Voynich Marginalia and Later Hands: Month Names, Notes and f116v

The Voynich Manuscript is not written in one moment.

That sounds obvious once you say it.

Someone made the main book.

Someone later numbered its gatherings.

Someone later numbered its leaves.

Someone wrote month names beside zodiac diagrams.

Someone marked colours on plant drawings.

Someone wrote or erased an ownership name.

Someone appears to have experimented with character tables in a margin.

Someone wrote mixed, partly legible material on the final verso, f116v.

And someone much later added pencil letters to several cosmological panels.

The manuscript is therefore not one voice.

It is an object that accumulated readers.

This matters because the most tempting readable material is often the material most likely to belong to a later layer.

A Romance-looking month name beside Aries feels like a key to Voynichese.

It is not.

A German-looking fragment on f116v feels like the final page finally confessing the language of the manuscript.

It has not done so.

Later writing can tell us who encountered the manuscript, how they interpreted it and what they tried to do with it without telling us what the original text says.

This article is about keeping those historical layers separate enough that each can become useful evidence.

Quick Read

One-sentence answer: the Voynich Manuscript preserves several kinds of writing outside its principal unidentified script—including folio and quire numbers, zodiac month names, colour annotations, ownership marks, marginal notes, later pencil letters and the mixed writing on f116v—and these additions are crucial evidence for the manuscript’s custody and reception history but must be analysed as separate chronological and palaeographic layers rather than treated as automatic keys to Voynichese.

The principal Voynich script is not the only writing on Beinecke MS 408.
Quire marks and folio numbers were added after the main manuscript production and appear to belong to different hands or moments.
The old numbering systems are historically valuable because they preserve evidence of now-missing material.
Later Romance-language month names appear in the centres of the zodiac diagrams from Pisces through Sagittarius.
These month names are in a different hand from the principal Voynich text and should be treated as later annotations.
The month-name spellings have been compared with French, Occitan, Spanish and neighbouring Romance forms; a northern French interpretation has influential support but should not be converted automatically into manuscript provenance.
Individual Latin-alphabet colour annotations appear on some early herbal pages, including a form read as rot (“red” in German).
The faded or erased ownership inscription of Jacobus de Tepenec on f1r belongs to the custody history, not original authorship.
Faded character tables on f1r have been interpreted as a possible later decryption attempt.
Several pages contain marginal sequences, numerals or mixed writing that do not belong straightforwardly to ordinary running Voynich text.
F66r contains partly legible non-Voynich writing near a human figure; readings are disputed.
F116v, the last surviving page, contains a mixture of apparent Latin/German-looking forms and Voynich-script words and has generated many incompatible proposed readings.
The mixed writing on f116v has not produced an accepted decipherment key.
Later pencil letters a, b and c on some cosmological pages may be extremely late additions and are evidence of later handling rather than original structure.
Marginalia are strongest when used to reconstruct readers and manuscript history, not when forced to decode the original system.

In other words:

the manuscript contains both the unknown system and traces left by people who came after it.

Marginalia Are Part of the Object’s Biography

A manuscript is not frozen when the original scribe finishes.

Readers annotate.

Owners sign.

Librarians number.

Students doodle.

Collectors test hypotheses.

Binders add marks needed for assembly.

These later traces can become historically precious.

A name can anchor ownership.

A language can reveal where a later reader came from.

A numbering scheme can preserve evidence about lost leaves.

A failed decryption table can show that the text was already unreadable to an earlier owner.

But marginalia can only do these jobs if we resist absorbing them into the original layer.

A note written a century later does not become fifteenth-century authorial intent simply because it is on the same parchment.

First Separate “Different Hand” From “Different Scribe in the Main Text”

Voynich discussions use the word hand in two different ways.

One problem asks how many scribal hands produced the principal Voynich text.

Lisa Fagin Davis proposed five main hands in influential palaeographic work, while a 2026 critique disputes whether the variation requires five discrete scribes.

That is the internal-scribe problem.

A different problem concerns writing plainly added outside or after the main production:

folio numbers;
quire marks;
month names;
ownership inscriptions;
pencil letters;
mixed marginal writing.

