The Voynich Ninja

A few different things I've read lately have reminded me that  in searching for similar-looking manuscripts we generallly ignore the fairly-obvious fact that Beinecke MS 408 doesn't present in the style of a formal book of the fifteenth century.

Re-reading d'Imperio's comment on the Friedmans' being refused a university grant to study this manuscript, two things were very noticeable - first, how little they knew about the manuscript fifty years ago, when Elegant Enigma was written and secondly - that the university scholars who dismissed the book as trivial were making a valid point.  It doesn't look like a formal medieval book at all.

Then, listening to Efraim Lev's discussion of the Cairo geniza documents, I noted that he constantly distinguishes between "a book" which is properly set out, all written in one language and so on, and other sorts of documents including hand-made notebooks, letter, medical prescriptions and many others.  The second group which are 'trivial' in that sense are very often written in a mixture of languages and scripts. 

This recalls that idea of 'multiple' and 'overlapping' types of text (which some see as multiple ciphers), as well as the Currier A and B which are usually called "languages" in quote marks.

As far as palaeography goes, and especially orthography, it is as well to remember that while the "real book" was produced in an environment where spelling was more-or-less standardized, other people wrote as they heard the words, and if they spoke dialect, that's the sound they wrote.

I should like to see more solid reasons for believing that the plant-pictures in the Vms are intended to form a 'herbal'; I should also like to see some effort made to look outside the narrow limits of European herbals of the formal sort, and I should also like to know whether the usual techniques for decrypting a cipher presume standard and consistent orthography.

This isn't my area, but I'm fascinated by the form of non-orthodox scripts, so I'll find a few and write a post. A short one.

A mixture of languages and scripts, has been tried by me. If you mix languages, the analysis of the text shows more and more a universal picture.
If you use a raw and pure language, the analysis show very distinct features. That is also very logical knowing that many European languages are related to each other: if you mix them, specific features will diminish.

Your second possibility of phonetic scripture was also examined by me. This results in even more distinct characteristics than a natural language. Although that would make the language somewhat more difficult to match with a known language, i will notice. 
The same for dialects, give me any dialect, substitute the letters also if you want, but i think i can easily see the language base. 

In the VMS language, the main problem lies in the fact that in a normal language, almost every letter that exists in the text, always occurs at least the same percentage on the first position of a word, as it does on another position. Second big problem is that the "flow" of the letters is not correct (in my language diagram watch the flow of the yellow bars). And last but not least, the letters on the last positions are not consistent with other languages.

Currier A and B are almost the same. Despite of some words that occur more in the other, the language construction is statistically exactly the same.

If we talk about a book, it would have: 

page numbers, ok check
logical paging, ok check
a story to tell, ok check
consistent writing, (left-to-right, paleo typeface) ok check



>> plant-pictures in the Vms are intended to form a 'herbal'
Well, i think Marco showed us many drawings, similar to those in the VMS in many threads. Or what do you mean?

What would be more interesting, yesterday i found only 1 herbal in the VMS that occurs twice.
Now i can try to compare the text and see if i can find similarities. My main question is: are there more "double herbals" inside the VMS?

>>..herbals of the formal sort, and I should also like to know whether the usual techniques for decrypting a cipher presume standard and consistent orthography.
I do not understand what you mean, could you elaborate?

I should have been more specific about d'Imperio's work.

The structure and direction of the research reported by, and expanded upon by d'Imperio was determined by William Friedman's two study-groups.  d'Imperio's book was published in 1978, when the National Security Agency printed it. Some of her references to Brumaugh and to Tiltman are dated in the 70s, but I take those as filling a couple of gaps.
Others may feel differently.
-----------------
Davidsch,

Part 1 of the promised post is up: You are not allowed to view links. Register or Login to view.

I think it includes answers to these questions.

Thanks for asking them.

D.

Diane, I started to read your blog until I reached this paragraph...

"About the Voynich manuscript’s written text, one assumption  is near-universal: that if it conceals a plain text, that plain text will conform to “book-standards”:  with its language-use, grammar and orthography (spelling) clear, consistent and so on."

