Voynich Comparator | The Vienna Dioscurides: When a Medical Plant Book Survived a Thousand Years of Use, Translation and Repair

A manuscript can be six centuries old before the Voynich parchment even exists.

Then it can still be actively used.

Annotated.

Translated.

Repaired.

Rebound.

And carried forward as a medical reference.

The Vienna Dioscurides—Österreichische Nationalbibliothek Cod. med. gr. 1—is one of the most important controls in the entire Voynich comparator programme because it makes deep transmission visible.

The manuscript was made at the beginning of the sixth century for the Roman aristocrat Juliana Anicia. It preserves Greek medical and pharmacological material centred on the writings of Pedanius Dioscorides and contains more than 400 illustrations of plants and animals. Its history did not end when the book was finished. Multilingual annotations and translated plant names show prolonged use. In 1406—roughly the opening of the Voynich parchment window—the scholar Johannes Chortasmenos arranged for the ancient codex to be extensively restored and rebound.

That single chronology changes how we should think about medieval source material.

A fifteenth-century medical culture could be physically handling, repairing and consulting a book whose information lineage was already ancient.

Quick Read

The Vienna Dioscurides is held by the Austrian National Library as Cod. med. gr. 1.
It was produced at the start of the sixth century for Juliana Anicia.
UNESCO describes it as a masterpiece of late-antique book art and a major carrier of botanical and pharmacological knowledge.
The manuscript contains more than 400 illustrations of plants and animals.
Its principal textual layer transmits Dioscoridean materia medica along with material from other authors.
Multilingual annotations and translated plant names reveal centuries of active use.
In 1406 Johannes Chortasmenos arranged an extensive restoration and new binding.
The manuscript therefore had a major conservation event almost exactly when the Voynich parchment was being produced.
Its survival demonstrates that a fifteenth-century manuscript can inherit visual and pharmacological traditions vastly older than its own physical support.
It does not prove any direct relationship to Beinecke MS 408.

Comparator Firewall

Dioscorides tradition ≠ Voynich source. Ancient plant iconography ≠ Voynich ancestry. A tradition surviving into the fifteenth century proves availability and longevity, not copying.

This distinction is central.

The Vienna Dioscurides is useful because its history is exceptionally well documented.

We can watch an old knowledge object survive.

We do not need to invent a relation to Voynich in order to learn from it.

The Physical Book Is Sixth Century; the Knowledge Is Older

Dioscorides lived in the first century CE.

The Vienna manuscript was produced centuries later.

Even the Dioscoridean text itself drew on earlier pharmacological knowledge.

So one codex already contains several time layers:

older pharmacological traditions → first-century synthesis → sixth-century manuscript host → medieval annotation → 1406 restoration → later collection history

This is an essential Voynich lesson.

Radiocarbon dates parchment.

It does not date the origin of every idea represented on that parchment.

A 1420 manuscript could carry a diagram tradition from 1200, a pharmacological list from 900, a translated authority from antiquity and a local fifteenth-century compilation layer simultaneously.

Therefore:

object date and information-lineage date must remain separate axes.

The Images Are Not Merely Decoration

The Vienna Dioscurides is famous for its plant illustrations because the images participate in identification and medical transmission.

This makes it a deep comparator for the Voynich herbal pages.

But the right comparison is not:

Which Voynich plant looks most like a Dioscorides plant?

The stronger questions are:

Which visual features survive across known copies?
Which details are repeatedly redrawn?
How do captions remain attached to images?
What happens when a plant name is translated?
Can the same visual lineage survive while language changes?
Which errors or distortions become inherited markers?

These questions turn the tradition into a real copying-control system.

More Than 400 Images Create a Real Visual Corpus

A single plant resemblance is weak evidence.

A tradition containing hundreds of identifiable images lets us study transformation statistically.

We can examine:

silhouette stability;
root simplification;
leaf number and orientation;
flower placement;
relative scale;
colour conventions;
caption position;
page layout;
image-order stability;
shared mistakes.

A strong Voynich ancestry claim should survive comparison at this level.

It should not depend on one memorable silhouette.

Multilingual Plant Names Show How Identity Survives Translation

The Austrian National Library notes entries and translations of plant names in different languages.

This is enormously useful.

The plant stays the referent.

The label changes.

A later user can add a new linguistic interface without redrawing the botanical object.

This proves that visual and linguistic ancestry can decouple.

The same image can support several naming systems across time.

For Voynich this means:

even if an image lineage were identified, the language beside the image would still need its own evidence.

Use Leaves Traces

A manuscript used for centuries changes physically and intellectually.

Readers annotate.

Names are glossed.

Bindings wear.

Margins acquire evidence.

