(11-05-2025, 05:00 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.it would appear strange to use two different versions of the same character this way.
My hypothesis is that the alphabet is an invented one and that the authors intended that no-one should ever be able to read the manuscript. And because of this the authors were not obliged to have to follow any particular standard, and hence no obligation to use
m in place of
r on every occasion. At the end of a line they had a choice to use one or the other. I see that the authors had similar choices with other characters throughout the text. Earlier, in
You are not allowed to view links.
Register or
Login to view.
I gave my view that
iin and
in words were the same. The authors minds were sometimes in single stroke mode, and at other times in double stroke mode. Similarly with
r and
m. Sometimes they wanted to add the downswing, and at other times not to do so.
@oshfdk
If it's not too much trouble, can you add more colors to the chart in Post #50 for the categories 0.50 and above, 0.75 and above, and 0.90 and above?
Then is it possible to rearrange the glyph sequence based on highest scores?
(11-05-2025, 06:06 PM)dashstofsk Wrote: You are not allowed to view links. Register or Login to view.My hypothesis is that the alphabet is an invented one and that the authors intended that no-one should ever be able to read the manuscript. And because of this the authors were not obliged to have to follow any particular standard, and hence no obligation to use m in place of r on every occasion. At the end of a line they had a choice to use one or the other. I see that the authors had similar choices with other characters throughout the text. Earlier, in
You are not allowed to view links. Register or Login to view.
I gave my view that iin and in words were the same. The authors minds were sometimes in single stroke mode, and at other times in double stroke mode. Similarly with r and m. Sometimes they wanted to add the downswing, and at other times not to do so.
A good question then, why some symbols can have alternate loopy forms, like m replacing r, but some others not, like there is never a loop+downstroke form of n, as far as I know. I've seen at least once in the MS a tiny loop at the top of n, but I don't think I've seen a full downstroke.
[
attachment=10589]
I don't think these forms are just a whim, I think there is a system.
(11-05-2025, 07:07 PM)R. Sale Wrote: You are not allowed to view links. Register or Login to view.@oshfdk
If it's not too much trouble, can you add more colors to the chart in Post #50 for the categories 0.50 and above, 0.75 and above, and 0.90 and above?
Then is it possible to rearrange the glyph sequence based on highest scores?
I created this chart a while ago (~2023), and even if I find the code, I think it will be just easier for me now in our modern AI age to just create a new Javascript interactive version. Do you have any particular statistics you would like to see? I can compute the relative probabilities (that was my name for what dashstofsk calls the affinities, I agree that the affinities sounds better) and context cosine similarities.
(11-05-2025, 08:45 PM)nablator Wrote: You are not allowed to view links. Register or Login to view.Maybe they are badly written m or i+loop+downstroke, and the last image is a+loop+downstroke.
Gosh, I guess I was not paying attention. Are these usually transcribed as
m? Most of them look like a sloppily written m, but some of them like ainaiŋ look very deliberate and carefully written to me.
This actually leads to the question of whether
r and
n are the same?
Given that affinities exist, is there more info on the high affinity examples? Do examples occur in pairs?
(11-05-2025, 07:07 PM)R. Sale Wrote: You are not allowed to view links. Register or Login to view.If it's not too much trouble, can you add more colors to the chart in Post #50 for the categories 0.50 and above, 0.75 and above, and 0.90 and above?
Then is it possible to rearrange the glyph sequence based on highest scores?
I've asked Claude to create a Javascript implementation of the affinity/relative probability computation. You can enter any substring (EVA Latin) and hit Enter and it will produce a list of characters that precede/follow this substring with higher and lower observed count than what one would expect if characters were put randomly. The transliteration used is based on:
Code:
#=IVTFF Eva- 1.7
# ZL transliteration file, updated from EVMT project
# Version 1r of 11/04/2020
With some automated simplifications, for example, with uncertain characters the first option is selected, rare characters are replaced with ?, etc.
You can try it out here: You are not allowed to view links.
Register or
Login to view.
Note that the computation made by this script is not accurate for the characters already included in the search string, especially if the search string is very common. E.g., for a string like "or." the computed conditional probability of a preceding "r" (that would produce "ror.") will be lower than the correct value, since the algorithm doesn't account for r's already present in the search string when computing the total number of available r's.