Egerton MS 747 as a Voynich Comparator

Egerton MS 747 changes the Voynich comparison problem because it forces us to stop looking at medieval medicine one page at a time.

It is easy to compare one Voynich plant with one medieval plant.

It is much harder—and much more useful—to compare systems.

The British Library catalogues Egerton MS 747 as a parchment codex of 147 folios, produced in southern Italy, specifically Salerno in the catalogue record, and dated broadly to about 1280–1350.

Its contents do not stop with an illustrated herbal.

They continue into compound medicines, dosage, substitutions, weights and measures, synonyms, and supplementary pharmaceutical material.

Egerton 747 is therefore not just a picture comparator. It is a control for what an integrated medieval medical information system can look like when identity, prescription, quantity, substitution and terminology survive in the same codex.

Quick Read

British Library shelfmark: Egerton MS 747.
Catalogue date: c.1280–1350.
Catalogue origin: southern Italy (Salerno).
Material: parchment.
Extent: 147 folios.
Layout: generally two columns of up to 55 lines.
Script: Gothic.
ff. 1r–106v: Tractatus de herbis, here attributed in the manuscript to Bartholomaeus Mini de Senis.
ff. 106v–109r: additional plant images.
ff. 109v–111r: an added forecasting fragment with a lunar calendar.
ff. 112r–124v: Antidotarium Nicolai.
ff. 124v–125v: De dosibus medicinarum.
ff. 125v–127r: a quid pro quo substitution list for unavailable ingredients.
ff. 127v–128v: fragments on weights and measures.
ff. 128v–146r: synonyms for plant names and medical ingredients.
ff. 146v–147v: a supplement to the Antidotarium Nicolai.
The British Library calls Egerton 747 the oldest surviving copy of the Tractatus de herbis.

Comparator Firewall

Egerton MS 747 demonstrates what one real medieval medical knowledge system looked like. It does not establish that Voynich is medical, Salernitan, Italian, Latin, derived from Egerton, or organised by the same modules.

The point of the comparator is to make medical claims harder, not easier.

If someone proposes that Voynich is a medical manual, Egerton gives us a historically grounded control against which that proposition can be tested.

The Manuscript Is a Stack, Not a Single Genre

The most important feature of Egerton 747 may be the coexistence of several medical functions inside one physical book.

The codex moves through different operational layers:

Identify a simple medicinal substance.
Read about its properties and uses.
Move into compound remedies.
Determine dosage.
Substitute when an ingredient is unavailable.
Use weights and measures.
Resolve synonym problems across names.
Consult supplementary pharmaceutical material.

This is an operating sequence.

It does not prove that every medieval practitioner used the volume linearly, nor that one owner used every section. But it shows that these information problems were sufficiently connected to be gathered into one codex.

Layer 1: Identification — The Tractatus de herbis

The first major component is the Tractatus de herbis, extending through more than one hundred folios.

The British Library describes large miniatures of plants and some animals, people and landscapes set into the text area and illustrating the substances discussed, from Aloen through Zuchara.

This is already more sophisticated than “plant picture + paragraph”.

The page joins:

visual recognition;
named entry;
textual description;
therapeutic use;
ordered retrieval.

The image and prose are not redundant.

They solve different parts of the same problem.

Layer 2: Compound Medicines — The Antidotarium Nicolai

At folio 112r the codex changes function.

Instead of primarily asking what a simple substance is and what it can do, the Antidotarium Nicolai enters the world of compounded medicines.

This is a different information problem.

A compound remedy requires coordination across multiple variables:

ingredient identity;
ingredient sequence;
quantity;
preparation;
combination;
administration;
indication.

The move from simples to compounds is therefore not merely a change of subject.

It is a change of computational burden.

Layer 3: Dose — Medicine Becomes Quantitative

The presence of De dosibus medicinarum immediately after the antidotary is significant.

Once a medicine is compounded, “how much?” becomes unavoidable.

A dosage layer converts qualitative knowledge into controlled execution.

Identity without quantity is not yet a safe executable prescription.

