+- The Voynich Ninja (https://www.voynich.ninja)
+-- Forum: Voynich Research (https://www.voynich.ninja/forum-27.html)
+--- Forum: Analysis of the text (https://www.voynich.ninja/forum-41.html)
+--- Thread: how to deconstruct the VMS text and label words repeated twice or more (/thread-598.html)
how to deconstruct the VMS text and label words repeated twice or more - don of tallahassee - 25-05-2016
Hello,
I take this opportunity to reveal my latest work on deconstructing the Voynich Manuscript text words and label words.
It shows the results when my proposed system, code group tables and sequence are used for deconstructing the more than 2,000 words that appear twice or more in the VMS.
12 of the repeated VMS words do not deconstruct as I envision and show.
The results are at:
You are not allowed to view links. Register or Login to view.
Notice that many of the individual codes in Groups II through V! are often either the only representative(s) of those code.glyph(s) to be found in any of those groups, or, if duplicates or triplicates, often separated by one or more intervening groups (thus making them easier to identify). On close examination, the groups should be found to be very different from one another. Slightly more than a total of 400 codes are used to deconstruct the 2,000+ words.
The duplicates and triplicates were what made me have to figure out the sequence of deconstruction - without a consistent deconstruction sequence, duplicated codes in VMS words might be arbitrarily assigned (by the person doing the deconstruction) to the wrong group.
I think many of the remaining single-appearance words can also be deconstructed using the same system, code group tables and sequence.
Please also notice that no glyphs are added, changed, left out, moved, or exchanged for other glyphs - such changes seem to be unnecessary.
I encourage others to offer any reasoning they may be able to think of as to why almost all of the repeated VMS words seem to deconstruct in this manner. Or to check my work and results.
Thank you.
Don of Tallahassee
RE: how to deconstruct the VMS text and label words repeated twice or more - don of tallahassee - 26-05-2016
I'd like to make a few remarks about my proposed scheme to deconstruct the words found two or more times in the Voynich Manuscript.
The words are found throughout the manuscript. They comprise about 85% (rough estimate) of all the words in the manuscript - I've not been able to find any definitive number of exactly how many total words are in the VMS. The number varies from source to source. Nor have I added together all the times each of the duplicated words appears, to get a tally. Suffice it to say that a majority of the total words in the VMS deconstruct as I envision. Twelve exceptions out of more than 2,000 seems to show them to be exceptions, not a reason to chuck the whole scheme.
I don't need to change, move, eliminate, exchange or add any glyphs to make these perceived results happen. My main reason for this is that such seemingly questionable operations are not necessary, so why do them? There aren't any other examples of codes from the period that require these operations.
I use the words in the VMS that are repeated two or more times to help insure that few mistakes in transcribing the words are perpetuated because of ambiguities in the glyph identities.
The sequence of code groups (I, II, III, IV, V, VI) is followed in each example. The codes are found in the same places in each word, when present.
The sequence of deconstruction steps is followed for each word.
In all but the first group, the code groups are fairly small, and, within each word part (prefix - midfix - suffix), mostly exclusive. Yes, I know there are a few exceptions like a and o glyph characters, but not many within each word part.
I am not the first to think of such a scheme. Both Mrs. D'Imperio and Professor Stolfi beat me to it, and probably many others. And Roger Bacon.
I know many of the leading lights in the group trying to solve the puzzle of the VMS do not like that my scheme seems to show some way to construct and deconstruct the VMS words. For some of them, it would obviate years of work trying to find a solution using more modern techniques to account for an early Fifteenth Century manuscript. For others, that fact that the scheme cannot be made to fit any known language's words as a direct deciphering/transposition scheme of some sort or another is anathema. That's okay. They can think whatever they like. I'm only trying to convince those with open minds.
I don't think there are any known languages, actual or invented, that have words with the characteristics the VMS words show in construction/deconstruction.
The results seem to show a scheme with fairly simple relationships between the pieces that make it up. In actual use, it would be easy for anyone in on the secrets to be able to read/decode directly from the page without intervening steps or equipment. Only a key to the meanings of the code groups would be needed, maybe one that was memorized. 400 or 500 or 600 codes might not be too hard to memorize, especially if they might be coordinated with the code meanings in some way such as the codes being simple abbreviated reminders of the actual words being encoded. A code book or cheat sheet would make matters even simpler.
I don't know a lot about early Fifteenth Century codes. I don't think the VMS words directly correspond with any other known coded words of that date or before. Maybe I'm wrong about this - does anyone know of one?
I don't know of any other schemes by other people that seem to show a consistent way that the VMS words might be formed or constructed or any consistent way that the words might be deconstructed.
