The Voynich Ninja

ReneZ: I should say that I think we can be very confident that visually we have a "4" shape as the left hand part of the gallows characters with a top left hand loop, even if we cannot be sure of the meaning, if there is any meaning other than its usage.

I think the visual identification of the "4" shape should be amenable to rigourous demonstration.

Link to image:

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

(14-05-2019, 05:46 PM)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.Link to image:

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

Thank you!

I think my answer would be: it's a mix of the two. Or rather, a spectrum which runs from pointed left loop to rounded left loop.

This is a tricky issue, by no means reduceable to plain image recognition.

A shape might have been designed in one way, then systematically carelessly inscribed in another way - similar but not exactly 1:1. For the inventor of the alphabet this might have been one and the same glyph. For us these introduce the question of difference.

So one needs to look where and how the underlying design betrays itself. Image recognition will tell nothing about that.

(14-05-2019, 02:18 PM)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view....


It is the case that it is very common to have diplomatic cipher symbols which are made up of a number connected to a letter on its right like "o" or "p" (greek letter rho). It seems to me that this kind of symbol is not very common with the more standard alphabets we find with different natural languages which superficially have some similarities like the Glagolitic; that is why this identification of the "4" is important as it has implications for the source of the Voynich script, which obviously has implications for the source of the Voynich manuscript.

It is very common in languages that use Latin characters for a number to have an "o" shape connected to the right-hand side. Most of the time it means it is an ordinal number or an abbreviation (e.g., quarto, tertio femto)

Anton: I suppose the question I would pose then is if we can say on that basis in your opinion what any Voynich shape is

So do you think the "4" in the "4o" is really a "4"?

Do you think the "o" is really an "o"?

Do you think the "a" is really an "a"?

In short do you think we can say anything about how any individual Voynich glyph/symbol/character looks?

If I draw a picture of a triangle is it reasonable for someone to look at it and say that it is a picture of a triangle or can we say nothing about how any drawn image looks?

If we can't say what a symbol looks like how can we possibly say if two symbols are the same or different?

How can we justify the identification of any Voynich glyphs at all?

To me all these are doable and from a theoretical point of view I think it is worthwhile to refer to image recognition, that is of course not to say that there aren't other clues as well.

---

Anton: The question of difference is an interesting one as to how we measure the amount of natural drawing variation of the author.

If we can operate within a framework where in a sense we have a measure of natural drawing variation, so that we can tell if the difference between two Voynichese symbols is within the natural drawing variation of the author or not and thereby have a way of determining if they are two distinct symbols or the same symbol drawn slightly differently. I mean by this if we can formalise and make more rigourous the process of symbol identification.

It would be nice to move these questions from the subjective realm to the objective realm.

By all means come up with an algorithm that classifies them. As long as it's the software doing the work, there's no reason not to, and such an algorithm could be generalized to other letters.


But also keep in mind that some differences in shape might not be meaningful. Yes, there might be a relationship between the gallows and EVA-q, but the variations themselves might not matter, just as the difference between Latin/Cyrillic/Greek "o" written as a circle compared to "o" written as an oval is unimportant—they mean the same thing.

Hi Mark,

My views on the Voynich alphabet are more or less in line with the general "Curve-Line" idea expressed some years ago by Brian Cham, which we tried to develop further in this thread: You are not allowed to view links. Register or Login to view.  (which is yet all very raw and should be considered as work in progress).

In short, I believe that Voynichese glyphs are constructed by way of combining elementary shapes, mostly in the "base shape plus tail modifier" fashion. And next to that we see that more complex glyphs, such as benched gallows, are essentially combinations of less complex glyphs.

Whether this visual combination is merely a way to construct (= to invent) symbols of the (invented) alphabet, or it goes farther and this graphical superposition is a mapping of underlying actual superposition - this is an open question.

In English, "d" looks like superposition of "c" and "l", but that does not mean that "d" equals "c AND l". On the other hand, I recall one historical numeric cipher were graphical superposition did mean logical operation (arithmetic sum if I remember correctly).

The difference between a picture of a triangle (of a circle, of an apple, whatsoever) and a glyph is in that an apple or a triangle is an object of a material world or an abstract notion. A symbol of an alphabet is neither of those. It is a sign. Triangles depicted in a different fashion are different objects of the class "triangle". Letters depicted in a different fashion can be different letters or they can be the same letter just depicted differently. You can write a letter in various ways, this is the matter not only of personal handwriting but also of style. It will be one and the same letter none the less. Image recognition will not reveal the letter (the sign) behind the glyph. It is useful when we already know the alphabet, the exact full set of signs which comprises it. We do not know the Voynich alphabet, that's the biggest problem.

Quote:The question of difference is an interesting one as to how we measure the amount of natural drawing variation of the author.

With this I agree, and if results are tied with the context, that might reveal interesting trails and hints. The problem is with the methodology, as always. We have search tools, but we do not know what should we be searching for.

Anton: I think when you mention whether a "d" is a "c" plus "l" that is relevant. The symbol "4" is a unit which is probably not divisible into smaller known common symbols. In addition the symbol "4" is a known and common symbol of the time. I think therefore it is reasonable to ask if a gallows character is composed of a "4" symbol or not.