This point is useful for Voynich because one of the long-standing questions is where its numbers and quantities are.

If the manuscript encodes recipes, pharmacology or dosage-heavy practice, a model must eventually explain how quantity is represented—or explain why quantity is unnecessary for the proposed task.

Layer 4: Substitution — What Happens When Reality Refuses the Recipe?

Egerton 747 includes a quid pro quo list explicitly intended for situations in which a required ingredient is unavailable.

This is one of the most revealing modules in the book.

A written recipe lives in a world of ideal ingredients.

A practitioner lives in a world of seasons, trade routes, cost, spoilage, geography and missing stock.

The substitution list is the bridge between the prescription and the world.

It says, in effect:

If X is unavailable, use Y under the accepted rule.

That is not decorative knowledge.

It is resilience engineering for pharmacy.

Layer 5: Weights and Measures — Shared Units Make Distributed Practice Possible

The codex then preserves fragments on weights and measures.

Why would this matter?

Because a recipe cannot travel reliably if every reader interprets quantity differently.

Weights and measures allow a remedy to move across people, shops and cities with less ambiguity.

They create an interoperability layer.

For Voynich, this gives us another falsification pressure. If a hypothesis requires precise compounding but provides no account of quantitative representation, it has an unresolved operational gap.

Layer 6: Synonyms — The Same Drug Can Have Too Many Names

Egerton’s large synonym section runs from folio 128v to 146r.

This is not a trivial appendix.

Medieval Mediterranean pharmacology crossed languages, regions and commercial networks. A single substance might have:

a learned Latin name;
a Greek-derived name;
an Arabic-derived name;
a local vernacular name;
a merchant name;
a shop synonym;
a spelling variant.

The synonym problem is therefore a patient-safety and supply-chain problem, not merely a linguistic curiosity.

If the physician prescribes one name and the apothecary stocks another, the system needs a crosswalk.

Egerton contains one.

The Codex Connects Knowledge to Execution

When the modules are viewed together, a deeper structure appears:

name → identify → select → combine → quantify → substitute → execute

That is why Egerton 747 is such a powerful comparator.

It shows that a real medical book can contain several representational systems because medicine itself requires several kinds of information.

Pictures solve one problem.

Synonyms solve another.

Weights solve another.

Substitution solves another.

Recipes solve another.

A manuscript can look heterogeneous because the real-world task is heterogeneous.

This Helps Us Reframe the “Six Voynich Sections”

Voynich is commonly divided by modern observers into herbal, astronomical/astrological, biological, cosmological, pharmaceutical and recipe-like sections.

Those divisions are useful descriptive handles.

They are not proven medieval genres.

Egerton 747 warns us against assuming that visual difference means unrelated subject matter.

A single practical domain can require multiple visual and textual modes.

That means Voynich’s section changes could, in principle, reflect:

different modules of one system;
different genres bound together;
different production phases;
different users;
different source traditions;
or some combination.

Egerton does not tell us which answer is correct.

It tells us not to assume the answer from page appearance alone.

The Added Lunar Material Is Another Important Warning

Egerton includes an added forecasting fragment and lunar calendar in a different mid-fourteenth-century Gothic cursive hand.

This is a visible example of temporal layering.

A medical codex can acquire material after its main production.

Therefore a later astronomical, calendrical or marginal layer does not automatically define the original function of the whole book.

Voynich research needs the same caution with foliation, quire numbers, month names and marginalia.

One binding can preserve several times.

The Attribution Problem: Do Not Overread the Name

The British Library notes that the Tractatus de herbis is here attributed to Bartholomaeus Mini de Senis and that a colophon associated with the name appears over an erasure.

That is exactly the kind of detail that should slow interpretation.

A name in a manuscript can identify an author, copyist, owner, corrector, source, dedicatee or later intervention.

The presence of a name is evidence.

Its role still has to be established.

This is directly relevant to Voynich marginalia and readable later-hand material.

What a Medical Voynich Hypothesis Should Be Asked to Explain

Egerton gives us a useful challenge list.

