Abhishek S.
Shipping in public. Listening in private.

Abhishek

I lead women’s Indo-Western & Premium at Max Fashion. I also wrote the AI that runs the buying floor.

Rare profile. Category operator who ships production code.

Senior Buying Leader · Max Fashion Women’s Indo-Western & Premium · 530+ India stores NIFT ’12 · Twelve years on the floor

read the CV start a letter
abhishek@bengaluru ~ %
>role: senior buying lead
>dept: women’s indo-western + premium
>floor: 530+ stores india

Voynich Manuscript — Five Scribes vs. the Hoax Hypothesis

The discovery that the Voynich Manuscript was produced by five distinct scribes is not merely a codicological fact. It is the strongest single piece of evidence against the hoax hypothesis — and it is underappreciated, even in the specialist literature. The question of whether a collaborative manuscript can be a coordinated hoax forces into view every assumption about what medieval deception would have required.

The Five-Scribe Structure

Lisa Fagin Davis's 2020 paleographic analysis identified five distinct hands in the manuscript:

Scribe Primary domain Sharing
Scribe 1 Herbal (shared) Collaborates on all herbal folios
Scribe 2 Astronomical + astrological Sole author of this section
Scribe 3 Balneological Sole author of bathing-pool section
Scribe 4 Herbal + pharmaceutical (collaborative) Works alongside Scribe 5
Scribe 5 Herbal + pharmaceutical (collaborative) Works alongside Scribe 4

Two features are remarkable. First, the section specialization: each scribe covers a domain that corresponds to a coherent knowledge area. This mirrors institutional production — a scriptorium where different scribes copy out the sections of their competency. Second, the low error rate: the scribes made few mistakes, meaning they wrote confidently in the Voynichese script without second-guessing themselves. Producing invented text with low-error fluency across 240 pages is not what a hoaxer generating on the fly looks like.

Why Five Scribes Argues Strongly Against Hoax

1. The coordination problem

A lone eccentric hoaxer has one challenge: produce plausible-looking fake text. A five-person collaborative hoax has a fundamentally harder problem: all five must apply the same generating method consistently to produce a unified statistical profile.

The Voynich text follows Zipf's Law (word frequencies drop as 1/rank), Brevity's Law (more common words are shorter), and Heap's Law (vocabulary grows as a power function of text length) — all three simultaneously. These are emergent properties of natural language; they are not properties of randomly generated text, and they are not easy to deliberately engineer.

For one hoaxer with a systematic method (e.g., Rugg's grille): plausible. For five collaborating hoaxers to each produce Zipfian output in their sections while maintaining consistency across sections: this has never been demonstrated, and the coordination cost is extraordinary.

2. The fluency argument

The scribes wrote with few errors. Writing an invented script with near-zero hesitation and correction implies one of two things:

The first interpretation is supported by the 2024 Cryptologia paper (Zipf + Brevity + Heap simultaneous compliance) and by the 2025 Naibbe cipher demonstration that a medieval cipher mechanism can produce Voynich-like statistics. The second interpretation — five people separately internalizing a fake-language generation method to the point of low-error fluency — requires each person to have practiced producing Zipfian statistics, which is not a natural human skill.

3. The use-wear evidence (Davis, 2025)

Fagin Davis's 2025 materiality study reported that the Voynich parchment is soft — a texture found in books that have been "heavily thumbed." Books that are decorative objects, shelf displays, or long-term archival items do not develop this texture. Books that are read, consulted, referenced, and carried do.

A hoax intended to be sold as a curiosity or to deceive a buyer would not necessarily accumulate this kind of use-wear. The wear pattern suggests the manuscript was used as a reference document, not displayed as an artifact.

4. The Roman-letter marginal evidence (Davis & Easton, 2024)

Multispectral imaging of folio 1r revealed a hidden inscription: Roman alphabet / Voynich characters / offset Roman alphabet — a single-shift Caesar cipher, one of the simplest decoding attempts possible. This is visible evidence that an early owner (probably 17th century, consistent with the Baresch-Marci-Kircher correspondence) attempted actual decipherment, not mere authentication. Decipherment attempts happen when people believe a text carries content. They do not happen with manuscripts known to be decorative hoaxes.

