The Voynich Ninja

(03-04-2024, 05:55 PM)nablator Wrote: You are not allowed to view links. Register or Login to view.

(03-04-2024, 05:11 PM)merrimacga Wrote: You are not allowed to view links. Register or Login to view.Can we all agree that the VM is written fluidly by more than one hand?

Certainly not.

I wonder why anyone would describe the text as "fluidly written". If fluidly means written "You are not allowed to view links. Register or Login to view." then the VM's writing is the opposite, a mass of glaring irregularities: inconsistent spacing between words, so much that it is often impossible to decide where words are, with small unexpected gaps every few glyphs just because it is not weird enough (look at the first "word" fachys, it has 3 half-spaces), a wobbly baseline incomprehensibly changing direction often several time per word, often to avoid a gallows glyph on the next line, frequent unidentifiable glyphs on a continuum between r-s, o-a-y, etc.

While I certainly agree with all the points of @nablator, I also find that the script would certainly allow for fluent writing.
In that sense I agree with @merrimacga. I find this important enough to highlight, because in that respect it is quite different from all ciphers at the time. Most if not all cipher alphabets consist of individual characters that do not even allow fluid writing.

It is hard to see how this aspect was not part of the design of the writing / alphabet.

The points that @nablator makes become obvious as soon as one starts transliterating.
It is not equally obvious in all pages, but it is also not constrained to one hand or one set of illustrations.
(This is a subjective conclusion, and it will be worth coming up with some more quantitative result - that is something I have been working on quite intensively, but interrupted a while ago).

The (writing) baseline jumps were already addressed in a thread of its own.
I have seen several pages where a distinct vertical jump happens on several consecutive lines, at roughly the same horizontal position.
Plenty of cases (as nablator also says) where the writing on one line seems to be affected by the line below.

Another interesting example that just struck me a few days ago is f53r.
The first part of the first four lines form a nice compact block. All the rest is going in a different direction, with different spacing. As a whole, it is still more 'organised' than many other pages.

I have always considered the irregularities that @nablator mentions to be traits of easy, fluid, writing.

If the writing system is hard it would show up also as more defined and rigid row/page layout as one must devote more effort to get things right.
When it is an easy system, your lines and words can go anywhere because you are writing not just layouting; text just flows too easily around the figures and line/word spaces, directions seem to not pose any problem.

Of course the calligraphic skills of the writer contribute on this and at some points I doubt all of them were very skillful on writing books. I also believe that the writer(s) knew the "language" and were not simply and blindly transcoding something they did not uderstand, so the resulting text needed also to be relatively easy to read after writing, to make also corrections, and this is intriguing to me: the underlying language was known to the persons, the writing system was relatively easy and fluid to produce the VM but it was potentially also easy, when sufficiently experienced, to read the end product. How.

(04-04-2024, 08:01 AM)Scarecrow Wrote: You are not allowed to view links. Register or Login to view.I have always considered the irregularities that @nablator mentions to be traits of easy, fluid, writing.

It is clear that scribes were not skillful, some of them especially, there was also the need to break the flow to get ink every now and then... but that many irregularities is evidence, in the context of this thread, of uneasiness, of not being able to write more than a few glyphs (a "vord" at most) without breaking the flow, looking elsewhere. To me, this strongly suggests that:
- the text of the VM is not a copy of an identical written text, that would be much easier to write fluidly without difficulties,
- the functional (meaningful maybe) chunks of text are often smaller than a vord, the scribes often put a small gap (or vertical offset) consciously or unconsciously to separate them,
- the scribes didn't always know (or care) where spaces should be: spaces don't matter.

All this together is consistent with a verbose substitution cipher or similar, where the ciphertext is written one glyph or several glyphs at a time, requiring a break (not on the page) after each functional chunk to get the next from a source that cannot be memorized (therefore not a simple substitution cipher, something a bit more elaborate).

Quote:This strongly suggests to me that:
- the text of the VM is not a copy of an identical written text, that would be much easier to write fluidly without difficulties
- the functional (meaningful maybe) chunks of text are often smaller than a vord, the scribes often put a small gap (or vertical offset) consciously or unconsciously to separate them

I fully agree. 