Yes, I know the elephant in the room with my scheme is the first group of codes being so large. I can't help that. The Group I codes seem to be in the words where I say they are. I didn't put them there. There is the possibility that they need to be decoded further, maybe into more groups or sub-groups. I doubt it.
The other code group tables seem to be more reasonable in size and membership. And they are consistently found in the VMS words where I say they should be found.
Check my work. I didn't fudge anything or leave out anything or change any words to make them fit my ideas (unless by accident or error or typo). I don't work that way. Heck, with the VMS words, I don't need to work that way.
Please notice I am not giving or implying any meanings with my deconstruction scheme. I am not even giving any hint of the language behind the words. Neither is relevant here. This is about the scheme.
The structure upon which the codes in the code groups are hung to form words is very rigid. The code group representatives always appear in the same order, if present - Group I, Group II, Group III, Group IV, Group V, Group VI. The order of deconstruction is always the same. The words are formed the same way and from the same code groups throughout the VMS. This includes the Voynich A and Voynich B and label words - all of them.
A word or two about the VMS words that only appear once in the VMS - most of them also appear to deconstruct in the same manner using the same code tables.
The others that don't deconstruct successfully can mostly be made to deconstruct if more codes are added to the tables. This is most evident in the VMS words which contain glyphs not yet found any of the code groups - of course new codes would have to be added to include, into the scheme, words containing those glyphs.
Some VMS words that only appear once in the manuscript are blurry, indistinct or have other reasons for not being legible or having questionable transcriptions. I am presently working on a glyph/EVA transcription of the VMS showing which words are legible and which have questionable aspects to help alleviate this problem. Then I will tackle deconstructing the single-appearance VMS words. It'll be a while.
If it weren't that only twelve repeated words seem to fail to deconstruct, I would also look upon my proposed scheme with more doubt. But the other 2,000+ repeated words do deconstruct so consistently, don't they?
Whatever the reader decides about what it all means, I think I give a pretty good rationale, and show the proof, for thinking that my proposed scheme has a basis in fact, not supposition.
That's because I am trying to show what is real and true and can be proven.
I didn't write the Voynich Manuscript, but I think I can show how the words in it were constructed out of smaller parts. And I don't need to change/add/omit/rearrange/exchange any gylphs, nulls, spaces or words to do it.
I prefer to go by what's in the book.
Thank you.
Don of Tallahassee
RE: how to deconstruct the VMS text and label words repeated twice or more - -JKP- - 27-05-2016
(26-05-2016, 04:22 PM)don of tallahassee Wrote: You are not allowed to view links. Register or Login to view....
I know many of the leading lights in the group trying to solve the puzzle of the VMS do not like that my scheme seems to show some way to construct and deconstruct the VMS words. For some of them, it would obviate years of work trying to find a solution using more modern techniques to account for an early Fifteenth Century manuscript. For others, that fact that the scheme cannot be made to fit any known language's words as a direct deciphering/transposition scheme of some sort or another is anathema. That's okay....
Sorry I didn't respond sooner, Don, but I've been swamped with work (no sleep for days).
You won't get any arguments from me, Don. I've been saying for a long time it doesn't fit the STRUCTURE (and I've often used caps when saying this) of a natural language and that "the rules are strict". I also alluded to some of those rules in my blog about how to write Voynichese so it looks more like Voynichese (it's a blog, aimed at a general audience, not all of whom are interested in decoding the script, they just want to write it so it looks like Voynichese and I wanted to encourage interest in the manuscript, so I tried to keep it short and based it only a fragment of the text, BUT the concepts of construction (a rule-based combinatorial method) are at the heart of how I see the text).
Also, I get the feeling he came to his conclusions independently because he has good observation skills and has been following the "logic" of the script for a while (rather than trying to impose something upon it), but if you look at Wladimir D's charts, you will see he is doing much the same thing but using a flow-chart style. Different approach, same idea.
I haven't had a chance to look at Torsten's system yet (or to try his algorithm as I'm on a Mac and I think it was written for Windows?) but from glancing at his thread (Torsten, sorry I haven't given it more attention), I suspect he's evaluating the text in a similar manner, as well.
So, while there may be detractors out there, I think you will find you also have support for your approach.
I also have charts that are more comprehensive (and specific) than what I alluded to in my blog (didn't want to let the whole cat out of the bag, just the nose). Mine are constructed from a slightly different point of view from your charts and from Wladimir's charts but yield very similar results.
