Greetings,
My name is David Leimbach and although I've had an interest in the Voynich Manuscript for many years this is my first time writing about it.

I'd like to draw attention to some issues that may be misdirecting our intuition and a solution to that misdirection. 

Here I try to take a look through a scribes eyes and build a small profile regarding glyph consistency.  Due to the severe lack of errors and the consistency with which glyphs are written in the text I don't think it's controversial to suggest each scribe was very skilled at their work but with some caveats.

It is clear there is great consistency among some glyphs but a lack of consistency among others.
Gallows glyphs are all over the place in height and form.  Meanwhile many other glyphs look almost rubber stamped in their consistent size and shape.  I would suggest this means we have many commonly known glyphs that scribes would have spent their entire careers writing and some new unusual glyphs, but to us and our scribe these are not the same sets of glyphs.

I propose Voynich glyphs are not all as unusual as they appear to us.  I believe our scribe would be very familiar with some glyphs and that familiarity would help them write many consistently, but would actually know them as numbers.

Even the most skeptical must admit:  the EVA-y is unmistakably the number 4.  If this is a 4, could EVA-o be a 0?  And then where are the other numbers?  I propose many of the glyphs that are being considered as letters, should instead be read as numbers.

Our scribe as would any scribes of the time see these glyphs EVA-o,i(and n),r(and m),l,v,d(and g),y and see them immediately for what they are.  The numbers 0,1,2,4,7,8,9.  Of course 3,5,6 have no place in any self respecting book of appreciation for lumpy women so I must leave these to you to decide

    Number images from Ms. Barth. 24, c. 1460s, Rhein region.

These characters of note can easily be read as:
EVA-n is the number 1 at the end of a "word"
EVA-m is the number 2 at the end of a "word"
EVA-g is the number 8 at the end of a "word"
These all resemble the suggested number with an additional descender or end of word flourish and all appear only at the ends of words.

         

Furthermore EVA-i as the number 1 appears in the middle of words, but never at the end.  EVA-n appears only at the end of words.  Here's a little bit more evidence as I interpret this bit from folio 5v.  EVA-n is usually made in a single pen stroke.  You see two identical words, except on the second one it appears the scribe forgot the tail, finished the 1 and then added it in an additional stroke.  The fact that the first stroke would by itself be an EVA-i adds to this impression.

         

This leaves gallows and our new alphabet is much reduced:  EVA-a,c,s,e,h,q,x,z

My reference for numbers is taken from Ms. Barth. 24, c. 1460s, Rhein region.  Here is a sample (page 4)

   

Unfortunately seen through these eyes does not give the VM such a sexy look that we're used to seeing anymore.  To see the VM as a medieval scribe might and replacing gallows with characters on my keyboard leaves us looking at it like this.  Here is the first paragraph of 1r, for illustration I've also swapped the gallows
EVA-f = &
EVA-p = $
EVA-t = #
EVA-k = %

&ach9s 9%a4 a2 a#a111 sh04 sh029 c#h2es 9 %02 sh0489 s029 c%ha2 02 9 %a12 ch#a111 sha2 a2e c#ha2 c#ha2 8a1 s9a112 she%9 02 9%a111 sh08 c#h0a29 c#hes 8a2a111 sa 00111 0#ee9 0#e0s 2040#9 c#ha2 8a111 0#a111 02 0%a1 8a12 9 chea2 c#ha111 c$ha2 c&ha111

The VM would have appeared much more familiar to a reader at the time it was written as a book of codes to be deciphered and not seen as a language to be translatable into any other written or spoken language. 

Clearly this approach to looking at the glyphs has lots of implications, first of which is that there are no translatable words in the manuscript as commonly thought.  Instead we're looking at codes which refer to some unknown keys that we must somehow reconstruct.

Thank you for your consideration.

In f67r2 (see Beinecke scan You are not allowed to view links. Register or Login to view.) there is a detailed band around the presumed zodiac chart, which has a discontinuous/inconsistent pattern of blocks that I would assume has some symbolic purpose. I searched a bit but haven't found any discussion about it.