These are later-hand problems.

The two should not be confused.

A five-scribe model of the main text does not imply five later annotators.

A later month-name hand does not become “scribe six”.

Different chronological layers require different questions.

The Quire Numbers Are Later Metadata With Early Historical Value

Quire marks occur mostly near the end of gatherings.

They help identify how the codex was assembled.

They are not written in Voynichese.

They are not original section headings.

Yet they may be among the most useful later additions in the entire manuscript because missing quire numbers align with missing physical material.

This tells us that the numberer knew a fuller version of the book.

The marks therefore belong to reception history and physical reconstruction simultaneously.

They are a perfect example of a later annotation revealing earlier physical history.

But their function is still bounded.

Quire number 8 tells us a gathering identity.

It does not tell us that the gathering was called “Herbal Chapter Eight” by its maker.

The Folio Numbers Are Another Later System

The folio numbers appear in the upper-right area when leaves are folded in.

They run through the codex and skip missing folios.

The style of these numbers is less securely dated than the quire marks and has been assigned broad possible date ranges by different scholars.

The important point is not the exact year.

It is that foliation is a later administrative layer.

This immediately corrects several bad habits.

“Folio 75 begins the biological section” is modern location language.

It is not evidence that the original maker numbered a chapter 75.

Folio numbering helps us communicate.

It should not become the manuscript’s own conceptual voice.

The Zodiac Month Names Are the Most Seductive Readable Annotation

The central zodiac drawings carry later month names written in ordinary alphabetic forms.

They are conventionally read as forms corresponding to March, April, May, June, July, August, September, October, November and December.

The spellings have been compared with several Romance languages and dialects.

A northern French reading has become particularly influential in modern discussion, although localisation remains an interpretive question rather than an inscription saying “written in northern France”.

The month names are important because a later reader apparently recognised the zodiac sequence in conventional calendrical terms.

They are dangerous because readable language feels like a bridge into unreadable language.

The bridge is not automatic.

The month-name writer may have encountered the book decades or centuries after its production.

That person could annotate a manuscript whose main language they did not understand.

A later readable label identifies a later reader more safely than it identifies the original writer.

Why the Month Names Still Matter Enormously

Separating the month names from Voynichese does not make them irrelevant.

They can tell us several things.

A later reader could identify the central zodiac signs.
The signs were legible enough iconographically to support conventional month association.
The manuscript circulated through or reached a reader using a Romance-language naming convention.
The annotation postdates the original diagram and therefore creates a minimum sequence of historical events.
The later reader apparently did not replace or translate the surrounding Voynich labels into the same ordinary script.

The last point is especially interesting.

If the annotator fully understood the main text, why add only month names?

There are many possible answers.

Perhaps the months were useful navigation aids.

Perhaps the annotator did not understand the main script.

Perhaps they understood some parts but not others.

The annotation cannot decide among these.

But it gives us a real later-reader event to explain.

The Month Names Are Not a Provenance Certificate

Suppose the month spellings are correctly localised to northern France.

What follows strongly?

A northern-French-writing person encountered and annotated the manuscript.

What does not follow automatically?

The parchment was produced there.
The Voynich script was created there.
The plants were drawn there.
The original author spoke that dialect.
The book was commissioned there.

Books move.

Readers travel.

Later annotations can belong to a geography entirely different from original manufacture.

This is why provenance should be built as a chain, not inferred backward from one readable marginal hand.

Colour Notes May Preserve Workshop Behaviour

Some early herbal pages contain small ordinary-alphabet annotations near plant parts.

One form has been read as rot, German for red.

This creates a fascinating possibility.

The annotations may be colour instructions.

If so, they could belong to the production workflow rather than much later reception.

A draughtsperson or scribe might indicate which pigment should be applied later.

A painter could follow those notes.

This would create a layered page-production process:

drawing → annotation → painting

But even here, caution is necessary.

The annotations are sparse.

Their exact date relative to the main text is not perfectly established.

And one readable colour word does not automatically identify the geographical language of the whole manuscript.

