The Voynich Manuscript is often discussed as if its greatest problem were that nobody has found the right word.
But before there were words, there were materials.
Animal skin.
Ink.
Pigment.
Water.
Quills.
Hands.
Folds.
A codex is not an abstract message. It is matter organised into a message-bearing object.
Physical evidence can tell us what the manuscript is made from, when some materials were produced, how some layers were applied, and whether obvious anachronisms are present. It cannot tell us what the text means simply by being scientific.
Quick Read
- The Voynich Manuscript is written on parchment; protein analysis has identified tested parchment as calfskin.
- Radiocarbon dating of multiple parchment samples places the animal material in the early fifteenth century, commonly summarised as 1404–1438 at about 95% probability.
- Radiocarbon dates the parchment, not the exact moment every word was written.
- McCrone Associates examined and sampled inks and pigments in 2009 using microscopy and elemental analysis.
- The tested writing and drawing materials were compatible with historical manuscript materials rather than presenting an obvious modern chemical anachronism.
- The main writing ink showed characteristics consistent with iron-gall-type historical ink systems.
- Different coloured paints contain mineral and other components broadly consistent with premodern pigment practice.
- Later non-destructive work at Yale expanded examination through imaging and spectroscopic approaches.
- Multispectral imaging can reveal faded, erased, offset or otherwise difficult-to-see marks, but enhanced visibility is not translation.
- No physical test has identified the language, author, exact place of manufacture or semantic purpose of Voynichese.
The First Question: What Is the Writing Surface?
The manuscript is written on parchment rather than paper.
Parchment is processed animal skin. It has physical memory. Thickness changes. Flaws survive. Hair-side and flesh-side differences can remain. Folds preserve structure. Damage can reveal where sheets were cut, sewn or removed.
Later protein-based testing identified calfskin in sampled Voynich leaves. Yale reported that amino-acid/protein work on multiple folios found calf parchment, with the codex requiring roughly fourteen or fifteen calfskins.
This matters because the manuscript begins to acquire a biological supply chain before we ever ask who wrote it.
Animals had to be raised.
Skins had to be processed.
Sheets had to be selected, cut, folded and prepared.
The book therefore represents material investment before it represents semantic mystery.
Radiocarbon Dating: Powerful, Precise—and Easy to Overread
In 2009, four parchment samples from different parts of the manuscript were radiocarbon dated at the University of Arizona.
The results were consistent with one another and are commonly summarised as an early-fifteenth-century range of about 1404–1438 at the reported 95% confidence level.
This result changed the Voynich debate permanently.
It made a direct Roger Bacon authorship impossible for the surviving codex because Bacon died more than a century before the animals supplying the parchment.
It also made a simple twentieth-century manufacture scenario far harder to sustain.
But radiocarbon does not date the sentence.
The test dates biological material. Parchment can in principle be stored before writing. Therefore the safe claim is:
the sampled parchment belongs to an early-fifteenth-century material horizon.
That is strong.
“The manuscript was written in 1421” is not what the laboratory measured.
Why Four Samples Matter
A single old scrap can be misleading.
Testing parchment from multiple areas makes a composite modern-forgery scenario involving randomly acquired old skins less economical.
The compatible results do not prove that every leaf came from the same herd, workshop or town. They do support a coherent material period across the tested parts of the codex.
Again, science narrows.
It does not automatically identify.
The 2009 McCrone Examination
McCrone Associates examined and sampled the Voynich Manuscript in January 2009 under Yale’s authority.
Their report is one of the most valuable public scientific documents in Voynich research because it records observations rather than merely summarising a mystery.
The codex was measured at about 23.5 centimetres high, 16.2 centimetres wide and roughly 5 centimetres deep. The writing was described as brownish-black and variable in darkness. Letter size differed between pages but was generally consistent within a page.
The examination used microscopy and elemental methods on selected inks and paints.
The job of this work was not to solve Voynichese.
It was to ask material questions:
- What elements are present?
- Do different inks behave similarly?
- Are pigments consistent with historical recipes?
- Are obvious modern synthetic materials present?
- Can writing and drawing layers be compared?
The Writing Ink
Under ultraviolet examination, the writing showed properties suggestive of iron-gall ink, one of the most common historical European writing technologies.
Elemental analysis of selected writing samples identified combinations including iron, sulphur, potassium and calcium, with variation among samples.