The ring is partially obscured by a damaged edge on the left side of the page. There are 44 blocks visible, but based on the consistent width of each block the diagram should have had exactly 50. There are minimally seven distinct types of blocks, mostly the same for each type, but with some distinctions.

Is there anything similar in other manuscripts? Does any of this stand out as familiar symbolism to anyone more knowledgable?





The outer band of f67r2:

   

The four quadrants of the band, unrolled for easier viewing:

   

The various block types:

   
(There are three instances of the "bubble" type, but only one of them is clear. The top-right and center-right instances are both visually difficult to parse. There are either mistakes/imperfections in each, or both of those have different bars on the edges than the top left bubble block. There are also three small ink dots within one of the pins blocks in the bottom right, which is unique and possibly intentional.)

Also of interest is that the number of separating "bars" between each block varies around the ring, in threes and fours, possibly a single five in the top right, and a single two in the bottom center (possibly a mistake? in that block one of the two center 'spikes' is floating, which is unique)
   

I believe the Voynich Manuscript is a "House Manual" written down by an illiterate yet intelligent person who came up with their own phonetic written language for a spoken dialect at Castello del Buonconsiglio, in Tyrolia in approximately 1420. This was a German-speaking court ruling over an Italian-speaking population. A servant working here would hear German from the Bishop, Latin from the Priests, and Ladin/Italian from the staff.

I took this idea and started analyzing gallows characters, they appear disproportionately at the beginning of paragraphs and the beginning of lines. This makes sense for someone transcribing their own speech. Then I looked for the ways words might end, short vowels, hard vowels, and long fading sounds. Each one seems to be depicted in the text differently. If we stop treating the symbols for the different sounds ending words and instead start collapsing them into a single sound, the vocabulary of the "cypher" makes more sense.

Here is where I'll probably lose the interest of the pros here, but, I fed this information into a modern LLM. I'm fine with this thread being relegated into the AI slop bucket. But. When fed into the Zodiac illustrations you get A-bur-ril for April, You get Oct-ember, (which could be an illiterate person trying to remember the Latin suffix incorrectly). They get Mars for March, and Mai for May. 

When I feed random sections to the AI with this idea in mind. I get text that looks like this.

Line 1 (EVA Transcription)
fachys ykal ar ataiin shol shory ctthy res y kor sholdy

1. Analysis & Prefix Stripping:

fachys: No prefix. Root: Fach (German: division/plot) or Wachs (German: grow - phonetic swap f/v/w common in dialect). Selected: Wachs (Grow).

ykal: Prefix y- (Um...). Root: Kal. German Kahl (bare) or Ladin Cal (stalk/stem). Selected: Cal (Stalk/Stem).

ar: No prefix. Ladin Ara (plowed field) or German Hier (here - phonetic). Selected: Hier (Here).

ataiin: Prefix a- (often directional or just noise). Root: Taiin. Ladin Tai (cut) or Taia (cutting). Selected: Tai (Cut).

shol: No prefix. German Soll (Should/Must).

shory: No prefix. German Schor (Shorn/Sheared).

ctthy: Prefix ch- (Throat clear) or c- (Look/Ke!). Root: Thy. German Die (The).

res: No prefix. Ladin Res (Thing/Matter) or German Ries (Giant/Big). Selected: Res (The thing/plant).

y: Prefix y- (Stammer/Um).

kor: No prefix. Ladin Cor (Heart) or German Korn (Grain/Seed). Selected: Korn (Seed/Grain).

sholdy: No prefix. German Schuld (Debt/Duty/Must).

2. Draft Translation (Stream of Consciousness): "Grow... um, the stalk... here... cut [it]... should be shorn... [throat clear] the thing... um... the seed... must [be done]."

3. Alternative Check (Ladin vs. German Context):

Check: Could fachys be Ladin Fasces (Bundles)?

Alt Translation: "Bundles... um, of stalks... on the field (ara)..."

Decision: The "Grow/Wachs" root fits the start of a herbal manual better than "Bundles". I will stick with the German root for the verb.