Still, production annotations are one of the few places where ordinary language may touch the manuscript-making workflow directly.

The Tepenec Name Is Ownership Evidence, Not Authorship

On f1r, an erased or faded inscription has been identified with Jacobus Horčický de Tepenec.

This is among the manuscript’s most important provenance clues.

It anchors the book to a person in the early modern Central European custody story.

But ownership marks are frequently misunderstood.

A name written in a book can mean:

owner;
institution;
recipient;
reader;
cataloguer;
donor.

It does not automatically mean author.

Tepenec therefore belongs in the custody chain.

He should not be converted into the manuscript’s creator merely because his name is physically closer to the object than most other historical names.

The Faded Character Tables on f1r May Record an Earlier Attempt to Solve the Text

The first folio also contains faded or erased character sequences in the margin.

They have been interpreted as a possible decryption attempt by a later owner.

If that interpretation is correct, the marginalia tell us something extraordinary.

Someone before modern Voynich scholarship was already treating the script as a problem to be mapped into another system.

This would place failed decipherment inside the manuscript’s own reception history.

But a failed character table is not a surviving key.

The faded marks may reflect speculation rather than knowledge.

And because they are later, they cannot simply be privileged over modern analysis as though chronological closeness guaranteed access to the original system.

A seventeenth-century guess can still be a guess.

Pencil a, b and c Marks May Be Extremely Late

Several cosmological pages contain small alphabetic letters such as a, b and c written in pencil near corners or panels.

These are plainly unlike the main Voynich script and are thought to have been added by a later owner, possibly very late in the manuscript’s history.

This creates a simple warning.

A modern or early modern annotation can accidentally look like a diagram key.

Someone sees a, b and c around a foldout and imagines original sequence markers.

But if the marks are late, the apparent original architecture is an illusion created by reception.

This is why every mark needs a chronological layer before it gets a semantic job.

F17r Shows Voynich and Ordinary Writing Can Occupy the Same Marginal Space

Near the upper margin of f17r is a largely unreadable comment that includes faint material in or resembling the Voynich script.

Ultraviolet examination has helped reveal parts that are difficult to see normally.

The exact reading remains disputed.

What makes the example valuable is not a proposed translation.

It shows that later or marginal writing can mix graphic systems.

This complicates the idea that every Voynich-shaped word must belong to the original main text and every ordinary-alphabet word must be entirely unrelated.

A later reader may copy Voynich forms while writing an ordinary-language note.

A solver may transcribe a token beside a guess.

The physical mixture can be evidence of interpretation rather than original bilingual composition.

F49v and f76r Show Detached Character Sequences

Some margins contain sequences that sit apart from ordinary text.

On f49v, detached characters appear with Arabic numerals 1–5.

On f76r, another detached sequence occurs.

These features invite the word “key”.

Perhaps they are sequence exercises.

Labels.

Numerical mappings.

Later annotations.

Or original structured writing outside prose.

The safe first conclusion is topological:

not all Voynich-script material is laid out as ordinary continuous paragraph text.

That creates a category worth studying without prematurely assigning a cryptographic function.

F66r Contains One of the Most Tempting Mixed-Language Fragments

Near the bottom of f66r, beside a human figure and other objects, there is writing that has been interpreted as partly German-looking.

Different researchers have proposed different readings.

No reading has become secure enough to carry the language of the whole manuscript.

This is precisely the kind of fragment that creates overconfidence.

A partly familiar word is easier to “read” when we already expect a language.

Then the expected language helps choose ambiguous letters.

The chosen letters appear to confirm the language.

The circle closes.

A better approach records the visible strokes, compares palaeographic alternatives and preserves multiple readings until independent context discriminates them.

Marginal writing is often hard precisely because it is informal.

Informality should lower confidence, not increase imagination.

F116v Is the Most Famous Trap at the End of the Book

The final surviving verso, f116v, is unlike ordinary Voynich pages.

It contains several short lines in a mixture of forms that appear partly Latin-alphabet, partly German- or Latin-like, alongside words written in the Voynich script and small marginal drawings.

