Voynich Comparator | Historia Plantarum MS 459: What a Northern Italian Medical Encyclopedia Does With More Than 500 Images

There is a dangerous comparison in Voynich research.

We find another medieval manuscript full of plants.

It is Italian.

It is near the Voynich parchment window.

It contains medicine.

And suddenly resemblance begins behaving like ancestry.

Biblioteca Casanatense MS 459 is exactly the kind of manuscript that can tempt us into that mistake.

It is also exactly the kind of manuscript that can prevent it.

The Historia Plantarum is a richly decorated late-fourteenth-century northern Italian encyclopedia of the natural and medical world. The Biblioteca Casanatense describes it as containing plants, animals and minerals with particular attention to medical properties. More than 500 illustrations fill the codex, often accompanied by scenes of daily life. It was made in the orbit of Gian Galeazzo Visconti for Wenceslas IV and is associated in part with the Lombard artistic world of Giovannino and Salomone de’ Grassi.

Everything about it makes it an excellent Voynich comparator.

Not because it tells us where the Voynich was made.

Because it tells us what a known, image-rich, medically relevant, northern Italian knowledge object can look like when the historical channels are visible.

Historia Plantarum MS 459 is not a candidate parent of the Voynich Manuscript. It is a calibration object for scale, visual organisation, patronage, medical natural history and the difference between resemblance and demonstrated relationship.

Quick Read

Biblioteca Casanatense MS 459 is generally dated to the late fourteenth century.
It was produced in a northern Italian/Lombard courtly environment associated with Gian Galeazzo Visconti and Wenceslas IV.
The Casanatense describes it as an encyclopedia of natural science containing plants, animals and minerals with attention to medical properties.
It contains more than 500 illustrations.
Many images include scenes from daily life rather than isolated specimens alone.
Its decoration is associated in part with Giovannino and Salomone de’ Grassi.
Its high-status patronage and luxurious execution form a useful negative control against assumptions that every image-rich medical manuscript should look alike.
Its plant images can be compared against other known herbal traditions at feature level rather than by overall “Italian” style.
Shared northern Italian context does not establish Voynich provenance.
The comparator is most useful when it tells us what Voynich does differently.

Comparator Firewall

Late-fourteenth-century Lombardy + medicinal plants + rich imagery ≠ Voynich source, author, workshop or provenance.

MS 459 enters the comparator programme as a known historical object.

We know enough about its production environment to use it as a control.

That is a different epistemic job from saying the Voynich came from the same place.

The First Difference Is Scale

The Casanatense describes more than 500 illustrations of plants, animals and minerals.

This is not simply a “herbal” in the narrow modern sense.

It is a large natural-medical world.

That matters because the Voynich is often partitioned into modern-looking categories—herbal, astronomical, biological, pharmaceutical, recipes—as though medieval compilers necessarily respected those boundaries.

MS 459 gives us a known counterexample to overly narrow classification.

A late medieval knowledge book can comfortably place plants, animals, minerals, health and scenes of human activity inside one visual-intellectual environment.

That does not prove the Voynich is an encyclopedia.

It proves that mixed natural-medical content is historically normal enough that we should test it before declaring the Voynich’s diversity anomalous.

The Second Difference Is Social Visibility

MS 459 belongs to a world of visible elite patronage.

The Casanatense links its execution to the court of Gian Galeazzo Visconti and to Wenceslas IV. Its decoration is lavish enough to preserve the fingerprint of a major artistic environment.

The Voynich does not present itself this way.

It lacks an obvious dedicatory portrait, conventional title apparatus, visible author, explicit patron, heraldic programme or surviving commission record.

This contrast is useful.

If someone proposes a high-court Voynich production model, MS 459 shows what one nearby high-court natural-medical manuscript can look like.

The theory then has to explain why Voynich suppresses or lacks so much of that public apparatus.

Luxury Is Not the Same Variable as Knowledge Complexity

This is an important control.

A book can be intellectually complex and materially modest.

A book can also be intellectually complex and materially luxurious.

MS 459 occupies the second quadrant.

The Voynich appears much closer to the first.

Therefore the absence of courtly polish cannot be used to infer intellectual simplicity, fraud or amateurism.

Material register and knowledge architecture are separate axes.

More Than 500 Images Create a Statistical Control

A large illustrated corpus lets us move beyond cherry-picked pairs.

Instead of asking whether one Voynich plant resembles one Casanatense plant, we can ask:

How are roots simplified?
How often are flowers enlarged?
How stable are leaf counts?
How are stems bent to fit the page?
How often are plants shown with animals or human scenes?
How much visual detail survives between related illustrations?
How much distortion comes from page design rather than failed observation?

The larger the control corpus, the less power one attractive image has to mislead us.