Quote:- the scribes didn't always know (or care) where spaces should be: spaces don't matter.

I wouldn't draw any conclusions about spaces. Omissions or incosistencies using spaces can have many explanations. Ican just for get them or use them incorrectly, or not being well versed in writing, I might have problems using them already. One reason can still be that spaces do not matter but that would be quite exceptional imho.

Quote:All this together is consistent with a verbose substitution cipher or similar, where the ciphertext is written one glyph or several glyphs at a time, requiring a break (not on the page) after each functional chunk to get the next from a source that cannot be memorized.

Definitely the the writing system used looks like verbose cipher or dictionary, glyphs have dependency with each other. While glyphs can stand alone in "labelese" does not necessarily mean that they can be used alone but there must have been quite simple rules to find the correct bindings of glyphs to construct vords so for me the scribes needed to be able to memorize the source to be able to do the transformation, and then write it down with needed extras (gallows, benches).

I have many times played with the idea that the glyphs bind together to form parts-of or real words and then concatenated together, bigram-trigrams anagrammed. Like the famous qo could bind not-well-so but oq as so-not-well (and last never used as does not make sense). daiin could be d+a+i -> do-as-said and i+n=here, or d+a=something, iin->something else and combined make sense. I have no idea if anything like this would fit in any statistics, it is just my own idea.

(04-04-2024, 12:19 PM)Scarecrow Wrote: You are not allowed to view links. Register or Login to view.- the scribes didn't always know (or care) where spaces should be: spaces don't matter.

Spaces matters in a cipher based on n-grams. Lets say that the cipher contains the folloing n-grams: 8  a  r  ar  8ar . They have differentes sizes and they also share glyphs. Spaces would be the only way to distinguish them. 8ar  8ar would be equivalent to 8ar8ar but not the same that 8ar 8 ar and not the same that 8ar  8  a  r. 

(04-04-2024, 01:32 PM)Juan_Sali Wrote: You are not allowed to view links. Register or Login to view.Spaces matters in a cipher based on n-grams. Lets say that the cipher contains the folloing n-grams: 8  a  r  ar  8ar . They have differentes sizes and they also share glyphs. Spaces would be the only way to distinguish them. 8ar  8ar would be equivalent to 8ar8ar but not the same that 8ar 8 ar and not the same that 8ar  8  a  r.

They could matter, but then spacing would be far more consistent (no two transliterations agree). Reasonable conclusion: spaces don't matter, they are only there to mislead: vords are not words. Voynichese is a fake agglutinative language created by dropping a lot of spaces in a semi-systematic manner similar to what was commonly done in Latin manuscripts: spaces were frequently omitted, especially after short function words. MarcoP posted an example a few days ago: You are not allowed to view links. Register or Login to view.
2nd line: aquo = a quo, interra = in terra, incelo = in celo.

(04-04-2024, 05:16 PM)nablator Wrote: You are not allowed to view links. Register or Login to view.Reasonable conclusion: spaces don't matter, they are only there to mislead: vords are not words.

Possible, but in my opinion "spaces don't matter" and "they are only there to mislead" are still two different things. The second indicates an intention of which we cannot be sure it was there.

Spaces are largely consistent, but I agree that the scribes have been very inaccurate / sloppy with them.
The best we can hope is that they do not play a big role in finding the solution.
(They would play a major role for people placing a lot of trust in them, as this would then reduce their chances of success).

It has been observed that the spaces in the text are consistent with the label words, but this is not exactly proof that the spaces are meaningful. The label words could (as one hypothesis) simply be words picked from the text. But there are too many different options at all levels to pick one favourite method how all this could work. At least for me.

When looking for repeated strings of text (another anomaly in the Voynich MS) one will find cases of such strings with the spaces in different places. That is a thing that happens *much* less in known meaningful texts, so here is again an indicator that spaces are meaningless, or at least, mostly inaccurate.