Identity: where are substances named or distinguished?
Selection: what determines which material is used?
Combination: how are compounds encoded?
Quantity: where are dose or ratio relationships?
Preparation: how are actions represented?
Substitution: is there evidence for interchangeable ingredients?
Synonymy: could multiple labels or word families encode alternative names?
Retrieval: what page or section architecture helps users find an entry?
Execution: can a trained reader turn the information into a reproducible action?

A Voynich medical theory does not need to reproduce Egerton’s exact interface.

But the more practical the proposed function, the more of these operational problems it must solve somehow.

Why the Number Problem Becomes More Important, Not Less

The Voynich Manuscript has no securely identified ordinary numeral system in its main script.

That does not rule out medicine.

Quantities might be encoded by words, signs, repeated tokens, layout, omitted conventions or a system not yet recognised.

But Egerton shows why the question cannot simply be ignored.

A compound pharmaceutical workflow eventually reaches dosage and measurement.

If Voynich’s “pharmaceutical” pages really encode compounding, the quantity channel is one of the strongest places to search for structured regularity.

Why the Synonym Problem Could Mimic Vocabulary Families

Voynich words often form dense near-neighbour families.

One temptation is to interpret those as grammatical inflection.

Another is to interpret them as generated variants.

Egerton reminds us of a third information problem: synonyms and terminology management.

A medical system may carry multiple names for related or identical substances.

That does not mean Voynich word families are synonyms.

It means synonym-heavy technical registers belong in the control suite before morphology is declared.

What Egerton 747 Can Legitimately Teach Voynich Research

A medieval medical codex can contain several tightly connected information modules.
Plant identification is only one layer of medicine.
Compound remedies create a different textual structure from simple-drug descriptions.
Dosage and measurement are operational necessities for many pharmaceutical tasks.
Substitution lists connect ideal recipes to real supply constraints.
Synonym lists solve cross-language and cross-market identity problems.
Calendrical material can be added later without defining the original codex.
Names and colophons require role analysis before authorship is assigned.
A genuine medical hypothesis should explain how information becomes executable.

What Egerton 747 Does Not Prove

that Voynich is a medical codex;
that Voynich was produced in Salerno;
that its herbal pages derive from the Tractatus de herbis;
that its “pharmaceutical” jars encode the Antidotarium Nicolai;
that its stars or circles are a lunar calendar;
that Voynich word families are synonyms;
that missing obvious numerals can be dismissed;
that a similar multi-section architecture establishes shared provenance.

A Better Medical-Comparator Test

Define the proposed Voynich medical task before comparing manuscripts.
List the information that task minimally requires.
Map how Egerton solves each requirement.
Map how several unrelated medical codices solve the same requirement.
Ask whether Voynich contains a plausible functional equivalent.
Record missing modules as negative evidence.
Do not require Voynich to imitate Egerton’s surface design.
Require any proposed interpretation to predict cross-section behaviour.

This changes the question from “Does Voynich look medical?” to “Can the proposed system actually perform a medical job?”

Student Lens: A Book as an Operating System

Primary: imagine making a medicine safely. What information would you need besides a picture of the ingredient?

Lower Secondary: arrange cards labelled Identify, Measure, Substitute, Combine and Name. Which must happen before which?

Upper Secondary: compare a medieval medical codex with a modern medication information system and identify invariant functions despite different interfaces.

JC and adult readers: model Egerton as a modular information architecture with typed transitions between entity resolution, formulation, measurement, substitution and execution.

Research and Primary Sources

Continue Through the Voynich Comparator Library

The Final Idea

Egerton MS 747 is valuable because it makes medieval medicine operational.

A plant is not enough.

A name is not enough.

A recipe is not enough.

The system needs identification, combination, quantity, substitution, terminology and execution.

That is the standard a serious medical interpretation of Voynich should eventually face.

Do not ask only whether Voynich resembles a medical manuscript. Ask whether the proposed interpretation can perform the work a medical manuscript had to perform.

Voynich Comparator | Egerton MS 747: A Medieval Medical Stack From Plant Identification to Dose, Substitution and Compound Remedy