This is compatible with historical iron-gall-type ink systems, which can involve iron salts reacting with tannin-rich plant extracts.
But “iron-gall ink” does not mean “Italy”.
It does not mean “Prague”.
It does not mean “medical manuscript”.
Iron-gall inks were used across huge stretches of Europe and across centuries.
Material compatibility is not geographical uniqueness.
Text Ink and Drawing Ink
One of the most interesting questions is whether text and drawings were made with the same ink or in separate stages.
Microscopic and elemental observations can compare line quality, composition and layering, but apparent similarity does not automatically prove one pen, one person or one moment.
Two scribes in one workshop can use the same ink.
One scribe can use different ink batches.
A single ink batch can darken differently depending on concentration, substrate, ageing and stroke thickness.
The better question is local:
does one line pass over another?
does paint cover text?
does writing route around an existing drawing?
These relationships can help reconstruct production sequence. That is why The Order of Making deserves its own specialist article.
The Pigments
The manuscript contains blue, green, red, brown and yellowish painted areas of varying density and application quality.
McCrone’s report analysed selected colour samples and found components consistent with historical manuscript pigments.
The exact chemistry differs by colour and sample. The important public-level conclusion is not that every pigment recipe has been traced to one workshop.
It is that the tested pigments did not present the simple chemical signature one would expect from a crude modern fabrication using obviously anachronistic synthetic colourants.
That does not mean all paint was applied at the same time as the text.
Colour could be added later in the production process.
Some areas may have been retouched.
Different painters could use the same pigment family.
Material identity and production identity are different claims.
Colour Is Not a Legend
A laboratory can tell us what a colour contains more readily than what the colour means.
Green may indicate water.
Or plant tissue.
Or a category.
Or visual separation.
Or simply the colour an illustrator chose.
The chemical identity of green pigment does not decide which semantic function applies.
This is a recurring Voynich rule:
knowing what a mark is made of is not the same as knowing what the mark means.
Protein Analysis: What Calfskin Changes
Protein-based identification moved the discussion from generic “vellum” toward a biological species: calf.
That has practical implications.
Enough skins were required that the manuscript represents meaningful material cost.
It also allows researchers to compare the Voynich production with known manuscript-making economies.
But calfskin was not unique to one city.
Species identification narrows manufacturing choices more than it narrows geography.
Multispectral Imaging: Seeing What Normal Light Hides
A camera records only part of what a manuscript can reveal.
Different wavelengths interact differently with parchment, ink, pigment, erasure, stains and faded marks. Multispectral imaging captures the object under selected wavelengths and computationally combines or separates responses.
In 2014, a specialised imaging team photographed selected Voynich folios multispectrally. Later public access to some of this material gave researchers new views of faded writing, underdrawing, offset and pigment behaviour.
Multispectral imaging can do extraordinary things.
- increase contrast between ink and parchment;
- recover faded marks;
- separate visually similar pigments;
- reveal erased or overwritten traces;
- highlight offsets and show-through;
- help distinguish later writing layers.
It cannot automatically read an unknown script.
Better visibility can solve a transcription problem.
It cannot by itself solve a semantic problem.
Imaging Can Change the Alphabet
This is where material science meets textual analysis directly.
If a faint stroke becomes visible under enhanced imaging, what looked like one glyph may become two.
A supposed rare sign may become a damaged common sign.
A suspected ligature may prove to contain separable strokes.
An erased annotation may reveal a later ownership or reading layer.
This means every statistical study ultimately rests on physical visibility.
The “digital text” is downstream from the parchment.
See EVA, Transcription and the Segmentation Problem and The Alphabet Problem.
Can Materials Tell Us Where Voynich Was Made?
Sometimes material science can contribute to provenance.
Parchment preparation methods can vary.
Pigment recipes can vary.
Trace-element profiles can sometimes narrow sources.
Animal populations can carry geographic signals.
But medieval trade complicates easy conclusions.
Prepared parchment moved.
Pigments moved.
Recipes moved.
Craft workers moved.
A material can be compatible with a region without being unique to it.
So a chemically plausible northern-Italian production environment remains a possibility, not a custody chain.
See The Broken Provenance Chain.
Can Materials Tell Us Whether Voynich Is a Hoax?