Because it sits at the very end of the codex, the page has been treated as a possible key, colophon, ownership note, medical instruction, charm, mnemonic or scribal test.

It has generated many proposed readings.

They disagree substantially.

That disagreement is evidence about our confidence.

If the same strokes can be segmented into different ordinary-language phrases depending on expectation, the page has not yet supplied a stable decipherment anchor.

The strongest safe conclusion is modest:

f116v preserves mixed writing produced in a historical context close enough to the manuscript to be extremely interesting, but its exact reading, authorial relationship and function remain unresolved.

The last page is not necessarily the key just because modern books often put answers at the end.

Why f116v Does Not Automatically Tell Us the Language of Voynichese

Suppose f116v contains German.

Several possibilities remain.

The main text encodes German.
A German-speaking later reader wrote on the page without understanding the main text.
A multilingual owner added a note.
A scribe copied a formula from another source.
The page combines several chronological layers.

The existence of German-like material cannot distinguish these by itself.

To make f116v a language key, we would need a stable mapping between its ordinary-language forms and the Voynich-script forms on the same page.

That mapping would then need to work across ordinary Voynich text elsewhere.

No such accepted bridge exists.

This is a classic case where contact is not equivalence.

Two scripts sharing a page is weaker evidence than one translating the other.

A Bilingual Key Has a Much Higher Standard

A true bilingual or biscriptal key is one of the great prizes in decipherment.

The Rosetta Stone is famous because related content appears in multiple writing systems.

But co-presence is not enough.

To use f116v similarly, we would need evidence that:

the ordinary-language text and Voynich-script words refer to the same content;
their alignment can be established;
a repeated Voynich form maps consistently to a repeated ordinary-language element;
the mapping predicts forms elsewhere.

Without alignment, two scripts on one page may simply be two scripts on one page.

The historical proximity is exciting.

The cryptographic relationship remains unproven.

Later Readers Can Misunderstand the Manuscript Too

There is a subtle bias in historical research.

We assume people closer in time to an object understood it better.

Sometimes they did.

Sometimes the knowledge chain had already broken.

By the seventeenth century, Georg Baresch was already seeking help to understand the manuscript.

Marci sent it to Kircher because he believed Kircher might decipher it.

This tells us the opacity is old.

Therefore a later annotation may record a misunderstanding, guess or partial recognition.

The month-name annotator could correctly recognise zodiac signs while understanding none of the surrounding text.

A decryption-table writer could be completely wrong.

Historical age raises interest.

It does not grant infallibility.

Annotations Can Reveal Function Even Without Decoding

Suppose the zodiac month names are later.

They still show that a later user wanted calendar navigation.

Suppose colour notes belong to production.

They reveal division of labour or staged painting.

Suppose character tables are failed decoding attempts.

They reveal how a historical reader approached the script.

Suppose quire marks are administrative.

They reveal assembly management.

Meaning is not the only information marginalia can carry.

Use can survive after content is forgotten.

Reception can survive after original authorship disappears.

This is why marginalia belong in the core research map rather than being treated as noise around the “real” manuscript.

Palaeography Is the Bridge Between Stroke and Historical Layer

How do we decide that two writings belong to different hands or periods?

Not by mood.

Palaeography examines:

letter formation;
stroke order and direction;
pen angle;
spacing;
abbreviations;
numeral forms;
ligatures;
comparative dated hands.

Material analysis can add ink information.

Physical position can add sequence.

Together these can distinguish a later annotation from the main production more reliably than simply saying “it looks different”.

This is particularly important on f116v, where informal writing and damage make confident letter identification difficult.

A decipherment claim built on one ambiguous marginal letter should carry the uncertainty of that letter all the way through the argument.

UV and Multispectral Imaging Can Reveal Writing—Not Automatically Its Meaning

Faded or erased writing can become more visible under ultraviolet or other specialised imaging.

The Tepenec inscription is a famous example of physical enhancement helping historical interpretation.

But better visibility creates a new temptation.

Once a stroke appears, people want it to become a letter.

Once a letter appears, people want it to become a word.