4. Final Translation (Folio 1r, Line 1):

"Grow... [eh]... the stalk here. Cut it... [it] should be sheared... [kch] the thing... [eh]... the seed is mandatory."



I've applied this to most of the EVA text and it's all pretty legible. If there is interest I can post the document I have the whole thing in.

Would anyone here be willing to sanity check me?

Thanks
Zbigniew "Z" Ciupek

I have seen many pictures of these pages on Wikipedia, and the first thing that appeared in my mind was that this is a distillation—a process map—from Southern Europe, near the Mediterranean Sea. My approach assumes the author was writing sounds, not correct spelling. I am originally Turkish and our language is written as it sounds. My son who grew up in Denmark and Germany always struggles the same way when it comes to write proper Turkish words. Either he tries to write German words in Turkish sound or visa versa. I have decrypted Voynich Manuscript by my own human ideas at 02.01.2026 with the help of AI just to apply Latin linguistic methods since I can not speak Latin but know the way they sound.

Moving away from complex cryptographic substitution theories, this hypothesis suggests the text is a phonetic transcription of Latin and Romance dialects written by an author with dyslexia or a similar learning difference. When read phonetically, the text reveals a coherent technical manual describing volcanic distillation, thermal bathing, and herbal processing, likely localized to the island of Ischia (Aragonese Castle) in the 16th Century.

Please see attached PDF to be able to view my full hypothesis.

Attached Files

  Voynich_Dyslexic_Phonetic_Theory.pdf (Size: 94.99 KB / Downloads: 24)

First I want to say I did use AI (Claude) for counting and translating latin but it's a human theory:


I think the Voynich Manuscript uses a specialized shorthand for herbal recipes, like medieval "chemical notation." I know there are others with similar theories, but I think it might be the right direction. 
Not a letter-to-letter mapping but it uses morphological encoding. 

Original text:

fachys ykal ar ataiin shol shory cthres y kor sholdy
sory ckhar or y kair chtaiin shar are cthar cthar dan

Translation:

"Take fresh herbs from root pieces, extract small amount of oil essence from flowers.
Prepare flower water (infusion) with chopped compound root and leaf powder pieces."

The Rosetta key:

Each word = [PREFIX] + [ROOT] + [SUFFIX]

PREFIXES (what you're doing):
Examples (not exhaustive, I don't have them all) 

sh- = extract/essence (25 occurrences)
ch- = compound/mixed (17×)
d- = dried/powder (18×)
o- = with/oil (19×)
s- = water / liquid
y- = of / from

These almost always appear at the start of words.

ROOTS (what ingredient):

ol = oleum (oil)
ar = radix (root)
or = flos (flower)
a = aqua (water)
ai = folia → leaf (probable)
al =  salt / mineral
e= essentia / essence

Roots appear in the middle of words.

SUFFIXES (grammar/quantity):

-y = singular (43×)
-n = plural - 100% always final!
-iin = genitive plural (19×)
-ain = dative/for purpose (12×)
-ol = diminutive/small
-dy = dried state
-ar = locative (in / at / from

Suffixes appear only at the end of words.

Example: sholdy = [sh=extract] + [ol=oil] + [dy=dried] = "dried oil extract"

Why I think it's plausible:

1. 'n' appears in final position 100% of the time (37/37) and I don't think that's possible in natural language but perfect for a systematic suffix

2.Gallows characters never appear final - t is medial 96% of the time (27/28). These mark PROCESSES: boiling, grinding, mixing. Special characters traditionally called “gallows” are interpreted as process indicators and not letters.

Examples: t, k, ch, th, ph  - boiling, grinding, chopping, mixing

3.Low entropy (3.78 bits) - matches technical terminology, not prose

4. Follows Zipf's law coz it proves it contains real meaning,

5. Words cluster by context, the oil words appear together and the water words appear together

[Sorry of this has been asked before.  I tried searching the site but came up with nothing]

The central figure on the Rosettes page (f85v2, fRos) shows six towers with "onion" roofs, apparently supporting the starry heavens.

Has anyone found similar imagery in manuscripts from 1400 or earlier?  Or actual buildings with those features?

