Somewhere deep inside the bowels of Yale's Beinecke Rare Book and Manuscript Library – the Ivy League institution's own cemetery of lost books – lies a tome that experts have studied for centuries, but which has yet to be understood by a single soul.
The book has no known author or official title; Yale librarians simply refer to it as manuscript MS 408. But thanks to its peculiar language, symbols and diagrams – often strangely familiar, but insistently elusive in meaning – it has intrigued and frustrated anthropologists, linguists and mathematicians for centuries: even the elite cryptologists at the US National Security Agency drew a blank, after they spent years trying to decode it in the 1950s. And the time that some researchers have dedicated to the problem seems all the more remarkable given the possibility that, for all the complexity and consistency of the script it contains, it could simply be an elaborate hoax.
Written in an as yet undecipherable language, with unknown letters or "glyphs" arranged into a form of seemingly consistent but unintelligible syntax, the book is commonly referred to as the Voynich manuscript, after the Polish-American bookseller Wilfrid Voynich, who acquired it in 1912. Its history, however, begins long before.
Although the earliest suggested owner is Rudolf II, the 16th-century emperor of Bohemia, the first that we know of for sure is Georg Baresch, a 17th-century alchemist from Prague, who was so perplexed by the book that he sent it to Jesuit scholars in the hope that they might translate it.
They failed, but they did pass it on to the Roman Jesuit University, from where it was whisked away to Frascati, near Rome, in 1870 to keep it safe from Vittorio Emanuele's marauding soldiers. It was bought by Voynich, and then donated to Yale in 1969.
Recently, however, experts have arrived at what – in Voynich terms, at least – must count as a significant breakthrough: while we still don't understand a word, at least we know how old it is. Carbon-dating by scientists at the University of Arizona has allowed them to declare that the manuscript was prepared from animal skin in the early 1400s – making it roughly 100 years older than previously thought.
The tests were done after Yale finally allowed the scientists to snip off tiny pieces from four different pages, selected at random. "The results seem to show quite clearly that the parchment for the book is from the early 1400s, between 1404 and 1438," says Dr Greg Hodgins, who works in the physics and anthropology departments at the University of Arizona. "And the fact that all four sections were dated to the same time appears to discount suggestions that the manuscript was added to over a period of many years or centuries."
These results only show when the parchment for the book was obtained, not when it was written. However, previous ink analysis done by the McCrone Research Institute in Chicago suggests that it was placed on the parchment while it was relatively fresh.
"By coming up with such a narrow time frame, we've effectively eliminated most of the theories about who wrote it," says Dr Hodgins. "The carbon-dating result also allows us to focus on what kind of scientific knowledge and encryption was around in this period."
Professor Gonzalo Rubio, a specialist in ancient languages at the University of Pennsylvania, agrees that the carbon-dating result is significant. "This shows us that it's not a forgery," he says. "It wasn't written by Voynich himself, as some people suspected. It's a genuine artifact from the early 15th century." It also, he points out, eliminates the popular theory that the book was created by the noted 13th-century polymath Roger Bacon.
Most theories about the book's meaning are inevitably informed by its illustrations, says Prof Rubio. These pictures, drawn in various shades of green, brown, yellow, blue, and red ink, are – like the script used in the manuscript's 240 remaining pages – unique. Yet while the words cannot be read, the illustrations provide a clue about the nature of the book. They suggest that the book was a scientific text, mostly an illustrated herbal manual with some additional sections on astronomy, biology and pharmaceuticals. The script itself is widely believed to be about alchemy, the medieval science with metaphysical and magical overtones, whose practitioners sought ways to turn base metals into gold.
But what of the language it is written in? Some of the glyphs resemble Latin letters, suggesting a Voynich "alphabet" of around 20-30 different characters. These are arranged into word-like blocks up to 10 letters long, with a total word count of around 35,000. However, some scores of pages are thought to be missing.
Many researchers have speculated that the strange alphabet was used to hide information that might have been heretical, suggesting that it was produced by transforming a European language through a cipher. But simple ciphers of the type used in the 15th century would almost certainly have been cracked by now.
Other theories argue that the text's meaning is concealed in tiny markings on the individual characters themselves, or that it contains a naturally occurring, non-European language that has been rewritten with an invented alphabet. In 2003, Gordon Rugg, a computer scientist at Keele University, claimed to have created a bogus language similar to Voynichese by using a "grille" to place invented letters on a page at random. But doubts about the claims have since arisen. "I don't really think he proved or disproved anything," says Dr Hodgins.
One of the theories that has gained ground in recent years, says Prof Rubio, is that the manuscript employs steganography to conceal its contents. This means that some or even most of the text is nonsense, and that only parts or even individual characters form part of the language. If this method was indeed employed, in addition to a cipher, then translating the contents might be exceedingly difficult.
Put on the spot, Dr Hodgins says: "It's either a secret alchemical text, with the pictures telling a story – or, as some have suggested, it was created, or invented, to enable its author to profit from it by selling it as a precious manuscript."
Prof Rubio is more sceptical, noting that despite sharing key structural similarities with known languages such as English and Latin, there is at least one troubling aspect about the book's structure. "For me, there's something strange about the text," he says. "The things we know as 'grammatical markers' – things that occur commonly at the beginning or end of words, such as 's' or 'd' in our language, and that are used to express grammar, never appear in the middle of 'words' in the Voynich manuscript. That's unheard of for any Indo-European, Hungarian or Finnish language.
"I would venture that the manuscript was a prank, something made for fun or even done to send up alchemy texts around in the period. But the thing is so unique, it's hard to know: there's nothing to compare it with."
He also denies the theory that if there were language present in the book, it would have been decoded by the experts. "The NSA cryptologists were working in the 1950s," he points out. "Computers have moved on since then." But he says that establishing once and for all whether there really is anything to be understood would require an up-to-date and dedicated team of linguists and computer scientists working together.
"I don't think that's going to happen," he concludes, "because experts might spend 30 or 40 years of their lives on this and still not come up with an answer. And I don't think people will be prepared to do this."
In short, it looks like the mysteries surrounding the planet's strangest manuscript are set to remain for a good while to come.
Also check out these sites for more information on the manuscript.
Day, Michael. 2011. "The Voynich Manuscript: will we ever be able to read this book?" The Telegraph. Posted: May 24, 2011. Available online: http://www.telegraph.co.uk/science/science-news/8532458/The-Voynich-Manuscript-will-we-ever-be-able-to-read-this-book.html