In 1921, a London rare book dealer named Dr. Wilfred M. Voynich toured the United States with pages from a mysterious ancient manuscript that he claimed would “startle the scientific world.”

"When the time comes," Voynich said, "I will prove to the world that the black magic of the Middle Ages consisted in discoveries far in advance of 20th-century science.”

The Voynich Manuscript—as it came to be known—appears in many ways to be a typical medieval treatise like those produced by alchemists in the 14th and 15th centuries. The manuscript’s pages are crowded with intricate drawings of plants and herbs, diagrams of stars and planets, and even some racy doodles of nude women bathing.

But the Voynich Manuscript veers from the “typical” in one very significant way. It’s written in an entirely unknown and indecipherable language. The manuscript contains over 200 pages written in an alphabetic-looking script that is either an encoded text, an invented language or an ingenious medieval hoax.

Over the centuries, countless cryptologists have searched for the secret key to unlocking the Voynich Manuscript—including World War II codebreaker Alan Turing and the FBI—but none have succeeded. The world’s leading expert in the Voynich Manuscript calls it “the main unsolved problem in the history of cryptography.”

Science journalist Garry Shaw, who researched the Voynich Manuscript for his book, Cryptic: From Voynich to the Angel Diaries, the Story of the World's Mysterious Manuscripts, says there’s something about the inscrutable manuscript that “sucks you in.”

“People love a good mystery,” says Shaw. “And like all good mysteries, it seems like if you put enough effort into [the Voynich Manuscript], you could probably decipher it. Maybe if I spend a bit of time staring at these images, staring at this writing, maybe, just maybe I can figure out what's going on here. I think that's what grabs people, really.”

Voynich, the rare book dealer, came into possession of the manuscript in 1912, when he bought it from the Jesuit College at Frascati near Rome, according to Yale University's Beinecke Rare Book and Manuscript Library, which has housed the manuscript since 1969. There was a letter tucked into the manuscript that claimed to identify the author of the mystery volume as Englishman Roger Bacon, a 13th-century Franciscan friar.

“Voynich called it the ‘Roger Bacon cipher manuscript’ because he believed it was created by Bacon, who had a strong association with magic and also wrote about secret writing methods,” says Shaw.

The “Bacon theory” was one of hundreds in circulation until 2011, when researchers at the University of Arizona used radiocarbon dating to confidently date the calfskin pages of the Voynich Manuscript to the early 15th century, 200 years after Bacon. The ink also dated to the 15th century, making it highly unlikely that the manuscript was a modern forgery written on ancient parchment.

The full text of the Voynich Manuscript is viewable online, courtesy of the Beinecke Library. The Voynich Manuscript has more than 230 pages, some of which can be folded out to reveal multi-page composite images—a rare feature for a 15th-century text.

The physical manuscript is roughly the size of a paperback novel, just 6.2 inches by 8.8 inches. Even though its language is unreadable, the Voynich Manuscript appears to be a compendium of knowledge about the natural world. There’s no table of contents, but mystical images throughout the Voynich Manuscript suggest at least six chapters or “sections”:

As early as the 17th century, people thought that the Voynich Manuscript was an encrypted text. In 1639, an alchemist in Prague named Georgius Barschius was in possession of the manuscript, but couldn’t make heads or tails of its alien text. Barschius sent copies of the manuscript to the best cryptographic mind of his day, Athanasius Kircher, who claimed to have translated Egyptian hieroglyphics.

“Now since there was in my library, uselessly taking up space, a certain riddle of the Sphinx, a piece of writing in unknown characters, I thought it would not be out of place to send the puzzle to the Oedipus of Egypt to be solved,” wrote Barschius.

Encoded letters and manuscripts were common in the medieval period, especially among diplomats trying to protect state secrets. But Shaw points out that most medieval codes are relatively easy to crack using simple substitution ciphers (each letter or number represents a different letter or number, like a decoder ring).

“Some alchemist manuscripts were also ciphered, but in those cases they're not enciphering full manuscripts, just individual words or sentences,” says Shaw. “So you don't get massive amounts of encipherment. And again, those ciphers are normally very easy to break, because they’re intended to stop prying eyes from looking at this information and stealing it.”

If the Voynich Manuscript is a cipher text, then it’s more complex than any that have existed in ancient or modern times. A simple substitution code is out of the question. The writing appears to consist of a legitimate alphabet with letters and words, but scholars can’t even agree how many individual letters or symbols there are, with claims ranging from 34 to 70.

One of the first modern cryptographers to attempt to crack the Voynich code was William Newbold, a professor of Latin and philosophy at the University of Pennsylvania, who, in the early 1900s, spent the last years of his life hunched over a microscope trying to make sense of the hand-inked script.

“Newbold’s solution was that it was a type of microscopic shorthand,” says Shaw. “He believed that if you looked at every single Voynich character with a microscope, you could see little shorthand Greek writing all the way around it. The theory was quickly debunked after Newbold’s death, since it would have required Roger Bacon—who Newbold claimed was the author—to have invented the microscope in the 13th century.”

Another popular theory is that the text of the Voynich Manuscript isn’t a code, but an entirely new language. Statistically, the distribution of words in the Voynich Manuscript matches the natural distribution of words in a real language. It’s not a random sequence of symbols. There are clusters of words that only appear in the “herbal” section, for example, but not the “astrological” section.

In the 1940s, William and Elizabeth Friedman were two of the preeminent codebreakers of their day, responsible for deciphering Japan’s “Code Purple” during World War II. After decades of studying the Voynich Manuscript, including analysis with early computers, the Friedmans were convinced that it was an invented language.

Shaw says that there was indeed an “invented language craze” that swept Europe in the 17th century, but the dating of the Voynich Manuscript to the early 15th century makes the Friedmans’ theory highly unlikely.

Despite centuries of intense interest in the Voynich Manuscript, the meaning of the text remains elusive. Serious scholars and obsessed amateurs have claimed to translate the text into nearly every known ancient and modern language, but those claims have all proven false.

According to René Zandbergen, a foremost Voynich expert, “If you just heard or read that the [manuscript] has been solved, then rest assured: The text of the Voynich MS has NOT YET been solved.”

One possible explanation for the continued failure to translate the Voynich Manuscript is that there’s nothing to translate. The text might be an elaborate medieval hoax.

“People in the past were just as interested in ripping people off as they are today,” says Shaw, citing a 17th-century cipher text in the British Library called The Subtlety of Witches that was decoded to reveal a series of random Latin words taken from the dictionary. “There’s an element of performative secrecy where people use the idea of mystery to try and sell something.”

Modern handwriting analysis shows that five different scribes worked on the Voynich Manuscript. Maybe they conspired to create a phony collection of knowledge locked away in a secret language of their own invention. Or maybe, as some Voynich fans have theorized, it’s an alien guidebook to planet Earth left behind by an ancient extraterrestrial visitor.

“I would love it to be some ingenious cipher associated with someone famous, or for it to be a wonderfully fantastic artificial language that predates artificial language craze by centuries,” says Shaw, who admits that he leans toward the theory that the Voynich Manuscript is a medieval hoax. “It’s such a fabulous mystery, though, and I hope I’m wrong.”