Towers just like those -- round, with ribbed onion roofs, topped with a tapered trumpet-shaped cone piercing a ball -- seem to be characteristic features of Medieval Russian Orthodox cathedrals, and their more recent "Revival" style.  What was the geographic extent of that style?  Did it reach Central Europe? 

It seems that the Russian towers were usually built into the cathedral.  The towers in the VMS, on the other hand, are free-standing, with characteristic "lobed swelling" at the base.  Is there any parallel to those bases, n imagery or actual buildings?

All the best, --stolfi

Specifically, I'm looking for depictions of women that have red cheeks, attire or hairstyle similar to those in the MS, and/or facial/body structure depicted similarly as in the MS.

I've found a lot of good candidates in southern Germany, particularly in the Cod. Pal. Lat. fonds of the Vatican Library (which was originally the collection of Heidelberg). 

I'm attaching pictures of two pretty strong resemblances. The first is Cod. Pal. Lat 1726, a "Mythological Miscellany" manuscript.

Note in particular the "nymph-like" lady in repose, together with doves that look kind of like the bird we see in the VMS:
You are not allowed to view links. Register or Login to view.

And the heavenly squiggle clouds.
You are not allowed to view links. Register or Login to view.
The posed people on the left hand pages also appear similar to the "Gemini" page of the VMS, which I believe someone (I think Koen?) has traced back to a possible link with manuscripts of the Wigelaf romance. But that's another post 


The other, Cod. Pal. Lat. 1709 is an miscellany manuscript that contains a lot of astronomical and mythological illustrations. 

Here is one of the Biblical plagues:
You are not allowed to view links. Register or Login to view.
(So maybe this is Moses and not a woman- but still, note the crown and cheeks)

Here is one depicting the influences of different planets:
You are not allowed to view links. Register or Login to view.

What other MSs like this have you found, and where are they from?[url=Google.com][/url]

I left the voynich conversation a quarter century ago or thereabouts. So, as if that was yesterday, here I am continuing about where I left off with page (and subsequently paragraph) initial words.

Transcription errors are very likely, but I do think the statistics generally do hold true when you look across the entire manuscript. The words below are among the ‘rare’ words unless you clip off the leading gallows, but I think there’s more to them.

f3v (bold below) is particularly intriguing with the first and third paragraph initial words being eerily similar to the way the second and fourth are structured.

I mixed up how I handled split gallows and word-adjacent words that might be prefixed by the first word. Using period or hyphen to join them. It’s pretty clear with long split gallows across several words. I included them as part of the paragraph or page initial words because I felt they were needed.

The rarity of these words doesn’t look quite as rare when only looking at paragraph initial words.

First 7 quires paragraph initial words:

