double gallows.  (I consider gallows : PFKT)  There are no double gallows, because the "formula" for gallows does not allow it, and these are mainly scribe errors, and sometimes something else is going on.

I've posted examples of VMS "combination" gallows on other threads, but it's appropriate here too. I think these two examples (and others like it) may be double gallows:

[attachment=2472]   [attachment=2473]

There are several reasons why I think these are double gallows, but there are three reasons in particular...

1) "Stacked" letters (vertical ligatures) do exist in scribal conventions. One sees stacking in early Latin texts, and in Greek texts, especially in numbers (I've blogged about this) but also letters. In fact, in Greek, stacking is not uncommon. It is a convention that goes way back to carved inscriptions. It saves space and looks cool. I've mentioned "prodromos" written with stacked letters in a previous post.

2) The letters that are stacked in the example are frequently stacked in Greek. If you look at the first example, the lower one looks similar to "pi" and the upper one looks similar to rho. Pi and rho are frequently stacked because they represented common numbers, which means they were stacked in both textual script and numeric script.

3) Benched characters also look like Greek stacked characters. In fact, the format is closer to the way they are usually stacked. Crossing an ascender with a pi character was a common way to stack letters.

Ligatures can look like this too. This is just one of many examples. It represents a ligature of n + g:

[attachment=2475]


I've been saying for a very long time that most of the VMS characters are based on Latin letters/abbrev./ligatures and some are based on Greek. The ones that are less common in Latin are frequently found in Greek. It's not only the shapes that are similar, it's the way they are combined. The IDEA of stacking and benching is similar, as well.

Here's an example of several letters stacked in Greek. The same concept is used in creating monograms in many languages:

[attachment=2474]


.
Classical literature was very popular in the 14th to 16th centuries and there were several revivals in teaching Greek. It was not uncommon to see Greek alphabets and numbers included in Latin texts. Scholars would have seen Greek shapes and conventions even if they didn't study the language, so the presence of Greek conventions might indicate knowledge of Greek or it might not. The majority of the VMS shapes are Latin, and many of the Greek conventions are also found in Latin text, so there's nothing specifically pointing to Greek, but there are signs of familiarity with Greek scribal conventions.