One thing I'd like to add to what you've posted... you've created a very good deconstruction, it's a good foundation. There are two things that have prevented me from posting what I have (I constantly wrestle with an unfortunate personality quirk that I want to figure it ALL out before talking about it):
- In the main text (which may differ slightly from the labels) there appears to be an additional dynamic that goes beyond the word-tokens, and that involves the construction of a line. In other words, there are rules for the "syllables" (what they are and where they are located) and there are ALSO rules for lines (you can particularly see this at the ends of lines) and I haven't fully puzzled those out yet. It's a challenging problem but may be important.
- There are a few "syllables" (deconstructed groupings) I have not resolved yet (and which might explain some of your "exceptions") because they are sometimes in between other deconstructed units, which makes it unclear which one of the glyphs belongs to a) the preceding "group", b) the following "group" or c) to itself (you may already have noticed this). These are a minority of cases, most of the tokens can be found in enough context throughout the manuscript to determine whether they belong on the right or left side or neither, and it probably won't be necessary to unravel this minority to figure out the VMS, but... it's still important, at some point, to resolve the length and composition of these ambiguous groups.
One more thing I have observed. It is my belief that these groups (for a few years, I've been thinking of them as "atomic" groups) have specialized functions. In other words, I believe there is a grammar involved (it's not as simple as stringing adjectives or nouns together), one that borrows concepts from natural languages, but which is specific to the VMS. In other words, imagine a constructed language that borrows some of the basic grammatical rules found in older languages in a more regular way than we see them in our mongrelized loan-word-evolved modern languages.
I have a list that includes some of the grammatical rules that I'm fairly sure of, but there are a few I haven't nailed down yet and once again, I am highly reluctant to post them until I have a more complete list/analysis.
RE: how to deconstruct the VMS text and label words repeated twice or more - don of tallahassee - 27-05-2016
Dear JKP,
Thanks for trying to understand my points. And for agreeing with some of them.
I think the matter of position in lines of words will depend most on what the subject of the book is and how the info is being transmitted (what each of the six code groups are doing or about).
The first group of codes (Group I) I think I have found is large enough for the kind of thing I think you are talking about. I don't know if the other groups (Groups II through VI) each have enough members to be able to do much in showing grammar.
But, I don't think I fully grasp your grammar parts idea. Is this like a base word with additions that show it is an infinitive because of an appended code and a verb because of another code? And the next word's base word is modified by showing it to be a noun and plural.
Or is it nearer to the base word 'pend' being modified by adding 'com' to the front and "ium" to the other end? The 'ium' ending would maybe be used for all past participles or some such?
Or am I missing some basic part of your explanation? Could you send me a draft copy of this part of your work, please. I'm always seeking new explanations that may work better than my own ideas. I won't be picky about the loose ends.
I think you imply/say the VMS is about a made-up language the author also developed in conjunction with the glyph idea and the code group idea.
Are you propounding some sort of readable text instead of lists of things? That would make more sense than my ideas of what it's all about.
If the VMS instead shows some type of early Fifteenth Century lists of things, one of the problems is trying to find things that can be included in a list of possible subjects for the VMS that would employ the different groups shown in each word. So far I've come up with herbal potion ingredients for medicinal potions (dill, seeds, 2 grains), alchemy ingredients in formulas (Philosopher's stone, the red kind, dry, 6 pinches), metallurgical ingredients in formulas (iron oxide, the red kind, 6 hundredweight), census/Domesday book entries for some city, duchy or kingdom or other such (serf/archer, 6, expert with longbow, with 200 arrows, with hand cart). Diane O"Donovan thought of modern-day auto parts lists. That got me to remembering the type of nomenclature used (and also encoded) in the old US Army lists (M151A2, Jeep, without top, with M1 machine gun mounting, camouflage paint color).
What other topics might have required a 220 page book to enumerate the elements?
To get back to the word position question - if The VMS is a list of things or group of lists of things, it may not matter so much where in a line a word appears, maybe not at all. If a VMS paragraph shows recipes, metallurgical processes, alchemy potions, or lists amounts and kinds of different soldiers/laborers, the placement of a word (ingredient) within any given line may be entirely at the whim of the author or compiler. And I think we both agree the author of the VMS was great at devising puzzling ways of doing things that helped to hide his purposes and secrets. The EVA = m glyph at the end of three consecutive lines on a page may be more the result of intentional meddling by the author than anything else. I notice there are lines with words ending in EVA = m that aren't at the end of a line on some pages. I think I even remember one such word appearing in the middle of a line with another one at the end of the line.
