The Voynich manuscript has resisted every code-breaker for more than a century

A small illustrated book bound in goatskin sits inside a climate-controlled vault at Yale University, its roughly 240 vellum pages filled with an alphabet nobody living has ever read. Named for the Polish book dealer who bought it in 1912, the volume’s parchment carbon-dates to the early 1400s, yet its script matches no known language, living or dead. Cryptographers who broke Axis codes in two world wars examined its pages and made no progress. Its pictures of unidentifiable plants, nude women in bathing pools, and elaborate star charts only deepen the question of what the book was ever meant to do.

A Manuscript With a Documented Paper Trail

Unlike many mystery artifacts, this one has a traceable chain of custody stretching back nearly four centuries. Its first confirmed owner was Georg Baresch, a 17th-century alchemist in Prague who was baffled enough by the volume that he sent copied pages to the Jesuit scholar Athanasius Kircher in Rome, hoping for a translation. The book eventually passed into Kircher’s own collection at the Collegio Romano, where it likely picked up the goatskin binding it still wears. It resurfaced in 1912 when Wilfrid Voynich, a Polish-born rare book dealer, purchased it along with a batch of other volumes from a Jesuit college near Rome. Voynich spent the rest of his life trying to get it deciphered and never succeeded. His widow, and later a New York rare book dealer, kept the manuscript until Yale University’s Beinecke Rare Book and Manuscript Library acquired it in 1969, cataloguing it simply as MS 408.

What the Radiocarbon Dating Actually Showed

In 2009, researchers at the University of Arizona radiocarbon-dated small samples of the vellum and got consistent results across every sample tested: the calfskin was produced between 1404 and 1438. Later protein testing in 2014, discussed in the manuscript’s Yale catalog record, confirmed the parchment came from calfskin drawn from at least fourteen or fifteen individual hides, and multispectral imaging showed the skin had not been written on before, ruling out the idea that the book reused an older, scraped-clean text. Chemical analysis of the ink found iron gall ink consistent with the period, and the blue, white, and red-brown paints were identified as ground azurite, an egg-white and calcium carbonate mixture, and iron-rich ochre, all inexpensive pigments common to the era. Stylistic details in the drawings point toward Italy during the Renaissance as a likely place of origin, though nothing in the text itself confirms that.

Six Illustrated Sections, None Fully Explained

Because the text cannot be read, scholars organize the manuscript by its pictures instead. A 126-page herbal section shows plants that do not match any known species. A 17-page astronomical section contains zodiac-style diagrams surrounded by nude female figures each holding what looks like a labeled star. A 20-page balneological section is filled with small nude women bathing in pools connected by pipework. A 14-page cosmological section includes a six-page foldout nicknamed the “Rosettes” page, showing nine circular islands linked by causeways. A 16-page pharmaceutical section pairs labeled plant parts with jar-like containers, and a 25-page recipes section breaks continuous text into short, star-marked paragraphs. Eight additional pages contain only text, and at least 14 folios are missing entirely from the surviving copy.

Codebreakers Who Struck Out

The manuscript has drawn the attention of some of the most capable cryptanalysts in modern history. American codebreaker William Friedman, his wife Elizebeth Friedman, researcher Prescott Currier, and British codebreaker John Tiltman all studied it at length, applying techniques developed for military cryptography, and none produced a verified translation. Friedman’s team went as far as transcribing the entire text onto IBM punch cards in the 1940s to make it machine-readable for statistical analysis, a significant undertaking for the era. In 1962, Elizebeth Friedman described further attempts at pattern-based decryption as “doomed to utter frustration,” a judgment that has largely held up against every subsequent claim of a solution.

What the Statistics Argue Against a Hoax

Modern computational analysis has added a layer of evidence that complicates the idea that the manuscript is meaningless gibberish. Word frequencies in the text follow Zipf’s law, the same statistical pattern found in real languages, where the most common word appears roughly twice as often as the second most common. A 2014 study led by Diego Amancio at the University of São Paulo used network analysis of word relationships and found the manuscript’s structure matched genuine books in about 90 percent of the comparisons run. Linguists Claire Bowern and Luke Lindemann measured a property called second-order conditional entropy and found the text’s character sequences were more predictable than any known natural language, yet still inconsistent with a simple substitution or polyalphabetic cipher. Separately, manuscript specialist Lisa Fagin Davis identified subtle handwriting variations indicating the book was produced by five different scribes, suggesting a coordinated group effort rather than the work of a single hoaxer.

Theories That Remain on the Table

Without a confirmed translation, every explanation for the manuscript remains a hypothesis. Researchers have proposed that it encodes a natural language using an unfamiliar cipher, that it represents an invented constructed language, that it is an elaborate hoax designed to fool a wealthy buyer, or that it functioned as a reference work such as a folk index of herbal and astrological knowledge. Its physical wear, described by conservators as soft and heavily handled, suggests it was used often and likely served some practical purpose rather than a ceremonial one. More than a century after Wilfrid Voynich rediscovered it, and roughly six centuries after it was written, the text remains exactly as unread as the day scholars first tried to crack it.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview