What is known for sure
Radiocarbon dating places the parchment in the first half of the 15th century, and the ink is consistent with that period. So the book really is old, ruling out the idea that it was forged later for sale.
- 15th centuryDating places the parchment in the first half of the century.
- 17th centuryCorrespondence about it survives: even then, people tried and failed to read it.
- 1912A rare-book dealer buys the manuscript in Italy, and it takes his name.
- NowIt is kept in a university library, has been scanned, and is available to anyone.
Its contents are divided into sections based on the drawings: plants, astronomical circles, bathing scenes with figures, and pages that look like recipes. No plant has been confidently identified as a real species.
Why no one can read it
The text behaves strangely. In some statistical respects, it resembles a natural language: words repeat in patterns similar to those found in ordinary texts. But in other respects, it resembles nothing known.
For example, there are almost no repetitions of a word with one word in between, and very few corrections: whoever wrote it never once lost their place. Some words appear only in one section, while others occur three times in a row, something that almost never happens in languages.
The manuscript has already been deciphered several times.
Claims of a solution appear regularly, almost every year. None has stood up to scrutiny: the proposed reading usually works for a few lines and falls apart across the rest of the text.
The theory that it is a meaningless string of symbols made for sale is still around, too. Its weak point is the same: the text's statistics are too similar to those of a language, and producing such a forgery in the 15th century would not have been easy.
Machine methods have been tried as well. Programs find a structure resembling language, but they do not agree on which language it resembles, and the results depend heavily on the researcher's initial assumptions.
Another difficulty is that there is only one text. Deciphering usually requires a second document with the same content in a known language, and here there is nothing to compare it with.
What is known for sure about the manuscript?
Why are claims of a decipherment not accepted?
Sources
- Radiocarbon dating of the parchment places it at the beginning of the 15th century
- the text's statistical properties partly resemble those of a natural language, but no proposed decipherment has been accepted by specialists
Next in the “Mysteries” series
2 articles in full on the site, 3 more in the app
