THE VOYNICH MANUSCRIPT
A 240-page codex written in an unknown script, carbon-dated to the early 1400s and filled with plants that do not exist. A century of cryptographers has failed to read a word of it.
.jpg){kind=link}
The Voynich Manuscript is a small handwritten book of about 240 surviving vellum pages, named for the rare-book dealer Wilfrid Voynich, who bought it in 1912. It is written left to right in a flowing script of some two dozen recurring characters that matches no known alphabet, and it has never been read. It now sits in the Beinecke Rare Book and Manuscript Library at Yale, catalogued as MS 408 [1].
The physical object is well understood even though the text is not. In 2011 the University of Arizona radiocarbon-dated the vellum to between 1404 and 1438, and analysis of the inks and paints found materials consistent with that period, which rules out a modern fabrication by Voynich himself [2]. The drawings sort into loose sections that scholars label by subject: a herbal section of plants that mostly cannot be matched to real species, an astronomical section of circular diagrams and zodiac figures, a biological section of small figures in what look like bathing pools, and a pharmaceutical section with jars and roots.
The writing behaves in ways that pull against easy answers. Statistical work shows that the script follows Zipf's law and carries word-to-word structure and section-specific vocabulary, the kind of patterning natural languages produce and random gibberish usually does not [3]. At the same time it has features no ordinary language shows, such as words repeated two or three times in a row and a narrow, rigid word structure. Computational studies have found the text statistically closest to real language while stopping short of identifying which one [4].
Every proposed solution so far has failed to convince other specialists. The most cited early claim, William Newbold's 1921 reading of it as the shorthand of the medieval scholar Roger Bacon, collapsed when the cryptographer John Manly showed that Newbold's method could generate almost any text the reader wanted. The mainstream position is not that the manuscript is meaningless, but that its content is genuinely undetermined: a real fifteenth-century book whose script has resisted a hundred years of cryptography, linguistics, and computation.
- [01] Clemens, R., ed. (2016). The Voynich Manuscript. Yale University Press. Full facsimile with the Beinecke provenance and codicology.
- [02] University of Arizona (2011). Radiocarbon dating of the vellum to 1404-1438, reported with the AMS laboratory results.
- [03] Montemurro, M.A. & Zanette, D.H. (2013). Keywords and Co-Occurrence Patterns in the Voynich Manuscript. PLOS ONE 8(6): e66344.
- [04] Reddy, S. & Knight, K. (2011). What We Know About the Voynich Manuscript. Proceedings of the ACL Workshop on Language Technology for Cultural Heritage.
The manuscript's unread text has drawn sharply different explanations. Each has a traceable origin and a documented response. They are laid out in full so you can weigh them yourself.
A real cipher or lost language
One camp holds that the script encodes a genuine message, either a natural language written in an unfamiliar alphabet or a cipher, and points to the language-like statistics as proof that meaning is present [2]. The response: the same statistics are consistent with several incompatible source languages, no proposed key has decoded a passage that other scholars can reproduce, and partial readings such as Stephen Bax's handful of plant names have not extended into running text. Language-like structure narrows the field but has not produced a single verified sentence in a century.
Meaningful-looking but empty
Gerry Rugg argued that a sixteenth-century hoaxer could have produced the text with a Cardan grille, a card with cut-out windows slid over a table of syllables to generate word-like strings quickly and cheaply [1]. This account explains the repeated words and the rigid structure that real languages avoid. Its weakness is the 2011 carbon date, which places the vellum around 1420, well before the grille method Rugg described was documented, so a hoaxer would need an earlier and unrecorded technique.
An invented private language
A third possibility is that the author built a constructed script for private or ritual use, real to its maker but tied to no outside language, which would explain why it resists every known key while still carrying internal structure. It cannot be ruled out, and it is also the hardest to confirm, since a one-person system leaves no second text to check against. Gordon Cheshire's 2019 proto-Romance reading drew wide news coverage and was rejected by Yale and by professional linguists as unsupported, a reminder of how often a confident Voynich solution has arrived and then failed review.
- [01] Rugg, G. (2004). An Elegant Hoax? A Possible Left-to-Right Table Cipher. Cryptologia 28(1). The Cardan-grille hoax proposal.
- [02] Bax, S. (2014). A Proposed Partial Decoding of the Voynich Script. Working paper proposing a natural-language reading of plant names.
- [03] Cheshire, G. (2019). The Language and Writing System of MS408. Romance Studies 37(1). A proto-Romance claim disputed by Yale and by linguists.
INFORMAL READER POLL, NOT DATA. WHICH EXPLANATION DO YOU FIND MOST CREDIBLE?