Voynich Pictures: What the Images Can and Cannot Tell Us

The easiest way to become certain about the Voynich Manuscript is to look at a picture.

The easiest way to become wrong is the same.

A plant looks familiar.

A root looks like an animal.

A circular diagram looks astronomical.

A human figure stands in a green pool, and the mind supplies a bath.

A tall container looks like an apothecary jar, and suddenly the entire page becomes pharmaceutical.

A cluster of towers appears on a foldout, and somebody sees a city.

None of these interpretations is foolish.

That is exactly why the problem is hard.

Good interpretations are usually plausible before they are proven.

The work is learning how not to confuse those two states.

A picture can be evidence before we know what it depicts.

That sentence is the foundation of this article.

Quick Read

One-sentence answer: Voynich images reveal strong visual organisation, recurrence, categorisation and page-level design, but resemblance alone cannot identify a plant, function, process, place, manuscript lineage or semantic meaning.

The conventional “herbal”, “astronomical/astrological”, “biological/balneological”, “cosmological”, “pharmaceutical” and “recipe” labels are useful modern descriptions, not decoded medieval chapter headings.
A recognisable zodiac sign is stronger evidence than a vague resemblance to an unknown plant because the iconographic category is more constrained.
Even when a category is recognisable, its function in this particular manuscript remains a separate question.
Plant images may be naturalistic, stylised, composite, diagrammatic, copied, transformed or intentionally information-dense.
A root resembling an animal or object does not prove symbolic encoding.
Human figures and liquid-like regions do not by themselves establish anatomy, bathing, gynaecology, treatment, alchemy or ritual.
Container-like forms do not establish a universal preparation code.
Stars can mark entries, identities, categories or celestial information; shape alone does not select among those possibilities.
The Rosettes foldout is clearly relational and geometrically organised, but “map”, “city” and “cosmos” remain interpretations.
Visual recurrence inside the manuscript can support internal reuse or shared design families without proving semantic identity.
The Carrara Herbal is valuable because it demonstrates what a highly resolved illustrated medical manuscript near 1400 could look like; it does not prove Voynich provenance.
Masson 116 and Sloane MS 4016 are especially useful because they show a documented visual transmission relationship, giving us a control for what copying can preserve and transform.
Egerton MS 747 and BnF Latin 6823 show that plants, substances, remedies, measurement, calendars, animals and other practical knowledge could coexist inside medieval manuscript worlds.
The strongest visual method is to move from geometry and morphology to comparison, then interpretation—not start with a name and search for confirming shapes.

The pictures are not useless because we cannot name them.

They become more useful when we stop demanding names too early.

First Describe the Line That Is Actually There

Take an unfamiliar Voynich plant.

Instead of asking:

What species is this?

begin with:

How many major leaves?
Are the margins smooth, serrated or lobed?
Are the leaves opposite, alternate or arranged another way?
Is there one stem or a branching structure?
What is the root geometry?
Are flowers or seed-like structures present?
Which parts are coloured?
Which parts repeat elsewhere?
Which components are attached and which are merely nearby?

These questions feel humble.

They are powerful because different observers can check them.

“Looks like a sunflower” depends heavily on expectation.

“Has five radiating lobes around a central circular form” is closer to shared evidence.

Description creates a common floor before interpretation creates disagreement.

A Name Compresses More Than You Think

Suppose we call a drawing “a jar”.

Immediately the word imports a function.

Jars contain things.

They store.

They preserve.

They sit in kitchens, pharmacies and workshops.

Now the adjacent plant fragment starts looking like an ingredient.

Now the nearby text starts looking like a recipe.

One noun has quietly built an entire causal chain.

Sometimes that chain may be right.

But the visual evidence initially established something more modest:

a repeated container-like form with a neck, body and decorative geometry.

The longer description sounds clumsy.

It also leaves alternative functions alive.

Recognising a Category Is Easier Than Recovering Its Job

Some Voynich imagery is genuinely recognisable.

Zodiac signs provide good examples.

A ram-like Aries figure, a bull-like Taurus figure and other conventional zodiac imagery place at least some diagrams inside a recognisable historical iconographic family.

This is stronger than saying an ambiguous root “looks like” an animal.

But even strong category recognition does not recover use.

A zodiac can support:

astrological interpretation;
calendrical organisation;
medical timing;
birth or personal records;
mnemonic arrangement;
illustrated reference;
or another practice known to its users.