Once a word appears, people want it to become a language.

Imaging improves the observation layer.

Palaeography interprets the strokes.

Historical context interprets the word.

Those are separate inferential stages.

Technology can recover signal without recovering semantics.

What the Later Hands Can Tell Us

The manuscript continued to be handled and interpreted after its original production.
Later readers recognised at least some zodiac iconography.
The codex was numbered in more than one later administrative layer.
Some now-missing leaves were known to later numberers.
At least one later owner’s name became attached to the manuscript.
Some ordinary-language writing may preserve production or reception vocabulary.
Later readers may have tried to decode or organise the manuscript.
The writing system was already difficult for historical owners by the seventeenth century.
Multiple scripts and languages can coexist physically without forming a bilingual key.

These are substantial contributions to the manuscript’s biography.

What the Later Hands Cannot Tell Us Alone

The language of the principal Voynich text.
The identity of the original author.
The original place of production.
The exact date of every annotation without independent palaeographic or material evidence.
Whether f116v is a key, note, charm or scribal exercise.
Whether month-name language matches original manuscript language.
Whether colour notes were written by the original team or a later user in every case.
Whether faded character tables preserve correct decipherment knowledge.
The semantic function of every marginal Voynich-script sequence.

Later writing narrows history more readily than it unlocks language.

A Better Marginalia Analysis

1. Locate the mark physically

Margin, diagram centre, binding edge, folio corner or page body?

2. Describe the script before the language

Latin alphabet, Voynich script, numerals, pencil, ink, mixed writing?

3. Establish the hand and layer

Main production, possible workshop annotation, later owner, librarian or modern collector?

4. Preserve uncertain readings

Do not force ambiguous strokes into the expected language.

5. Ask what the annotation does

Numbering, ownership, navigation, colour instruction, translation attempt, comment or unknown?

6. Only then ask what it means

Function can sometimes be established before exact language.

7. Do not project the later layer backward

A French month annotation does not make the main manuscript French.

8. Require a predictive bridge before using marginalia as a decipherment key

A proposed mapping must work beyond the annotated page.

This sequence makes marginalia historical evidence rather than semantic contamination.

What Would Count as a Real Marginalia Breakthrough?

Imagine multispectral imaging reveals a previously unreadable annotation on f1r.

Palaeographers independently agree on the letters.

The phrase names a known owner, workshop or location.

Documentary records independently place that person with the manuscript.

That would strengthen provenance enormously.

Or imagine f116v yields a stable ordinary-language phrase directly aligned with one Voynich-script phrase.

The mapping predicts multiple words elsewhere.

Independent researchers reproduce it.

That would be a genuine decipherment bridge.

The standard is high because the historical consequence is high.

A breakthrough does not mean “I can make the letters look like a sentence”.

It means the reading survives independent palaeography and then predicts beyond itself.

Primary School: Who Wrote Which Note?

Take a schoolbook.

Imagine the printed text belongs to the author.

A student writes a name inside the cover.

A teacher circles one sentence.

A librarian adds a barcode.

Ask a child:

Are all these marks part of the original book?

No.

But all tell us something about the book’s history.

That is exactly how to introduce Voynich marginalia.

Lower Secondary: Separate Author, Reader and Cataloguer

Give students three historical marks:

a zodiac month name;
a folio number;
the Tepenec ownership inscription.

Ask what each most likely tells us.

Month name → later reader interpretation.

Folio number → later book administration.

Ownership name → custody.

Now ask which tells us the original language.

None directly.

The learner discovers that a historical object can accumulate multiple information layers with different evidentiary jobs.

Upper Secondary: A Readable Word Can Be the Wrong Layer

Give students this proposition:

The month names are French, therefore Voynichese is French.

Ask them to identify the missing premise.

The argument assumes the month-name writer is the main-text writer.

But the hands differ.

Therefore the conclusion does not follow.

This is a clean lesson in temporal confounding.

Two variables coexist on the same page but belong to different historical moments.

Spatial proximity is not chronological identity.

JC and Adult Readers: Treat the Manuscript as a Stratified Site

At a higher level, marginalia make the manuscript resemble an archaeological site.