As Nablator mentioned above, the Bonaventura manuscript You are not allowed to view links. Register or Login to view. shows irregular spacing. These are a couple of occurrences of "de cuius"

[attachment=8403]

For a phonetic script, spaces are convenient, but not vital. Of course, one can also think of many different cipher systems that do not depend on spaces.

Inconsistencies in spacing basically rule out some forms of nomenclator where ol, chedy, olchedy correspond to totally different and unrelated words (e.g. ol=rain, chedy=bow, olchedy=boat). This could be another argument against the possibility of a word-to-word mapping (word-to-word requires clear words, and this is not the case).

Similarly, verbose ciphers that heavily depend on spaces like that You are not allowed to view links. Register or Login to view. are ruled out. But verbose ciphers can be designed in such a way that space ambiguity is not a problem (e.g. by avoiding cipher sequencies sharing the same prefix).

(05-04-2024, 09:27 AM)MarcoP Wrote: You are not allowed to view links. Register or Login to view.Similarly, verbose ciphers that heavily depend on spaces like that You are not allowed to view links. Register or Login to view. are ruled out. But verbose ciphers can be designed in such a way that space ambiguity is not a problem (e.g. by avoiding cipher sequencies sharing the same prefix).

Yes. Spaces may be both meaningful, as separators between functional units, and useless because they can be reliably restored thanks to a rigid positioning rule (partial relation order), so it is possible that a lot of them were omitted to make Voynichese look more like a natural languages, with longer words: that could be intentionally misleading. On the other hand it is possible that scribes just didn't care, because they knew that the functional units were safe, they could be identified with or without spaces. Just like the spaces in "de cujus" and "in terra" could be omitted because decujus and interra are not Latin words and there is only one way to insert a space that produces two words, so there is no ambiguity.

Here's a quick attempt at a description of the constructed language Athanasius Kircher set forth in his You are not allowed to view links. Register or Login to view. (1663) -- a date which places it three years before he's thought to have been sent the Voynich Manuscript.

Kircher's goal was to create a universal written language that could easily be converted into other languages (his dictionary furnishes translations into Latin, Italian, French, Spanish, and German).  Although his system is structured much like a code, its purpose was to make writings more rather than less accessible.

His book presents two different schemes in some detail.

In the first scheme, each word begins with a Roman numeral that identifies its category:

I through XXIII: various words arranged in alphabetical order (in Latin)
XXIV: countries
XXV: cities
XXVI: time expressions
XXVII: personal names
XXVIII: adverbs
XXIX: prepositions
XXX: pronouns
XXXI: forms of "to be"
XXXII: forms of "to have"

Next comes an Arabic numeral that identifies a specific word within one of these categories.  For example:

II-1: news
II-2: to play at dice
II-3: white
II-4: to feed
II-5: friend, etc.

XXV-1: Amsterdam
XXV-2: Antwerp
XXV-3: Avignon
XXV-4: Augsburg
XXV-5: Orleans, etc.

XXX-1: from them
XXX-2: from me
XXX-3: from us
XXX-4: from you (sg)
XXX-5: from you (pl)
XXX-6: I
XXX-7: of them, etc.

XXXI-1: I am
XXXI-2: you (sg) are
XXXI-3: he [she, it] is
XXXI-4: we are
XXXI-5: you (pl) are
XXXI-6: they are
XXXI-7: I was
XXXI-8: you (sg) were, etc.

A third element is optional and is drawn from a set of symbols that featurally represent grammatical distinctions of case, number, voice, tense, and mood.

[attachment=8405]

Kircher also provided a second scheme -- intended for letter-writing -- in which each word again begins with a Roman numeral that identifies its category but ends with a capital letter corresponding to some frequently used expression.  Many of these expressions involve seemingly trivial distinctions among options for writing "your most humble and obedient servant" and so forth, but we also have:

XXIV-A = you
XXIV-B = your
XXIV-C = your parents
XXIV-D = your relatives
XXIV-E = your brothers
XXIV-F = your sisters, etc.

Offered as one concrete example of an early (but not fifteenth-century) constructed language in which words have a very rigid structure.