They can eliminate some hoax scenarios more effectively than others.
A twentieth-century fake made on modern paper with synthetic ink would be easy to contradict.
That is not what we have.
The parchment is early fifteenth century.
The tested materials are compatible with historical manuscript manufacture.
The ownership trail predates Wilfrid Voynich.
Therefore the simple modern-forgery theory fails.
But material authenticity does not prove semantic authenticity.
A fifteenth-century maker could still create:
- a cipher;
- a private notation;
- an invented script;
- abbreviated language;
- mnemonic pseudo-text;
- generated text;
- or a deliberately deceptive work.
Old materials prove an old object.
They do not prove ordinary plaintext.
Can Chemistry Tell Us Who Wrote It?
Not directly.
One ink recipe can be shared by several scribes.
One scribe can switch ink.
Hand identification depends more strongly on palaeography: stroke formation, ductus, spacing, character construction and repeated habits.
Material analysis can support production grouping, but chemistry is not a signature in the ordinary forensic sense.
Can Science Ever Decipher Voynich?
Science can contribute enormously to decipherment without being the decipherment itself.
It can improve scans.
Clarify damaged glyphs.
Distinguish layers.
Constrain chronology.
Test whether pigments or inks belong to different production episodes.
Reveal erased writing.
Help reconstruct missing or altered forms.
Each reduces uncertainty.
But if the script encodes language through a transformation, semantic recovery still requires a model connecting sign behaviour to meaning.
Primary School: Material or Meaning?
Show a child a red pencil mark and ask two questions.
- What is it made from?
- What does it mean?
The first is a material question.
The second is a communication question.
A laboratory may answer the first perfectly and still know nothing about the second.
Secondary School: Build an Evidence Table
- Radiocarbon: dates biological material.
- Protein analysis: identifies animal species.
- Microscopy: examines strokes and layers.
- XRF/elemental analysis: detects elemental composition.
- Raman spectroscopy: can help identify compounds or pigments.
- Multispectral imaging: changes visibility across wavelengths.
Then ask what none of those methods can directly answer:
What does qokedy mean?
The student learns that methods have domains.
Reader Checklist: Before You Believe a “Scientific Proof” About Voynich
- What material was actually tested?
- How many samples?
- Which folios?
- Was the method destructive or non-destructive?
- What did the instrument directly measure?
- What inference was added after the measurement?
- Does the result date parchment, ink, paint or something else?
- Is the material unique to one location?
- Could trade move the material?
- Does the claim jump from compatibility to provenance?
- Does enhanced imaging reveal a mark or interpret it?
- What uncertainty remains?
Frequently Asked Questions
Is the Voynich Manuscript really medieval?
The parchment is securely consistent with the early fifteenth century according to radiocarbon testing. That makes the physical substrate medieval.
Is the parchment calfskin?
Protein analysis of tested folios identified calf parchment.
Does carbon dating prove the writing date?
No. It dates the animal material. Writing likely followed parchment manufacture, but the exact interval is not measured by radiocarbon testing of the skin.
Are the inks medieval?
The tested inks and pigments are chemically compatible with historical manuscript materials and do not present an obvious modern anachronism. That is stronger than saying chemistry independently dated every stroke.
Can multispectral imaging reveal hidden writing?
It can improve visibility of faded, erased, offset or otherwise difficult marks. Whether those marks can then be interpreted is a separate question.
Could the manuscript use old parchment but modern ink?
A modern-forgery scenario of that kind would have to explain why tested inks and pigments are compatible with historical materials and why the manuscript has a pre-modern ownership trail. It is not the economical explanation supported by current evidence.
Research Foundations
- McCrone Associates — Materials Analysis of the Voynich Manuscript (2009)
- Yale News — Voynich facsimile and scientific testing overview
- Yale Beinecke Library — Voynich Manuscript
- Voynich — eduKateSG Master Article
The Final Idea
The laboratory has already solved several Voynich questions.
It has told us that the physical manuscript belongs to an old material world.
It has narrowed chronology.
It has identified parchment species.
It has shown that tested inks and pigments behave like historical manuscript materials.
It has given us better ways to see faded layers.
Those are genuine victories.
The mistake is believing that because the machine is scientific, every question asked of it becomes scientifically answerable.
Science can tell us more clearly what the object is. Meaning still requires an explanation of what the object does.