The image can identify the symbolic vocabulary more confidently than the operational sentence in which that vocabulary was used.

Recognition answers “what family does this image belong to?”
Function asks “what was this family doing here?”

Why the “Herbal” Label Is Useful and Dangerous

A large portion of the manuscript contains one prominent plant-like drawing per page with surrounding or adjacent text.

Calling those pages “herbal” is sensible as orientation.

Known herbals also organise plant images and text.

The danger begins when the navigation label becomes a decoded function.

A plant image can participate in:

identification;
medical substance knowledge;
agriculture;
natural history;
symbolic teaching;
mnemonic systems;
recipe indexing;
collection records;
or a hybrid practice.

Until the text is understood, “herbal” should mean:

pages visually dominated by plant-like forms.

That keeps the category useful without pretending to know the medieval heading.

A Plant Drawing Is Not Necessarily a Portrait

Modern field guides train us to expect botanical pictures to behave like portraits.

The goal is species recognition.

Leaves, flowers, fruits and roots should support identification.

Historical manuscript images can have other priorities.

A copied image can preserve the memory of an exemplar more strongly than the appearance of a living plant.

A diagnostic feature can be exaggerated.

A useful root may be enlarged.

Different growth stages can be compressed into one image.

Several traditions can be combined.

A picture can become diagrammatic.

Therefore the question “what plant does this look like?” is only one possible way to interrogate the drawing.

Sometimes “what information has been exaggerated?” may be more productive.

Composite Plants Are an Explanation Family, Not a Licence

Because many Voynich plants resist confident identification, researchers often propose that they are composites.

Leaf from one species.

Flower from another.

Root transformed for symbolic or medicinal meaning.

This can explain almost anything if used carelessly.

That is the problem.

If every mismatch becomes evidence of compositing, the hypothesis can never lose.

A useful composite hypothesis should predict:

which components are likely to be borrowed;
whether the same components recur elsewhere;
how components are combined;
whether known manuscript traditions use similar assembly;
and whether the proposed combination serves a stable function.

“Composite” is not a magic word for a plant we cannot identify.

It is a testable model of how an image was made.

Roots Are Where Pareidolia Goes to Work

Voynich roots are visually generous.

They curl.

Branch.

Bulge.

Split.

Mirror.

Sometimes they look like legs, claws, heads, vessels, letters, tools or animals.

Humans are extremely good at finding familiar forms in ambiguous geometry. We see faces in clouds and animals in rocks.

Therefore a root resemblance should be treated as a candidate feature, not a decoded symbol.

The stronger question is:

Does the same supposedly symbolic shape recur systematically in the same structural role?

If yes, the symbolic hypothesis becomes more interesting.

If not, the resemblance may belong more to the observer than to the manuscript.

The Carrara Herbal Makes the “They Could Not Draw Plants” Explanation Weaker

British Library Egerton MS 2020—the Carrara Herbal—was made around 1390–1404 and contains exceptionally detailed plant imagery alongside an Italian medical text associated with Serapion the Younger.

Its historical value for Voynich research is not that its plants secretly decode Voynich plants.

Its value is that it demonstrates capability.

Near the Voynich parchment period, sophisticated manuscript makers in northern Italy could produce plant images with high observational resolution.

Therefore a universal explanation such as:

medieval artists simply could not draw plants accurately

is too crude.

Voynich distortion may still come from poor copying, inherited schemata, limited access to specimens or an unskilled artist.

But deliberate stylisation, diagrammatic compression and alternative visual functions must remain live as well.

Carrara broadens the plausible production world.

It does not identify the Voynich workshop.

Capability Proof Is Not Provenance Proof

This distinction deserves its own section because it rescues comparative research from two opposite mistakes.

Mistake one:

The comparison does not prove Voynich came from Padua, so the comparison is useless.

No.

A comparison can show what was technologically, artistically or intellectually possible.

Mistake two:

The comparison is striking, so Voynich must come from the same place.

Also no.

The correct relationship is narrower:

Known manuscript X demonstrates that practice Y existed in a relevant historical environment.

That sentence can transform our interpretation without inventing ancestry.

Masson 116 and Sloane 4016 Give Us Something Rarer: a Known Transmission Relationship

MS Masson 116 is an image-rich Tractatus de herbis manuscript with simple Latin captions and no continuous prose.