Plant Identification Should Be Tested Feature by Feature

Published botanical comparison of medieval cucurbit imagery illustrates the value of this approach.

Researchers compare leaf shape, tendrils, flowers, fruit geometry, colour, surface markings and relative proportions across manuscripts including Casanatense 459, Manfredus, Masson and Carrara.

This is much stronger than saying two plants “look similar.”

Voynich comparison should use the same discipline.

Break an image into attributes.

Score the attributes.

Measure rare shared distortions separately from generic plant geometry.

Then compare against unrelated controls to estimate coincidence.

Daily-Life Scenes Show That Medical Knowledge Can Be Contextual

Many MS 459 images do more than isolate a medicinal object.

They place natural materials inside human activity.

This matters for the Voynich because the manuscript’s human figures are often treated as a separate “biological” problem.

A comparator such as MS 459 reminds us that late medieval natural-medical books can integrate substances, bodies and activities visually.

Again, this does not solve the Voynich figures.

It widens the historically plausible design space.

A Known Courtly Manuscript Gives Us a Negative Control for the Voynich Paratext Problem

The Voynich lacks many ordinary navigational signals.

MS 459 reminds us that high-investment books usually had no reason to hide their status from intended readers.

Titles, patronage signals, recognisable scripts and conventional content architecture help a manuscript enter a social system.

The Voynich’s absence of obvious public-facing identifiers therefore remains a real explanatory problem.

A courtly-origin theory has to account for that absence rather than merely point to visual sophistication.

The Lombardy Trap

MS 459 is northern Italian.

The Carrara Herbal is Paduan.

Sloane 4016 is associated with Lombardy.

Roccabonella belongs to the Venetian world.

It is easy to accumulate these examples and then say northern Italy is “obviously” the Voynich environment.

That is exactly what the comparator programme must prevent.

Northern Italy produced many relevant manuscripts because it was an enormous, dense and innovative manuscript culture.

High comparator density can reflect the richness of the control environment rather than the origin of the target object.

A region producing many plausible comparators is evidence that the region could produce such books. It is not yet evidence that it produced this book.

What MS 459 Can Legitimately Teach Voynich Research

Late medieval northern Italian medical-natural knowledge could be represented at enormous visual scale.
Plants, animals, minerals and human activity could inhabit one encyclopedic programme.
Luxury patronage produces paratextual and artistic signals that can be compared against Voynich’s relative anonymity.
Large image corpora allow feature-level statistical comparison instead of isolated resemblance.
Plant images can preserve some diagnostic structures while idealising or stylising others.
Mixed content is not automatically evidence of a bizarre or incoherent book.
Visual sophistication does not require the same material or social register across manuscripts.

What MS 459 Does Not Prove

that Voynich was made in Milan or Lombardy;
that Voynich belonged to the Visconti court;
that similar plant forms indicate direct copying;
that the Voynich human figures have the same function as Casanatense daily-life scenes;
that an encyclopedic comparator makes Voynich an encyclopedia;
that northern Italian density outweighs contradictory evidence from elsewhere.

A Better Historia Plantarum Comparator Test

Create a structured catalogue of MS 459 plant, animal and mineral images.
Separate object images from contextual human scenes.
Encode plant morphology into comparable features.
Measure page-fit distortions and decorative transformations.
Compare against Manfredus, Carrara, Masson, Sloane and Roccabonella before introducing Voynich.
Estimate which similarities are common to northern Italian visual culture.
Identify rare shared anomalies that might carry genealogical weight.
Blind the evaluator to manuscript identity during feature scoring.
Introduce Voynich only after the transformation baseline is frozen.
Reject any “match” that requires more interpretive flexibility than known manuscript relationships.

Student Lens: One World Can Have Many Information Layers

Primary: sort pictures into plant, animal, mineral and human activity, then notice that one health book may need all four.

Lower Secondary: compare two illustrated encyclopedias and separate what is being represented from how it is organised.

Upper Secondary: distinguish similarity caused by common genre from similarity caused by direct descent.

JC and adult readers: model manuscript comparison as Bayesian evidence: generic regional features have low likelihood ratios; rare shared errors and ordered multi-image correspondences carry more weight.

Research Sources

Continue the Comparator Route

The Final Idea

Historia Plantarum MS 459 is valuable precisely because it is so tempting.

Northern Italy.

Late fourteenth century.

Plants.

Animals.

Minerals.

Medicine.

Hundreds of images.

It gives us almost everything needed to build a seductive Voynich story.

So the correct scientific use is the opposite.

Use it to measure how much of that story can be explained by the ordinary capabilities of the period.

The best comparator does not make the Voynich look more familiar. It tells us exactly which parts of the Voynich remain unusual after familiarity has been accounted for.