Quire 1
Fychys ?odar ?ydain Cpho-shaiin 
Kchys  Potoy 
Kydainy Kydain
Kaoiin Kchor
Tsheos Pcheol Tsheoarom Pcheoldom 
Koaiin Tchor
Kodalchy Pydaiin 
Pchaoiin Torchy 
Kshody Tshy 
Kocheor Tchody 
Foar 
Koar 
Pchodaiin Ksholochey 
Polyshy Kchor 
Pshol Tshosp Ctho..cthey 
Cthod.. soocth Pchar
——-
2
Tydlo Pshoain 
Fochar Pchor 
Pchocthy-shar Ycheor
Paiin Qotchy 
Tshol Tchol 
Polchody 
Torshor Shorodo 
Koair Foldaiin 
Pchodaiin Soshy 
Pdychodaiin 
Tshor 
Poror 
Pocheody Tchor Toror 
Pchraiin Pchocthy 
——
3
Fshody Tchom Ksheo 
Pchodol Kchor Sor Qoain Oal 
Pdrairdy Tchor 
Told Tolol 
Pchor 
Pochaiin Tod 
Kdchody Pchocthy Fchodees 
Faiis-ar Tshol 
Pchor Ochofychy Fchokshy 
Toldshy 
Pololshy Pchaiin Kchol Kchol 
Pysaiinor Fshor 
Pydchdom Qolkody Tshol 
Podairol Tshol 
Porory 
Tchodar Tochol Ksho 
——
4
Fcholdy 
Poeeaiin 
Psheoky Fcho 
Pchedar Pchdar Pchedy 
Ksor Kchey
Fochof 
Pchodar 
Kshol Tshoiin Pchol 
Poraiin Kcheol 
Koaiin Kochor Qokaiin
Okchesy Opchol 
Cthscthain oyshy 
Keedey Tshokeody Tolshso 
Podair Pcheeody 
Fchaiin Fcho 
Kcheodaiin Ksho 
——
5
Tshdar 
Tarar Tshdy 
Pcheoepchy Tcheo 
Kschdy Pchedar 
Cthoo-r-choly-cthy Paiin 
Parchor 
Pchafdan Podaiin 
Pcharosy Tchor 
Tocphol Pchotchy Koiin 
Kshody Qotor Todain 
Tolor 
Okchop 
Tedochshd Pchdaiin Pchdar 
Pdair Pardy 
Pchey Ksheo 
Pchedain Teeypchody Toees (and ‘title-word’ Pchedy)
——
6
Pshey.Kedealeey Shedey.polchedy 
Pcheody (keeshdal?)
Stho-ofaiin-cthaihcthy Pcho Pydaiin
Tcho..cto..sheey Posheor 
Tarodaiin Pshesy Pshdar 
Pdsairy Tolkshdy Tarchor 
Tshodpy Toy Pshy 
Tsho Yokalod Tsheody 
Pykydal Kolshol Tshol
Korary Qotol Yksheor 
Pcheocphy Tedy Tshdy Pchdair 
Pody 
Pchair Folr 
Psheot Pchodaiin
Pshdaiin 
Pcheodchy Pchedar 
——
7
Poshol Ksho Podaiin 
Kshor Fchochor Qotcho 
Psheor 
Tchy 
Tsholdchy 
Poshody 
Tdokchcfhy 
Pchor Pcheol 
Kodam
Tshor 
Podaiin Korare 
Pcheodar 
Podaiin Tchedar 
Kcheedchdy Okeed 
Otchal-chchsty Tchoky
Kcheat Kchokchy 

Anyone who's been around the Ninja (or Voynich Mss. related discussions in general) is familiar with the existence of certain glyph pairs with unusually high frequencies which are significant contributors to the low conditional second-order entropy of the text. To pick an obvious example, here are the 10 most frequent glyphs following Currier 'O'/EVA 'o' in running text lines in ZL_ivtff_1b.txt converted to Currier:

  kgram:     OF     OE     OP     OR     O8     OB     O2     OC     OA     OX
  (EVA):     ok     ol     ot     or     od     op     os     oe     oa   ockh
   Rank:      1      2      3      4      5      6      7      8      9     10
  Count:   5620   5242   3244   2531   2035    519    382    322    257    179
 REFreq: 0.2609 0.2434 0.1506 0.1175 0.0945 0.0241 0.0177 0.0150 0.0119 0.0083
RECmFrq: 1.0000 0.7391 0.4957 0.3451 0.2276 0.1331 0.1090 0.0912 0.0763 0.0644

Note the steep drop from OR (11.75%) and O8 (9.45%) to OB (2.41%) and O2 (1.77%).

If certain glyph pairs go together as a unit, uncertain spaces before and/or after may reflect an unconscious hesitation on the part of the scribe. Here are the 20 most frequent glyph pairs with an uncertain space after them:

  kgram:    OE,    AR,    OR,    AE,    CO,    89,    C9,    SO,    AM,    AT,
  (EVA):    ol,    ar,    or,    al,    eo,    dy,    ey,   cho,  aiin,   air,
   Rank:      1      2      3      4      5      6      7      8      9     10
  Count:    387    207    181    165    121    101     96     58     55     51
 REFreq: 0.2043 0.1093 0.0956 0.0871 0.0639 0.0533 0.0507 0.0306 0.0290 0.0269
RECmFrq: 1.0000 0.7957 0.6864 0.5908 0.5037 0.4398 0.3865 0.3358 0.3052 0.2761

