The Voynich Manuscript is a 240-page illustrated book written in a script that nobody can read, in a language that nobody can identify, about subjects that nobody can agree on. It has been studied by professional cryptographers, linguists, historians, and amateurs for over a century. None of them have cracked it.
It currently sits in the Beinecke Rare Book and Manuscript Library at Yale University, cataloged as MS 408, quietly resisting every attempt at comprehension.
What It Contains
The manuscript is divided into sections based on the illustrations. There appear to be botanical pages showing plants that do not match any known species. There are astronomical or astrological diagrams with circles, stars, and what might be zodiac symbols. There are pages showing nude women bathing in interconnected pools of green liquid. And there are pages of dense text with no illustrations at all.
The script consists of roughly 20-30 distinct characters, depending on how you count ligatures and variants. The text shows statistical properties consistent with natural language: word length distribution, character frequency patterns, and something resembling grammar. But it does not match any known language.
Radiocarbon dating places the vellum in the early 15th century, between 1404 and 1438. The inks are consistent with that period. Whatever the manuscript is, it is genuinely old.
The Theories
The leading theories fall into three categories, and honesty requires admitting that none is proven.
1. It is a real text in an unknown or encoded language. Some researchers believe the manuscript is written in a natural language using an invented script, possibly a constructed language or a highly unusual cipher. Statistical analysis by linguist Claire Bowern and physicist Marcelo Montemurro has shown that the text has entropy patterns similar to natural languages. If it is a cipher, it is extraordinarily sophisticated for its era.
2. It is a hoax or glossolalia. Computer scientist Gordon Rugg demonstrated in 2004 that text with similar statistical properties could be generated using a Cardan grille (a cipher device known in the Renaissance). Under this theory, the manuscript is meaningless but structured enough to appear meaningful. The illustrations would be deliberate nonsense designed to impress a buyer.
3. It is a specialized text in an unusual tradition. Some scholars have proposed that the manuscript is a pharmacological or alchemical text written in a private script by a practitioner who wanted to keep their knowledge secret. The botanical illustrations, while strange, might represent stylized or regional plants viewed through an unfamiliar artistic tradition.
What Has Not Worked
Every major decipherment claim has failed peer review. In 2019, a researcher claimed to have identified the language as proto-Romance. The claim was widely covered in the media and widely rejected by linguists. In 2017, a team claimed to have identified some words through AI analysis. The results were not reproducible.
The manuscript has also attracted conspiracy theories, connections to Roger Bacon, to the Cathars, to alien visitors. These do not merit serious discussion but they illustrate how powerfully a genuine mystery attracts speculation.
The NSA reportedly attempted to decode the manuscript during the Cold War. They did not succeed.
Why It Matters
The Voynich Manuscript matters because it is a genuine reminder of how much we do not know. We assume that with modern computing, machine learning, and the accumulated knowledge of all human languages, we should be able to read any text. This one proves otherwise.
It may be that the manuscript is unsolvable without additional evidence, a bilingual text, a related document, or a key. It may be that it is a hoax and there is nothing to solve. Or it may be that the solution is sitting right in front of us and we lack the framework to see it.
That uncertainty is uncomfortable, and it is honest.
The Voynich Manuscript remains undeciphered. It is fully digitized and available online through the Beinecke Library at Yale. Researchers continue to study it using computational linguistics and materials analysis, but no consensus decipherment exists.