Weird History
The Voynich Manuscript
A 600-year-old book filled with strange plants, astronomical diagrams, and bathing women — written in a language no one has ever cracked.
Trace the connectionsWhy this rabbit hole matters
The Voynich Manuscript is a 240-page illustrated book carbon-dated to the early 15th century, written in an unknown script and language that has defeated every attempt at decipherment for over 600 years. Its pages depict unidentifiable plants, astronomical charts, bathing women in strange pools, and elaborate cosmological diagrams — all accompanied by a flowing text that has statistical properties consistent with a real language but matches nothing known. Held at Yale's Beinecke Rare Book Library, it has been seriously studied by NSA cryptographers, academic linguists, and professional codebreakers. Nobody has cracked it.
First appearing in historical record in 1639 in a letter from Georg Baresch, a Prague alchemist who couldn't read it, the manuscript was purchased in 1912 by rare book dealer Wilfrid Voynich from the Jesuit College at Frascati. Carbon dating in 2009 placed its vellum creation between 1404 and 1438. The text follows statistical patterns of natural language — it has word-frequency distributions matching Zipf's law, consistent 'grammar,' and a vocabulary of roughly 35,000 word tokens — but matches no known language, code, or constructed language. The illustrations are equally baffling: the botanical section shows plants that match no known species; the astronomical section depicts cosmological models that match no known tradition.
Connections to follow
Decipherment Attempts and Failures
The history of Voynich decipherment attempts is a graveyard of reputations and careers. William Friedman, arguably the greatest cryptanalyst in American history, spent years on it privately and concluded only that it might be an artificial philosophical language. In 2019, Dr. Gerard Cheshire of the University of Bristol claimed it was proto-Romance language — the claim was retracted after linguists demolished it. In 2017, a paper claimed it encoded an Aztec herbal — also rejected. The pattern is consistent: each proposed solution explains some features while failing badly on others, suggesting either extreme complexity or the fundamental absence of a real message.
The Hoax Hypothesis
The most parsimonious explanation for the Voynich Manuscript may be that it is an elaborate hoax — a book of nonsense designed to look like an important alchemical or medical text in order to sell it to a wealthy collector. The leading candidate for its creation is Edward Kelley, the notorious alchemist-forger who claimed to channel angelic communication for John Dee in Prague in the 1580s. Computer analysis of the manuscript's statistical properties in 2004 found features consistent with 'glossolalia' — the non-random but meaningless babbling of people claiming to speak in tongues — suggesting a human creating the impression of language without its content.
Undeciphered Scripts Throughout History
The Voynich Manuscript belongs to a broader category of human communication that remains unreadable: undeciphered scripts and languages. Proto-Sinaitic (the ancestor of all alphabets) was only partially deciphered in the 20th century. Linear A, the Minoan writing system, remains completely undeciphered. The Indus Valley script, used by a civilization of millions, has resisted every decipherment attempt. The Rongorongo script of Easter Island was never decoded before its speakers died. Each of these represents a lost intellectual tradition — entire ways of thinking that may be permanently inaccessible.
The Statistical Analysis: Is It Even a Language?
The Voynich Manuscript's text has been subjected to exhaustive statistical analysis — and the results are simultaneously compelling and baffling. The text follows Zipf's law: in natural languages, the most common word appears twice as often as the second most common, three times as often as the third, and so on — and Voynich text matches this distribution. It shows consistent positional grammar (certain word-forms appear only at line beginnings, others only at ends), and its vocabulary distribution matches natural language better than cipher text. Yet the text has unusually low entropy — it is less random than natural language — suggesting either a highly constrained natural language, a constructed language, or meaningless text generated by a mechanical device that mimics language structure.
The Known Owners: Six Centuries of Provenance
Unlike most mysterious artifacts, the Voynich Manuscript has a reasonably well-documented ownership chain for approximately the last 400 years. The earliest confirmed owner is Holy Roman Emperor Rudolf II (1552-1612), an obsessive collector of curiosities, who reportedly paid 600 gold ducats for it. It passed through the collections of Jacobus Horcicky (Rudolf's physician), then to Jesuit scholar Athanasius Kircher in 1666 via letter from Bohemian physician Jan Marek Marci, who speculated it had been acquired from Emperor Rudolf. In 1912, bookseller Wilfrid Voynich purchased it from the Villa Mondragone collection in Frascati, Italy. He sold it to American book dealer H. P. Kraus in 1961; Kraus donated it to Yale's Beinecke Library in 1969, where it has remained. Radiocarbon dating of the vellum places its creation between 1404 and 1438.
Modern Scientific Tools and What They Have Found
Since the manuscript arrived at Yale, it has been subjected to an escalating series of scientific analyses. Radiocarbon dating (2009) placed the vellum manufacture between 1404 and 1438. X-ray fluorescence analysis identified the pigments — lapis lazuli, copper-based blue, red lead — all consistent with 15th-century European manuscript production. Multispectral imaging revealed erased text and different layering in some illustrations, suggesting revisions. In 2017, researchers found an 'inking order' suggesting a scribal process, not random generation. A 2019 analysis claimed to identify the text as encoded Hebrew with transposed consonants — a claim disputed by most Voynich linguists. None of these analyses has broken the text. The tools have established authenticity and narrowed the field of hypotheses without resolving the central mystery.