One of my earlier comparisons of the VMS words was to the early IBM punch cards with information to be encoded into designated fields using code tables for some/all fields. If no info was encoded for a field, the machine and the human reader could still figure out what was being encoded for the other fields because there were rules to consistently be followed for each card's set of fields. That is the same as the deconstruction sequence I envision, except the punch cards were read starting from the left - not in accordance with the right end, left end, middle mantra of the VMS deconstruction sequence.
Please let me know more about your take on things Voynich. You can send the work/results/drafts to me at:
You are not allowed to view links. Register or Login to view.
I think I have the other kind of personality - I've been trying to show all my work's results as it progressed since I figured out the sounds for the first fifteen or so glyphs five years ago (10 or 11 March 2011 - on the Voynich-List) and worked from there. (Yeah, I started my journey on the far side of nowhere to get to the deconstruction ideas.)
Thank you.
Don of Tallahassee
RE: how to deconstruct the VMS text and label words repeated twice or more - -JKP- - 27-05-2016
(27-05-2016, 06:30 AM)don of tallahassee Wrote: You are not allowed to view links. Register or Login to view....
I think the matter of position in lines of words will depend most on what the subject of the book is and how the info is being transmitted (what each of the six code groups are doing or about)...
Don, if it were only this (line ends varying based on content) then I probably wouldn't have mentioned it.
But this is something different. Just as the "words" have a certain way of being constructed, it appears that lines (or possibly the ends of lines) also have a certain way of being constructed. Look through some of the dense-text pages (especially scan down the right-hand sides) and you might see what I'm talking about.
But, I don't think I fully grasp your grammar parts idea. Is this like a base word with additions that show it is an infinitive because of an appended code and a verb because of another code? And the next word's base word is modified by showing it to be a noun and plural.
Or is it nearer to the base word 'pend' being modified by adding 'com' to the front and "ium" to the other end? The 'ium' ending would maybe be used for all past participles or some such?
Or am I missing some basic part of your explanation? Could you send me a draft copy of this part of your work, please. I'm always seeking new explanations that may work better than my own ideas. I won't be picky about the loose ends.
It's hard for me to answer this because I think I've figured out some of them but others are not so clear and until I get at least a couple more, I won't know if I have to tweak the first ones. It's a bit like a Rubik's cube where you have to be willing to tear apart what you have to get to the last step.
Here's what I can say... it appears to me, so far, that the rules are fairly constrained (just as the "syllables" are quite constrained), in other words, limited to basic necessities in the grammatical sense. I'd prefer to work this out in more detail before submitting it. As you've probably realized now from my previous post, I'm reluctant to post half-baked stuff (or to send it privately to others). Even writing a blog is hard for me because the blogging format, by necessity, means simplifying things, sometimes almost to the point of absurdity (papers are better, they can be longer and more detailed but I can't spare the time for papers right now).
Are you propounding some sort of readable text instead of lists of things? That would make more sense than my ideas of what it's all about.
Yes, I am. "Readable text" might not be the best way to describe it, but it goes beyond lists. These "atoms" have a relationship to each other and I believe some have specialized grammatical functions (like modification).
I created my own VMS font long ago (I've never used the EVA font) and also my own transcription of the text, and discovered, after working on this for a while, that I could type what was going to come next with a fairly high degree of accuracy before I had seen it. Not long parts, usually only 2, 4 or 6 characters, but that in itself says something about how the text has been constructed.
One of my earlier comparisons of the VMS words was to the early IBM punch cards .... If no info was encoded for a field, the machine and the human reader could still figure out what was being encoded for the other fields because there were rules to consistently be followed for each card's set of fields.
This is a VERY good example and it is, in fact, possible to figure out some of the faint, scratched, partially covered, and hard-to-read parts of the manuscript once you grok the rules. Not all—there is more to it than that, there are word-tokens in between the predictable sequences that are uncommon or unique, but the proportion of predictable sequences is quite high and I think people should take note of your analogy.
RE: how to deconstruct the VMS text and label words repeated twice or more - Koen G - 27-05-2016
There is one problem with these approaches that look into the structure of the VM language is that they aren't very easy to judge for the layman. Person a. comes up with a way to understand the structure that works, but so does person b. and c. and so on. For someone who isn't used to this kind of work, it's really hard to get a decent view on this.
What I would genuinely like to read is for example a comparison of the various proposed solutions by someone who's used to working with the Voynich structure, explained in a way that's understandable for those of us who are more used to focusing on the imagery. Are the analyses compatible, do they reveal common patterns, does one point out things that the other overlooked...
This might focus some of the attention towards structural issues, because I agree with JKP and others that fully understanding the structure is one of the first steps required. It would also be nice to have some more direct comparison between various theories. I just don't feel qualified to do this myself because so far I have spent like 90% of my research time on the imagery.