British Library Sloane MS 4016 is catalogued as a copy of Masson 116.

This pair is extremely useful because the relationship is not invented from visual resemblance alone.

We can study what happens when an image tradition is actually copied.

Some components persist.

Some proportions change.

Details can simplify.

Page framing can change.

Captions can be re-hosted.

The copy remains recognisably related across many units.

This gives us a standard for Voynich claims.

If someone proposes that one Voynich plant derives from a known herbal, look for a family of systematic correspondences, not one exciting leaf shape.

A Visual Family Can Be Real Without Being a Copy

Two images can resemble one another for several reasons.

One copied the other.
Both copied a common exemplar.
They belong to the same artistic tradition.
They solve the same representational problem independently.
They share generic iconography.
They use common geometry.
The resemblance is accidental.
The observer has selected only the matching features.

These explanations require different levels of evidence.

“Same tradition” is a weaker claim than “same exemplar”.

“Same exemplar” is a weaker claim than “same workshop”.

“Same workshop” is a weaker claim than “same owner”.

A useful hierarchy is:

resemblance → shared motif → shared tradition → transmission → workshop → provenance → ownership

Every arrow requires new evidence.

You cannot jump from the first node to the last because the pictures feel persuasive.

The “Pharmaceutical” Pages Might Be Retrieval Pages

Pages with repeated container-like objects and detached plant parts are conventionally called pharmaceutical.

That label makes sense if the visual objects are interpreted as jars and medicinal ingredients.

But consider another level of description.

These pages contain:

repeated vessel-like forms;
smaller plant components rather than many whole plants;
short textual units associated with image components;
visual repetition and categorisation.

Such architecture could support preparation.

It could also support indexing, classification, storage identification, ingredient retrieval or another reference task.

Without readable text, it is safer to say that the page morphology becomes more component-oriented and catalogue-like than the large-plant pages.

That difference is already meaningful.

We do not need to declare the exact warehouse before noticing the shelves.

Internal Visual Reuse Is More Interesting Than External Look-Alikes

When a plant component in one Voynich region resembles a component elsewhere in the same manuscript, the comparison has one advantage over an external match.

Both objects are unquestionably inside the same surviving codex.

That removes the need to prove an external transmission route.

But semantic identity still does not follow automatically.

A repeated root shape might mean:

the same plant;
the same visual template;
the same component class;
the same artist’s habit;
or a generic convention used for several objects.

The internal match becomes stronger when multiple independent features recur together.

Root geometry.

Leaf shape.

Colour.

Label relationship.

Page role.

One similar curve is weak.

A bundle of coordinated similarities is much more interesting.

A Repeated Picture Can Be the Same Object—or the Same Job

This distinction matters enormously.

Imagine a red circle appears on many modern diagrams.

Sometimes it represents the same object.

Sometimes red circles merely mean “warning”.

Voynich recurrence has the same ambiguity.

A repeated plant fragment could be a repeated ingredient.

Or it could be a repeated class marker.

A repeated star could represent the same celestial category.

Or it could simply delimit entries.

A repeated vessel could identify one substance.

Or it could indicate a storage or retrieval class.

Therefore recurrence tells us:

the maker considered this form reusable.

What was being reused—identity, category, function or style—remains the next question.

The Containers Did Not Survive as a Universal Process Code

One attractive idea is that different container shapes encode different operations.

One jar means powder.

Another means infusion.

Another means distillation.

Another means storage.

This is historically conceivable.

Apothecary practice used vessels for different purposes, and visual systems can encode process through equipment.

The problem is not plausibility.

The problem is predictive evidence.

A universal container code should create repeatable associations.

The same vessel class should align with the same independent textual or visual behaviour strongly enough to predict unseen cases.

Our retained work did not justify promoting container morphology to that level of decoded process semantics.

The container patterns remain worth studying.

The universal key does not.

Human Figures Create a Different Kind of Ambiguity

In the group of pages often called biological or balneological, human figures interact with pools, tubes, vessels and flowing forms.

The figures are recognisably human.

The systems around them are not recognisably ordinary anatomy.

That gap creates interpretation.

bathing;
medical treatment;
gynaecology;
reproduction;
body fluids;
thermal waters;
astrological medicine;
alchemical process;
metaphorical diagrams;
or an unfamiliar technical system.

Some of these have historical precedents.

That makes them candidates, not conclusions.