I have to disagree. I see no evidence that this is a near-universal assumption. Most people I've seen who carefully studied and commented on the text have remarked that if it is plain-text it does NOT use clear or consistent or familiar grammar, etc., and many have suggested it might be synthetic text specifically BECAUSE it does not follow the usual natural-language patterns.

I'm sorry. I wanted to read the rest but this statement stopped me in my tracks because it says the opposite of the general impression I've gotten from most of those who have been examining the text and goes against my own impression of the VMS text, as well.

JKP
First, thank you for reading - even as far as you did. It's nice to know the work isn't going into a cone of silence.

That first comment was as an introduction; the detail followed.

I've made this a lot shorter than the original answer.

Yes, I wrote:

Quote:"About the Voynich manuscript’s written text, one assumption  is near-universal: that if it conceals a plain text, that plain text will conform to “book-standards”:  with its language-use, grammar and orthography (spelling) clear, consistent and so on."
You are not allowed to view links. Register or Login to view.

Perhaps I should have put the "IF" in bold.

I think it might have been Col.William Friedman who first made that observation in relation to the Voynich manuscript: namely that any cipher text, when deciphered, must yield one and only one possible reading, otherwise it is not useful as a ciphertext.

The "plain text" is whatever message was written that was later enciphered, and since no-one has read the Voynich "plain text" (supposing we do have a cipher text), so we don't know what characteristics it will have, only the characteristics we observe.

Now - those who still think it is a cipher text *have* to suppose that the underlying plain text will be consistent in spelling, and will conform to standard or "book-standard" levels of grammar and consistent spelling and so on.  That's the basis from which people derive their confidence in using statistical breakdowns by word-length, and by counting incidences of a given string such as "qokedy".

If you have a different sort of text altogether: say, one with no verbs, no definite or indefinite articles, and where someone might spell "water" that way in one paragraph, but "iater" or "wyter" or "orter" or "oooaata" in other places, then the cryptologists approach is stymied, pretty much.

The point of my post was that in running tests from an assumption that any plain text will conform to standard slabs of prose or poetry, researchers have always looked to the "book-" sort of writings, never to the other type called 'trivial' or 'minor' writings.

One exception to that rule has been Don Hoffman's approach to the text - which I refer my readers to in the same post.

But otherwise I have never found any case where the researchers did not simply presume that the Voynich text (underneath it all) would produce a 'normal' sort of grammar and orthography: that's why they test the written part of the text against things like the Declaration of Independence, and not against (say) a few pages from an accountant's diary or the work of a navigator, or a handbook of weaving instructions.

In Friedman's time, such texts were considered beneath the dignity of anyone who wanted to think of themselves as an important historian, or important person generally.  I also made that point in the post.

This has implications for study of the imagery too.  Hunting formal 'book-' type writings may be to look in just the wrong places for the precedents of imagery in the VmS.

With all due modesty, I think the point valid, and new, and worthwhile.  But that's just me.

I agree that the underlay plain text (if any) would not necessarily be of the "book-style"; indeed I have an impression that it is very probably a "Gold-Bug" (referring to E.A. Poe's story) or "telegram" style - omitting words that can be reasonably reconstructed after decipherment. In particular, this impression is based on the observation (best followed with the help of Job's visual tool) of high density of label occurrences in supposedly "descriptive" folios (such as herbal folios).

And, indeed, this circumstance needs to be taken into account when comparing stats with any literature narratives such as "Origin of the species" or the Bible.

Returning to paleography, however: even if the orthography (e.g. abbreviation principles) are not book-style-standardized, then would not we expect them to be consistent throughout the corpus nonetheless (at least within the same Currier language)?

Anton,
Interesting question, but not one I can answer. I guess I'm not sure about how the units are formed, and from what, so while I would expect the original maker(s) to have been consistent in essentials, I haven't a clue about how they would define the "essential-" or the "-consistent."

That's why I used the example of the Bobbio Missal. For the maker, the essential and consistent was no more than the alphabet and the meaning conveyed by the spoken text.  Since he didn't care whether the spoken text sounded like his own dialect, or like Latin, or Latin spoken with the accent of his local dialect, his orthography wasn't aimed at producing a "book-" text, but rather a series of sounds which, being spoken, conveyed meaning.

It might sound as if I'm talking semantics, but I'm doing my best to describe a cast of mind.