5. The purpose problem

Every hoax has a purpose. The Voynich hoax hypotheses imply:

For a profit-motivated deception, five scribes using expensive parchment (paper was available in 15th-century Italy) is an enormous coordination and material investment with no guaranteed return. The Voynich's buyer chain (Rudolph II? Unknown owners between ~1600–1912) does not confirm it was ever successfully sold as encoded knowledge. The investment exceeds any plausible hoax return.

Steelmanning the Hoax Case

The hoax position has genuine force:

The hoax case survives but requires: (a) A single underlying method capable of producing all three statistical laws (Zipf + Brevity + Heap)
(b) Five people sharing and consistently applying it with near-zero errors
(c) A plausible motive that justifies expensive parchment and five-person labor
(d) An explanation for use-wear suggesting actual consultation

No published hoax hypothesis addresses all four. Most address (a) partially.

The Verdict (Confidence: emerging)

The five-scribe collaborative structure argues substantially against the hoax hypothesis. It does not close the case — institutional production of a sophisticated cipher or coded text is also collaborative. But it pushes the manuscript toward two most plausible interpretations:

  1. Institutional cipher: a scriptorium or intellectual community producing a real text in a real or constructed language under a cipher, with section specialization reflecting actual expertise divisions.

  2. Institutional constructed language: a community's internally developed communication system — not a hoax at all, but genuine coded discourse.

Both interpretations are consistent with five scribes, low error rates, use-wear, Zipfian statistics, and the marginal decipherment attempts. Neither requires a coordinated deception.

The lone-eccentric hoax (one brilliant faker with a method) remains possible. The five-person coordinated hoax is implausible enough that it should be ranked below cipher and constructed-language hypotheses.

Cross-Realm Connections

Cryptography × Epistemology: The five-scribe argument is an example of inference to the best explanation operating under genuine ambiguity. All the evidence is consistent with both "real language under cipher" and "sophisticated collaborative hoax." What distinguishes them is the cost of generating the evidence under each hypothesis. The cipher hypothesis has lower coordination cost, lower purpose cost, and makes better sense of the use-wear. Cross-link to concept fermi paradox: observing a statistically language-like signal does not tell you whether it is language — the same epistemological wall. See also concept godel incompleteness: without external ground truth, statistics alone cannot close the question.

History × Computing: The section-specialization pattern — five scribes each handling their domain — mirrors parallel computation in a scriptorium. The manuscript's production was not sequential but modular. This is the same organizational pattern that makes the Jacquard loom (tech jacquard loom) powerful: break the pattern into independent components that can be woven simultaneously. A collaborative encoded text produced modularly is structurally analogous to distributed computing, centuries before the concept existed.

Cryptography × ZKP: The Voynich achieves an accidental semantic zero-knowledge proof: it demonstrates the existence of a generating mechanism (coherent statistics, domain organization, authorial intent evident in section structure) without revealing what the mechanism is. The 2025 Gödel-in-cryptography breakthrough (concept zero knowledge proofs) constructs formal ZKPs where the existence of a simulator is unprovable. The Voynich is the accidental inverse: the existence of a key is unprovable with current tools, but the existence of something to be keyed is statistically self-evident. The manuscript proves information was produced; it refuses to prove what the information was.

Open Questions

  1. Has any research group run a quantitative coordination cost analysis — estimating how many collaborative trials would be needed to accidentally produce Zipf + Brevity + Heap compliance across all five scribes' sections simultaneously?

  2. The use-wear is a promising but unquantified claim. Is there a methodology (spectrometric analysis of page surface wear patterns, page-edge fiber fragmentation mapping) that would allow the Davis 2025 softness observation to be made numerically comparable to manuscripts of known reference and display histories?

  3. Could the section specialization pattern (Scribe 2 exclusively handles astronomy; Scribe 3 exclusively handles balneology) be tested against Medieval scriptorium production conventions — to determine whether this division-of-labor pattern is typical of institutional manuscript production or atypical enough to require explanation?

Key Facts

Key Sources

See Also