  kgram:    S9,    AN,    4O,    P9,    O2,    F9,    C8,    ZO,    OP,    C2,
  (EVA):   chy,   ain,    qo,    ty,    os,    ky,    ed,   sho,    ot,    es,
   Rank:     11     12     13     14     15     16     17     18     19     20
  Count:     34     25     24     20     20     19     19     18     14     14
 REFreq: 0.0180 0.0132 0.0127 0.0106 0.0106 0.0100 0.0100 0.0095 0.0074 0.0074
RECmFrq: 0.2492 0.2313 0.2181 0.2054 0.1948 0.1843 0.1742 0.1642 0.1547 0.1473

...and here are the 20 most frequent glyph pairs with an uncertain space before them:

  kgram:    ,SC    ,AM    ,ZC    ,FC    ,8A    ,AE    ,FA    ,OE    ,AR    ,89
   Rank:      1      2      3      4      5      6      7      8      9     10
  Count:    195    162    149    126    124    101     97     86     82     70
AllFreq: 0.0011 0.0009 0.0008 0.0007 0.0007 0.0006 0.0005 0.0005 0.0005 0.0004
 REFreq: 0.0910 0.0756 0.0695 0.0588 0.0579 0.0471 0.0453 0.0401 0.0383 0.0327
RECmFrq: 1.0000 0.9090 0.8334 0.7639 0.7051 0.6472 0.6001 0.5548 0.5147 0.4764
  kgram:    ,SO    ,4O    ,FS    ,OR    ,AN    ,PA    ,AT    ,PC    ,EF    ,AJ
   Rank:     11     12     13     14     15     16     17     18     19     20
  Count:     55     54     45     42     39     29     28     27     26     26
AllFreq: 0.0003 0.0003 0.0002 0.0002 0.0002 0.0002 0.0002 0.0001 0.0001 0.0001
 REFreq: 0.0257 0.0252 0.0210 0.0196 0.0182 0.0135 0.0131 0.0126 0.0121 0.0121
RECmFrq: 0.4438 0.4181 0.3929 0.3719 0.3523 0.3341 0.3206 0.3075 0.2949 0.2828

None of those counts are huge given the total number of glyph pairs in the running text, but there is at least a weak signal with regard to some of the most obvious candidates like OE, OR, AE, AR, and the various word-end specific A<x> combos like AM, AN, AT, AJ.

Of course, the above results need to be taken with an appropriate grain of salt given disagreements between transcribers regarding whether something is a clear or uncertain space, or whether there is an uncertain space in a given position at all. Nevertheless, thought it was worth throwing out there as something to think about.

Happy New Year to all readers & posters on the Ninja, and best wishes for a happy & healthy 2026.

Hello everyone,

I'm new to the forum but have been working on a quantitative analysis of the Voynich Manuscript. I've developed an approach called the "Nordico Paradigm" and would appreciate feedback from the community.

**Key aspects of the approach:**

1. **Methodology**: Quantitative, reproducible framework based on 130 folios (66.3% of the manuscript)
2. **Core metric**: Continuum Index (CI) = 0.72×Pₒ_norm + 0.28×Rᵥc_norm
3. **Main findings so far**:
  - Stylistic variation appears continuous rather than binary (Currier A/B)
  - Strong modular organization at the quire level (p < 0.000001)
  - Identification of an "extreme α" category (CI < 0.30)
  - CHOR system shows continuous distribution (0-82%), not binary presence/absence

**Open data & code**: All analysis scripts and data are available at: 
You are not allowed to view links. Register or Login to view.

**What I'm looking for**:
- Feedback on the methodological approach
- Suggestions for additional metrics or analyses
- Help completing the analysis of remaining folios
- Discussion of the implications for understanding manuscript structure

I'm here to learn from the community and contribute what I can. All constructive feedback welcome!

Best regards,
aliben

Attached Files

  10000Nordico_00001(1).pdf (Size: 257.1 KB / Downloads: 21)