Repairs enter the object.

The Vienna Dioscurides is therefore a good control for the Voynich Marginalia and Damage-and-Repair problems.

Later hands are not necessarily noise.

They can record the manuscript’s changing receiver community.

1406 Is a Remarkable Historical Coordinate

The major restoration arranged by Johannes Chortasmenos in 1406 is especially striking for our project.

That year sits essentially at the start of the Voynich parchment radiocarbon range.

This does not connect the two manuscripts.

What it does is establish a concrete historical fact:

while new fifteenth-century parchment books were being made, much older medical codices were still circulating, being studied and being repaired.

The fifteenth-century source field was therefore not limited to contemporary books.

It included old objects with deep information histories.

Restoration Can Change a Codex Without Changing Its Core Knowledge

A rebinding can alter the physical host.

Repair strips can enter the folds.

Leaves can be reinforced.

The book can receive a new material state while the core text and images survive.

This is a critical control for Voynich binding history.

Physical state and information state do not have to change together.

A later binding can preserve earlier content.

A missing early cover does not mean the current text block is late.

Likewise a repair should be dated independently from the writing it stabilises.

The Vienna Dioscurides Also Defeats the “One Original Author” Reflex

The physical codex is a compilation centred on Dioscorides but also includes material from other authors.

Later readers add further linguistic layers.

The manuscript is therefore not the expression of one mind at one moment.

Its coherence emerges from a medical information tradition and a long history of use.

This matters for the Voynich exemplar problem.

A complex medical codex may have:

multiple textual ancestors;
an inherited image tradition;
a local compiler;
one or more scribes;
later annotators;
later binders and conservators.

Searching for a single named “author” can therefore be the wrong abstraction.

This Is Also a Geographic Control

The Vienna Dioscurides belongs to an eastern Mediterranean/Byzantine history, not a Paduan one.

Yet it contains several features that might otherwise be over-weighted in a Padua argument:

medicinal plants;
large-scale illustration;
pharmacological organisation;
multilingual naming;
long-term scholarly use;
repair in the early fifteenth century.

None of those is geographically diagnostic.

Padua must be supported by evidence that distinguishes Padua from the wider medical manuscript world.

What the Vienna Dioscurides Can Legitimately Teach Voynich Research

Medicinal image traditions can survive for many centuries.
The age of a knowledge tradition can be far older than the manuscript carrying it.
Images can remain useful while labels and languages change.
Medical books can accumulate multilingual annotations through use.
A manuscript can be substantially repaired and rebound without losing its older information core.
Old medical codices were still active objects around 1406.
Large known image corpora allow transformation rates to be measured.
Visual ancestry and language ancestry must be treated separately.
Material history and intellectual history travel on different clocks.

What It Does Not Prove

that the Voynich derives from Dioscorides;
that any Voynich plant is directly copied from Cod. med. gr. 1;
that Voynichese is Greek;
that multilingual annotations imply Voynich is a translation aid;
that early-fifteenth-century restoration creates a relationship to the Voynich parchment window;
that ancient imagery explains Voynich’s hybrid-looking plants;
that Byzantine transmission localises Voynich outside or inside Italy.

The eduKate Deep-Lineage Test

Build a corpus of securely related Dioscorides witnesses across several centuries.
Align plant images by identified entry, not visual guess.
Measure stable and unstable morphological features.
Track captions and translated plant names separately from image change.
Record reordering and missing entries.
Estimate how much visual drift occurs after centuries of copying.
Freeze that transformation budget before testing Voynich.
Require multiple independent Voynich correspondences to fit the same lineage model.
Downweight traits common across unrelated herbal traditions.
Reject direct-descent claims that require more freedom than the known Dioscorides transmission itself.

Student Lens: Old Knowledge Can Live in a New Book

Primary: copy a grandparent’s recipe into a new notebook. The notebook is new; the recipe is older.

Lower Secondary: separate the date of a source idea from the date of the copy that preserves it.

Upper Secondary: track what changes when one illustrated reference work is copied across languages.

JC and adult readers: model manuscript history as multiple clocks—substrate, textual ancestry, image ancestry, annotation, binding and custody.

Research Sources

The Final Idea

The Vienna Dioscurides is older than the Voynich mystery by almost a thousand years.

Yet in 1406 someone still cared enough about it to repair and rebind it.

That fact changes the shape of the source field.

The fifteenth century did not begin with a blank shelf.

It inherited old books, old pictures, old names and old pharmacology—and kept translating them into new users and new material states.

If the Voynich carries old information in a fifteenth-century host, that would not be extraordinary. The hard problem is demonstrating which lineage, which transformation and which surviving invariants actually connect it to anything older.