The safer visual observations concern:

how figures are placed;
what they touch;
which conduits connect;
whether liquid-like areas share boundaries;
which motifs recur;
and how text is distributed around the scene.

Those relationships survive even while the biological story remains uncertain.

A Tube Is Not Automatically Plumbing

Whenever a line encloses a channel, modern eyes see pipes.

Pipes imply flow.

Flow implies material moving from A to B.

And suddenly an entire hydraulic machine appears.

But a diagrammatic connector can mean many things.

physical conduit;
anatomical channel;
conceptual relation;
boundary;
sequence;
genealogy;
movement;
containment;
or decorative framing.

The right first question is topological:

What connects to what?

Only later comes the semantic question:

What travels through the connection, if anything?

Geometry should not be forced to confess content it has not yet established.

The Zodiac Does Not Automatically Become the Book’s Master Clock

Medieval medicine and astrology often interacted.

This historical fact makes an integrated medical-timing interpretation of Voynich tempting.

Plants are gathered at certain times.

Remedies are administered under certain celestial conditions.

Bodies relate to zodiacal correspondences.

A plausible medieval world exists in which these things connect.

The Voynich Manuscript still has to prove that it instantiated that exact world.

The zodiac imagery establishes zodiac imagery.

It does not automatically establish:

a medical calendar;
a harvesting schedule;
a treatment timing system;
or a control layer governing all other page families.

Those higher-order relationships require independent cross-manuscript evidence.

Historical plausibility opens the door.

It does not walk the theory through it.

Stars May Be Celestial—and Still Function as List Markers

The later pages often called “recipes” contain short paragraphs preceded by star-like forms.

The visual word “star” immediately invites astronomy.

But symbols can have two lives.

A star can depict a star.

A star can also mark an entry.

Modern writing does this constantly.

An arrow can represent direction in a map and act as a bullet in a slide.

A heart can represent an organ or a “like” button.

Therefore Voynich star-like forms need to be studied by behaviour.

Where do they occur?

Do they consistently introduce entries?

Do different star morphologies correlate with different textual patterns?

Does the same star type appear in clearly diagrammatic celestial contexts?

Function comes from distribution, not from silhouette alone.

The Rosettes Foldout Is Definitely a Relationship Diagram of Some Kind

The Rosettes foldout is one of the strongest examples of visual structure surviving semantic uncertainty.

Several large circular regions occupy a connected expanded surface.

Paths, bridges, channels or bands appear to connect parts of the design.

Architectural-looking and radiating motifs create local centres.

The whole is not arranged like unrelated illustrations placed opportunistically on one page.

It has relational geometry.

That is strong.

What kind of relationship?

That is weaker.

geographical;
cosmological;
elemental;
processual;
mnemonic;
architectural;
or another abstract system.

The foldout tells us that relations matter.

It does not yet name the relations.

“It Looks Like a City” Needs City-Specific Evidence

Architecture-like forms are especially dangerous because cities provide such rich visual vocabulary.

Towers.

Walls.

Gates.

Roads.

Courtyards.

Once a viewer sees a city, every connector becomes a route.

A strong city identification would need more than architectural flavour.

We would want stable correspondences:

topographic relationships;
distinctive landmarks;
orientation;
documented visual conventions;
consistent labels if deciphered;
and ideally independent provenance evidence.

Without those, “city-like” is a morphology observation.

It is not an address.

Colour Should Be Treated as a Variable, Not a Caption

A pool is green.

Does green mean water?

A root is red-brown.

Does red mean heat?

A star is painted blue.

Does blue define a class?

Any of these could be true.

The way to strengthen the claim is distribution.

If green carries a semantic role, it should correlate with something else.

Page type.

Textual profile.

Object class.

Position.

Repeated motif.

Until then, colour is an observed variable with possible functions, not a decoded semantic legend.

Scale Is Not Innocent

Voynich plant parts are not necessarily drawn to natural scale.

A root can dominate the page.

A flower can become disproportionately small.

A human figure can be tiny beside an enormous pipe-like structure.

Modern viewers often interpret scale literally unless warned otherwise.

Historical diagrams frequently use scale rhetorically or functionally.

Important things become large.

Repeated things become schematic.

Spatial relations are rearranged to fit the page.

Therefore disproportion may indicate:

lack of naturalism;
copying degradation;
symbolic emphasis;
diagnostic emphasis;
or simple layout pressure.