RE: how to deconstruct the VMS text and label words repeated twice or more - don of tallahassee - 28-05-2016
Dear Koen Gh.,
I think I know what you mean. Being a retired carpenter, I cannot follow or understand some of the other approaches well enough to talk about them in depth. I can't even tell if any of them work successfully most of the time (maybe 98% of the time?) or not.
Maybe their authors will be able to tell you if their theories are similar and/or give similar results. Maybe they can point out flaws in my ideas or work or results. Or maybe other people will show me the error of my ways.
I don't believe anyone else has previously offered anything like my proposed code group deconstruction ideas in such detail or showing an apparent 99.5% success rate for such a large sample of the VMS words' consistent construction/deconstruction.
Thank you.
Don of Tallahassee
RE: how to deconstruct the VMS text and label words repeated twice or more - don of tallahassee - 02-06-2016
If my proposed deconstruction ideas do, indeed, account for a way to understand how a large group of the VMS words were constructed, ii would seem the following statements about the Voynich Manuscript are true.
The Voynich Manuscript glyph words are made up of parts - coded parts - of some kind and for some purpose..
One of the groups of coded parts (Group I) is present in all VMS words.
Only a single Group I code is found in any VMS word, as is also the case with the other code groups (Groups II - VI), if they are present at all.
There seems to be a consistent order and system that can be used to separate the codes in each word, in a logical way, to deconstruct the words.
There are a few words in the VMS (12 of the 2,000+ words found twice or more) which do not fit the deconstruction scheme, groups and sequence or operations, at all. I am ignorant of why they were not originally constructed by the author to fit.
There seem to be one of several different and distinct series of code groups contained in most of the shorter VMS words.
The one-glyph words contain only a Group I code.
The two-glyph words contain a Group I code plus either a Group V or Group VI code, as in:
Group I code, Group V code, or
Group I code, Group VI code.
The three-glyph words are usually constructed of:
Group I code, Group IV code, Group V code, or
Group I code, Group IV code, Group VI code, or
Group I code, Group V code, Group VI code.
VMS words made up of four or more glyphs all contain a Group I code and a Group V and/or Group VI code as part of their construction.
Yes, there are also other code groups represented in the other VMS words - this is about the most common groups of deconstruction code groupings.
Since it seems to show an almost universal pattern, the proposed deconstruction method, codes and sequence must be accounted for in any proposed meaning for the words in the Voynich Manuscript. I don't think the proposed solution works so widely by accident nor can it be excluded from consideration in finding any ultimate meaning of the words.
This proposed solution would also explain why the EVA = q glyph is only found on the left side of VMS words - it's only found in Group I codes.
Thank you.
Don of Tallahassee
RE: how to deconstruct the VMS text and label words repeated twice or more - don of tallahassee - 02-06-2016
I must make a confession.
I am not now sure the list I used of labels found two or more times in the VMS is a correct list.
I have been asked a question about the list I used but can no longer find, and so cannot check the original list.
If anyone knows of a list of labels starting with those found 10 times in the VMS, I would appreciate very much a head's-up. I alphabetized the portion of the list I used.
Thank you.
Don of Tallahassee
RE: how to deconstruct the VMS text and label words repeated twice or more - don of tallahassee - 24-07-2016
I have generated (counted up and done the math) a few statistics for the list of words repeated two or more times in the VMS and deconstructed at my Fumblydiddles.com site:
There are slightly less than 2,400 words on the list.
They account for most of the total words in the VMS - more than 85%.
The words on the list are composed of a Code I code followed by 1 or more (sometimes none) of about 55 other codes that seem to somehow modify the Code I codes.
Of the slightly less than 2,400 words, about 20% contain a modifying code from Group II or Group III.
There are 20 members of Groups II and III.
If those twenty members are removed from the total of about 55 modifying codes, the remainder is 35 codes.
This means about 1,900 of the 2,400 repeated words in the VMS can be constructed using a Code I code followed by from 0 to 3 other code group members of which there are only 35 total members.
And those 35 other codes seem to appear in a strict order - Group I, then Group IV (if present), then Group V (if present), then Group VI (if present). The codes from each Code Group are not represented twice in any word, nor do they appear in words in any other order.
The almost 2,400 words seem to show only a few modifying codes (35) with which a great majority of the repeated words (about 1,900 - about 80%) can be composed.
That seems like a small number to me, but the results show it to be true.
If some other solution is proposed by another person (or me), I think it has to make some sort of provision for the above statistics.
Thank you.
Don of Tallahassee