17 December 2003
Using a Cardan grille the manuscript could have been written in three months.
© G. Rugg
A strange sixteenth-century book may be cunningly crafted nonsense, says a computer scientist. Gordon Rugg has used the techniques of Elizabethan espionage to recreate the Voynich manuscript, which has stumped code-breakers and linguists for nearly a century1.
"I've shown that a hoax is a feasible explanation," says Rugg, who works at Keele University, UK. "Now it's up to believers in a code to produce evidence to support their ideas." He suspects that English adventurer Edward Kelley produced the Voynich to con Rudolph II, Holy Roman Emperor and collector of antiquities, out of a fortune in gold.
The explanation is plausible, but not conclusive, say Voynich scholars. "It's an excellent piece of work," says Philip Neal, a former medievalist based in London. "I haven't given up hope that the manuscript contains meaning, but this makes it less likely."
The Voynich manuscript is often described as the world's most mysterious book. It is hand-written in a unique alphabet, about 250 pages long, and contains pictures of unrecognizable flowers, naked nymphs and astrological symbols.
The manuscript first appeared in the late 1500s, when Rudolph II bought it in Prague from an unknown seller for 600 ducats - about 3.5 kilograms of gold, worth more than US$50,000 today. The book passed from Rudolph to noblemen and scholars, before disappearing in the late 1600s.
It surfaced again around 1912, when US book dealer Wilfrid Voynich bought it. The manuscript was donated to Yale University after Voynich's death.
No one has worked out whether Voynichese is a code, an idiosyncratic translation of a known tongue, or gibberish. The text contains some features that are not seen in any language. The most common words are often repeated two or three times, for example - the equivalent of English using 'and and and' - giving weight to the hoax theory.
On the other hand, some aspects, such as the pattern of word lengths and the ways in which characters and syllables occur with each other, are similar to real languages. "Many people have believed that it is too complicated to be a hoax - that it would have taken some mad alchemist years to get such regularity," says Rugg.
But this complexity could have been produced easily, Rugg demonstrates, with an encryption device invented around 1550 called a Cardan grille. This is a table of characters. Moving a piece of card with holes cut in it over the table makes words. Gaps in the table ensure different-length words.
Using such grilles on table of Voynichese syllables, Rugg has produced a language with many, although not all, of the manuscript's features. About three months' work would have been enough to produce the entire book, he says.
"It's an interesting angle, but it's too early to say whether it's correct," says Nick Pelling, a computer programmer based in Surbiton, UK, who also studies cryptography and the Voynich.
To prove that the manuscript is a hoax, one would need to produce entire sections using this technique, says Pelling. Tweaking the grilles and tables should make this possible, reckons Rugg.
It seems that the Voynich resists deciphering attempts because its author knew enough about codes to make the text plausible yet hard to crack.
The book appears to contain cross-referencing, just the kind of thing that cryptographers look for. The characters of Voynichese are also ambiguously written, so it is hard to work out how large the alphabet is, and drawing naked figures makes it impossible to date the text by styles of dress.
I haven't given up hope that the manuscript contains meaning
Philip Neal, medievalist
The chief suspect for producing the book is known to have used Cardan grilles. As well as a cryptographer and inventor of languages, Edward Kelley was a forger, mystic, alchemist, mercenary and wife-swapper. He travelled to Prague to meet with Rudolph in 1584, and may have sold him the manuscript then. Kelley was lost to history after escaping from prison at the end of the sixteenth century.
"If it's a hoax Kelley is the obvious candidate," says
Neal. But he adds that Rudolph bought many alchemical texts that
are far cruder forgeries than the Voynich manuscript. "Rudolph
was easily fooled. If the Voynich was a hoax by Kelley, it looks
a bit like overkill," Neal says.
Rugg, G. An elegant hoax? A possible solution to the Voynich
manuscript. Cryptologia, (in the press).