The Book No One Can Read

In a climate-controlled vault at Yale's Beinecke Rare Book and Manuscript Library sits a small volume, roughly the size of a modern hardback, bound in soft vellum and filled cover to cover with an alphabet nobody can read. Its pages carry drawings of plants that match no known species, star charts with unfamiliar constellations, and dozens of nude women bathing in green liquid that flows through networks of pipes. Every line of text is written fluently, confidently, in a script that resembles nothing else ever produced by human hands — and after more than a century of sustained attention from professional cryptographers, medieval historians, and computer scientists, not one word of it has been reliably translated.

This is the Voynich manuscript, catalogued at Yale as Beinecke MS 408. It is, by most measures, the most famous unsolved problem in the history of writing.

What the Object Actually Is

The manuscript measures about 23.5 by 16.2 centimeters and today runs to roughly 240 surviving vellum pages, bound into quires; codicologists estimate the original ran to around 272 pages before some were lost. The text totals roughly 38,000 words drawn from about 9,000 distinct word-forms, set down left to right in a fluid, unhesitating hand using an alphabet of 20 to 25 unique glyphs that correspond to no known language, ancient or modern.

In 2009, a radiocarbon dating team at the University of Arizona tested four separate vellum samples and dated the parchment to between 1404 and 1438 — placing its manufacture firmly in the early fifteenth century, during the reigns of the last Plantagenet and early Valois kings. Later analysis, published in 2014, established that the parchment was made from the skins of at least fourteen or fifteen calves, and multispectral imaging has ruled out the possibility that the manuscript is a palimpsest — a reused, scraped-down older document. The ink is iron gall; the paints include ground azurite for blue and a copper-based resinate for green, both materials consistent with the manuscript's stated age. Paleographic analysis by the manuscript scholar Lisa Fagin Davis has identified the handwriting of at least five distinct scribes, suggesting the book was the product of a scriptorium or a small working community rather than a single obsessive author.

Illustrations Without an Answer

The manuscript is conventionally divided into six sections based on their imagery. A herbal section of some 126 pages pairs individual plant drawings with blocks of text, but botanists have never been able to match the plants to real species. An astronomical section of circular zodiacal diagrams runs 17 pages; a balneological section of 20 pages shows nude women immersed in pools linked by elaborate plumbing; a cosmological section of 14 pages includes the so-called "Rosettes" foldout, a large diagram some read as a map of islands, castles, or volcanic terrain. A pharmaceutical section depicts isolated plant parts alongside labeled apothecary jars, and a final "recipes" section breaks continuous text into star-marked paragraphs, as if listing preparations or instructions.

Statistical study of the text has repeatedly found it behaves like real language rather than random noise: word lengths and frequencies obey Zipf's law, and the linguist Claire Bowern has noted that while the letter-to-letter entropy is unusually low, the manuscript's larger structural patterns mirror those of natural languages. Researchers have also identified two statistically distinct "dialects" within the text, dubbed Voynich A and Voynich B, which map roughly onto different sections of the book.

A Trail of Owners, Real and Legendary

The earliest confirmed owner is the Prague alchemist Georg Baresch, documented through a 1639 letter. The manuscript passed to Jan Marek Marci, who sent it in the mid-1660s to the polymath Jesuit scholar Athanasius Kircher, along with a cover letter repeating an already-old rumor: that the book had once belonged to Holy Roman Emperor Rudolf II, who supposedly paid 600 gold ducats for it, and that its author was the thirteenth-century English friar Roger Bacon. Both claims are now considered dubious by scholars, though a faded signature on the manuscript's first folio — possibly that of Jacobus Horcicky de Tepenecz, keeper of Rudolf's botanical gardens — hints that the Prague court connection may be genuine.

The book then sat largely forgotten in the library of the Collegio Romano for roughly two centuries, before passing into the private collection of the Jesuit father Pieter Jan Beckx at Villa Mondragone, near Frascati. In 1912, the Polish-American antiquarian book dealer Wilfrid Voynich purchased a group of manuscripts from the Society of Jesus, this one among them — which is how the book acquired his name. After his death it passed to his widow Ethel, then to Anne Nill, and finally in 1961 to the rare-book dealer Hans P. Kraus, who donated it to Yale in 1969, where it has remained ever since.

A Century of Failed Codebreakers

The manuscript has repelled some of the most capable codebreakers of the modern era. In the 1940s and '50s, a group of NSA and Army cryptanalysts that included William F. Friedman and Elizebeth Friedman — the couple credited with founding modern American cryptology — studied the text at length and made no verified progress. Earlier, in 1921, the American academic William Romaine Newbold claimed to have found microscopic Greek shorthand hidden in the ink strokes revealing a text by Roger Bacon describing biological cells and spiral galaxies; the medievalist John Matthews Manly demonstrated in 1931 that Newbold's "letters" were simply cracks in drying ink.

More recent decipherment claims have followed a similar arc of initial press attention and subsequent scholarly rejection. In 2017 the television writer Nicholas Gibbs proposed the text was abbreviated medical Latin; Lisa Fagin Davis countered publicly that his method "doesn't result in Latin that makes sense." In 2019, the University of Bristol research associate Gerard Cheshire claimed to have identified a "proto-Romance" language compiled by nuns for a Spanish queen; the university itself issued a statement distancing itself from the claim. Other researchers, including Gordon Rugg in 2003, have instead argued the text could be an elaborate, meaningless hoax generated with simple mechanical word-tables — though the specific device Rugg proposed, a perforated grille, was not invented until roughly a century after the manuscript's vellum was made.

Still Unsolved

What survives every round of debunking is the manuscript's sheer statistical strangeness: too structured to be pure gibberish by several measures, too alien to be any known language by every measure yet applied. In 2020, Yale published complete high-resolution scans of the manuscript online, opening it to worldwide analysis. In 2024, new multispectral imaging of ten pages revealed previously invisible marginal details. Yet the core mystery is unchanged. As the current scholarly consensus holds, the Voynich manuscript has never been demonstrably deciphered, and no proposed solution has survived independent scrutiny — which is precisely why, six centuries after someone filled its pages with plants that do not exist, it remains one of the most examined and least understood objects in any library in the world.