Again, several mechanisms can produce the same visual fact.

Orientation Can Manufacture a Match

Digital images make rotation effortless.

Turn a root 90 degrees and it resembles a fish.

Mirror it and it resembles a letter.

Crop it and it resembles architecture.

These manipulations can be useful exploratory tools.

They also multiply the number of possible accidental matches.

If a proposed identification only becomes convincing after arbitrary rotation, reflection and selective cropping, the evidentiary bar should rise.

A historical user encountered the image in a particular orientation within a physical page.

Our comparison should remember that embodied context.

One Spectacular Match Is Weaker Than Twenty Boring Matches

This may be the single best visual rule for the Voynich Manuscript.

A spectacular match attracts attention.

One root looks exactly like a lion.

One tower looks exactly like a famous castle.

One flower looks exactly like a New World species.

But if the theory predicts an historical connection, it should usually generate a family of correspondences.

Twenty ordinary features matching the same source tradition can be stronger than one breathtaking resemblance.

Why?

Because independent coincidences become harder to explain away.

Good comparison is cumulative.

It does not live or die on the most photogenic example.

The Best External Comparators Do Different Jobs

There is no single “closest medieval manuscript” that answers every Voynich question.

Different comparators are useful for different dimensions.

Carrara Herbal: high-resolution illustrated medicinal plants, vernacular learned text, northern Italian courtly context near 1400.
Masson 116: image-dominant natural-history corpus with simple labels and no continuous prose.
Sloane MS 4016: documented copy relationship to Masson 116, useful for studying visual transmission.
Egerton MS 747: herbal material alongside antidotary, dosage, substitutions, synonyms, measurements and calendrical/forecasting matter.
BnF Latin 6823, associated with Manfredus de Monte Imperiali: plants, animals, fishes, antidotary material and synonym traditions within a learned medieval medical/natural-history environment.

None is “the source of Voynich” simply because it is useful.

Together they map the surrounding possibility space.

That is a more defensible and ultimately more powerful job.

Egerton MS 747 Shows That Mixed Visual Worlds Need Not Be Strange

One reason Voynich looks so exotic is that modern readers expect subjects to be separated.

Botany is botany.

Pharmacy is pharmacy.

Calendars are calendars.

Egerton MS 747 reminds us that a medieval practical codex could combine plant imagery, antidotary material, substitutions, weights, measures, synonyms and lunar or forecasting material.

Therefore plants + remedies + timing is not an impossible medieval combination.

But notice the next step.

Because such a combination existed elsewhere, it becomes historically plausible for Voynich.

It does not become demonstrated for Voynich.

Comparison expands possibility.

Internal evidence selects among possibilities.

BnF Latin 6823 Reminds Us That “Natural History” Was Wider Than Plants

The manuscript tradition associated with Manfredus de Monte Imperiali includes plants alongside animals, fishes, antidotary material and medicinal synonym traditions.

This matters because modern readers often search for one subject label that explains the whole Voynich visual world.

Medieval learned categories could be broader, layered and practical.

A substance might be classified by appearance, source, medicinal use, synonym, preparation or inherited authority.

An illustrated codex could therefore organise a knowledge environment rather than a modern school subject.

This possibility is useful when reading Voynich.

But it is still a possibility.

The image does not become medicinal because another manuscript’s image is medicinal.

The Clean Plant → Jar → Body → Zodiac → Recipe Pipeline Was Too Clean

It is worth preserving one of the most useful failed ideas because it teaches how whole-manuscript reasoning should work.

The attractive story was:

identify plant → prepare substance → apply to body → choose celestial timing → record recipe.

Every individual arrow has some historical plausibility.

That makes the full story feel powerful.

Then the manuscript is asked to behave like the story.

The boundaries do not align cleanly enough.

Current page order cannot safely carry the whole sequence.

Container forms do not provide a decoded process legend.

Zodiac imagery does not independently establish master timing control.

Textual regimes and visual categories overlap without collapsing into the pipeline stages.

So the total model was demoted.

The individual relationships remain research questions.

This is what good failure looks like.

A Failed Grand Theory Can Leave Strong Local Questions Behind

Rejecting the full pipeline does not mean rejecting every connection inside it.

The right move is decomposition.

Do whole-plant pages and fragment pages share internal motifs?
Do particular vessel morphologies correlate with particular textual forms?
Do human-figure pages share vocabulary with zodiac pages?
Do star-marked entries connect statistically to diagram labels?
Do specific visual motifs recur across physically related bifolia?

Each question can be tested independently.

Some may survive even when the total story fails.

This is how a big hypothesis becomes useful after it dies.

It decomposes into smaller questions with sharper evidentiary boundaries.

Images and Text Should Cross-Validate, Not Co-Create the Same Guess

Suppose a proposed textual decipherment says a label means “rose”.

The picture looks vaguely rose-like.

Excellent?

Maybe.

But if the solver chose “rose” because the picture looked like a rose, the agreement is not independent.

For genuine cross-validation, one evidence stream should make a prediction about the other.

For example:

Derive a textual rule from non-plant material.
Freeze the rule.
Translate an unseen plant label.
Ask whether the predicted plant identity or property matches independently.

The same applies in reverse.

A visual classification should predict textual behaviour on pages not used to define the class.

Independent agreement is far more valuable than one clue teaching another clue what answer to give.

What the Pictures Can Tell Us With High Confidence

The manuscript contains several recurring visual families.
Plant-like forms dominate many pages.
Recognisable zodiac iconography appears in circular arrangements.
Human figures interact with bounded, connected and liquid-like forms in another major visual family.
Large circular and cosmography-like diagrams require specialised page architecture, including foldouts.
Container-like forms and detached plant components recur in a distinct page morphology.
Star-like forms recur both in diagrammatic and entry-like contexts.
Visual geometry is often relational rather than merely decorative.
Some motifs recur across non-adjacent pages.
Image and text are spatially coordinated.
Colour, scale, attachment and placement vary systematically enough to deserve measurement.

Those are substantial observations.

None requires a decoded word.

What the Pictures Cannot Tell Us Alone

The exact species identity of most plants.
Whether unusual plants are composites.
Whether roots encode symbolic objects.
Whether a container-like form is a pharmaceutical vessel.
Whether vessel morphology encodes preparation method.
Whether green pools represent water.
Whether human figures are patients, bathers, anatomical entities, allegories or another class.
Whether zodiac imagery controls medical timing.
Whether star-like bullets are celestial objects in every context.
Whether the Rosettes foldout depicts a city, landscape or cosmos.
Whether a visual similarity proves copying.
Whether a comparison proves Padua, Carrara ownership or another provenance.
What the accompanying text says.

The pictures narrow meaning.

They do not finish it.

A Better Visual Comparison Ladder

When comparing one Voynich image with another image, move through the following ladder.

1. Geometry

What are the measurable shapes, angles, counts and proportions?

2. Morphology

How are components constructed and attached?

3. Layout role

Where does the object sit relative to text and other images?

4. Recurrence

Does the same feature appear elsewhere under similar conditions?

5. External analogy

Do known manuscripts use a comparable form?

6. Historical relationship

Is there evidence of shared tradition, exemplar, workshop, place or ownership?

7. Semantics

Only now ask what the form means.

Most internet comparison reverses the ladder.

It begins with semantics:

This is a castle.

Then it searches backwards for geometry that agrees.

The ladder makes that much harder.

Primary School: Make Two Columns—See and Think

For a young learner, this entire article can be reduced to a sheet of paper divided in half.

Left column:

I can see…

Right column:

I think it might be…

Under “see”:

six leaves;
blue flower;
two branching roots;
three lines of writing.

Under “think”:

maybe medicinal;
maybe copied from another picture;
maybe invented;
maybe one real plant drawn strangely.

That child has learned to keep evidence and inference in separate containers.

Many adults never learn this reliably.

Lower Secondary: Build a Visual Feature Table

For Secondary 1 and Secondary 2, compare three images without naming them.

Record:

number of leaves;
leaf margin;
root branching;
flower geometry;
colour;
orientation;
page position;
label position.

Then ask:

Which two are most similar?

Why?

Now reveal the proposed identities from different researchers.

The student sees how naming can change what features feel important.

That is a practical lesson in observer bias.

Upper Secondary: Test a Famous Image Match

Choose any published claim that a Voynich image matches a known plant, city or manuscript.

Do not begin by deciding whether the claim is ridiculous or brilliant.

Score the reasoning qualitatively.

How many independent features match?
How many important features mismatch?
Was the candidate chosen before or after seeing the image?
Are rotation, mirroring or selective cropping required?
Does a wider family of images match?
Is there historical evidence connecting the two traditions?
Does the proposed match predict text or other imagery?