Layer A: parchment manufacture.

Layer B: principal drawing and writing.

Layer C: painting or production annotations.

Layer D: quire numbering.

Layer E: foliation.

Layer F: ownership and reader notes.

Layer G: modern conservation, cataloguing and pencil additions.

The exact number and order of layers varies by page.

The analytical principle is stratigraphy:

interpret each mark first in the historical layer that produced it.

Only then ask how layers relate.

This prevents later understanding from being projected backward into original authorship.

A Parent and Teacher Guide

Marginalia are an unusually accessible way to teach source criticism.

Ask who made the mark. Original scribe, painter, owner, reader, librarian or modern researcher?
Ask when. Same production moment or later?
Ask what job it performs. Navigation, ownership, interpretation, colour, numbering, attempted translation?
Keep readable language in its own layer. A later French or German word does not automatically reveal original language.
Preserve disputed readings. Informal handwriting is often ambiguous.
Use marginalia to reconstruct reception. What did later readers understand or fail to understand?
Demand prediction before calling something a key. A real key must work elsewhere.

The method is useful for annotated books, historical photographs, scientific notebooks, social-media screenshots and almost any source that has been touched by more than one person.

Reader Checklist: Before You Use Voynich Marginalia as a Decoding Key

Is the writing part of the main Voynich script or an additional hand?
What is the physical position of the annotation?
What palaeographic evidence dates or localises it?
Is the proposed reading stable across independent readers?
Does the writing belong to production, ownership, cataloguing or later interpretation?
Are month names being projected backward into the main text?
Is a readable word being used to infer manuscript provenance too strongly?
Could the annotation come from someone who did not understand Voynichese?
Does mixed writing on f116v align one-to-one across scripts?
Does the proposed mapping predict ordinary Voynich text elsewhere?
Are late pencil or foliation marks being mistaken for original diagram structure?
What remains true if the marginal reading is wrong?

The final question protects the manuscript from interpretations built on one unstable letter.

Frequently Asked Questions

Are the zodiac month names written in Voynichese?

No. They are ordinary-alphabet forms written in a later hand and are generally treated as later annotations.

What language are the month names?

They are clearly Romance-language forms. French, Occitan, Spanish and related possibilities have been discussed; northern French has influential modern support, but localisation remains debated.

Do the month names prove where the Voynich Manuscript was made?

No. They more safely localise or characterise a later annotator. The manuscript could have travelled before receiving them.

What is the Tepenec inscription?

It is a faded or erased ownership inscription on f1r associated with Jacobus Horčický de Tepenec. It is important provenance evidence, not authorship evidence.

What are the colour annotations?

Small ordinary-alphabet marks on some herbal pages may be colour instructions, including a form read as rot. Their exact relationship to the production workflow remains a research question.

Is f116v the key to the manuscript?

No accepted evidence establishes it as a key. It contains mixed, difficult writing that has generated many incompatible readings.

Does f116v prove the main text is German?

No. German-looking forms on a later or mixed page do not establish that the principal unidentified text encodes German. A predictive mapping across scripts would be required.

Are folio and quire numbers original?

They are later than the principal text and illustrations and appear to belong to separate numbering layers.

Why are later hands important?

They preserve custody, navigation, production and reader-response evidence. They help reconstruct what happened to the manuscript after it was made, including how quickly or completely its original knowledge context disappeared.

Research and Further Reading

The Final Idea

The marginalia of the Voynich Manuscript are voices standing around a silence.

One reader says March.

Another numbers a leaf.

Another leaves a name.

Another perhaps writes “red”.

Another tries to map characters.

Another leaves the strange mixed lines on the final page.

None has yet restored the lost original voice.

But they tell us something just as human.

People have been trying to understand this book for centuries.

Some recognised pieces.

Some organised it.

Some owned it.

Some probably misunderstood it.

The marks they left behind become valuable only when we let each voice belong to its own time.

The page is shared. The historical moments are not.

Voynich | Everything eduKate Knows and Tested | Marginalia and the Later Hands