The student is now doing something more sophisticated than “spot the difference”.

They are auditing an argument.

JC and Adult Readers: Ask Whether the Image Feature Is Identifying or Merely Compatible

This is the mature version of the visual problem.

A feature can be compatible with many causes.

A circular diagram is compatible with astronomy.

It is also compatible with calendars, cosmology, classification and abstract relational diagrams.

A jar-like object is compatible with pharmacy.

It is also compatible with storage, indexing and decorative classification.

An identifying feature is rarer.

A distinctive iconographic combination may strongly identify a zodiac sign.

A documented copying relationship can identify transmission more strongly than style alone.

The research question becomes:

Does this feature merely fit my theory, or does it distinguish my theory from the alternatives?

That is visual identifiability.

A Parent and Teacher Guide

The easiest teaching error is to reward the fastest recogniser.

“It is a sunflower!”

“It is a bath!”

“It is a map!”

The better classroom question is:

What would have to be true for that interpretation to become stronger?

For the plant:

Do multiple diagnostic features match the same species?

For the bath:

Do known bathing manuscripts show comparable structures?

For the map:

Do landmarks preserve spatial relationships?

This changes the classroom from a guessing contest into an evidence laboratory.

Reader Checklist: Before You Believe a Visual Voynich Claim

What exactly is visible before the object is named?
How many independent features support the identification?
Which important features do not fit?
Does the identification require rotation, mirroring or selective cropping?
Is the proposed object known from the relevant historical period?
Is the comparison about function, style, motif or provenance?
Has one kind of similarity been used to prove another kind of relationship?
Does the same motif recur elsewhere in Voynich?
If it recurs, does its context remain stable?
Do associated textual patterns also recur?
Is there a documented external transmission route?
Does the hypothesis predict unseen examples?
Could a different model explain the same image equally well?
What observation would make the visual theory fail?

If the last question has no answer, the claim may be a story rather than a testable interpretation.

Frequently Asked Questions

Are the Voynich plants real plants?

Many are clearly plant-like, but confident modern species identification is difficult for much of the manuscript. They may include naturalistic, stylised, transformed, copied or composite representations.

Are the roots secret symbols?

Possibly in some cases, but resemblance to an animal, tool or letter is not enough. A symbolic theory becomes stronger when the same form recurs systematically in comparable roles.

Are the jars definitely medicine jars?

No. “Pharmaceutical” is a useful conventional label for those pages, but the exact identity and function of the vessel-like forms remain unresolved.

Do jar shapes tell us preparation methods?

No universal preparation code has been established. Container morphology remains a potentially useful classificatory variable rather than a decoded process legend.

Are the human figures bathing?

Bathing is one plausible interpretation of some scenes, which is why “balneological” is often used. The images alone do not establish the complete activity or medical meaning.

Is the Rosettes foldout a map?

Not established. It has strong relational and architecture-like geometry, but map, city, cosmological and process interpretations remain competing possibilities.

Does zodiac imagery prove astrology?

It establishes recognisable zodiac iconography. The precise astrological, calendrical, medical or organisational role of that iconography remains unresolved.

Does the Carrara Herbal prove Voynich was made in Padua?

No. It is a high-value historical comparator showing sophisticated illustrated medical manuscript capability near the relevant period. It does not provide provenance for Beinecke MS 408.

Why are Masson 116 and Sloane 4016 so useful?

Because Sloane 4016 is catalogued as a copy of Masson 116. The pair lets us examine a supported visual transmission relationship and compare that with weaker resemblance-only claims.

What is the safest conclusion about the pictures?

They are highly organised information-bearing visuals whose categories and internal relationships can be studied rigorously even though most of their intended semantic roles remain undecoded.

Research and Further Reading

The Final Idea

The Voynich pictures are not waiting politely for us to attach modern nouns to them.

They belong to a visual system whose categories, habits and purposes were obvious enough to someone that they could make an entire manuscript around them.

Our difficulty is not evidence that the pictures meant nothing.

It is evidence that the bridge between their visual world and ours has broken.

So we rebuild slowly.

Shape before name.

Recurrence before identity.

Comparison before provenance.

Prediction before confirmation.

Evidence before story.

And when the story survives all five, we can finally trust it a little more.

Voynich | Everything eduKate Knows and Tested | What the Pictures Can and Cannot Tell Us