The Voynich text, a mysterious 15th century manuscript sometimes called "The world's most mysterious book" has never been translated. Not only are the words unfamiliar, but even the alphabet itself is unique among all the world's documents. This has lead many people, both amateurs and professionals alike, to believe that this 240-page document is a hoax.
New research shows, however, that the Voynich document holds a message in a language never seen before or since.
"The text is unique, there are no similar works and all attempts to decode any possible message in the text have failed. It's not easy to dismiss the manuscript as simple nonsensical gibberish, as it shows a significant [linguistic] structure," Marcelo Montumurro of the University of Manchester in England said.
This conclusion was based on statistical analysis of the text, which consists of counting the number of times that characters or sequences appear in a document. This basic technique is used in code-breaking on Earth and in searching for intelligent signals from outer space, hidden in the natural backdrop of electromagnetic radiation.
Montumurro decided to expand this method, in a way which had never before been tried. His team focused their study on the frequency of word clusters, hoping to uncover the presence of meaning through association. They measured the "entropy" of words - how often they clustered together - and compared that to a randomly-mixed sample. By measuring that difference and multiplying by how often the word is used, the researchers were able to quantify how much information the word contains.
Trials of the technique in 2009 using Moby Dick and On the Origin of Species as text documents picked up the word "whale" as being critical to the novel and "species" for the founding document of evolution. Certain words like this were identified in the Voynich argument, clustered into what appear (from pictures) to be sections of the document. For instance, the same words appear frequently in parts related to herbs and pharmacology, but not as often in chapters related to astrology.
"[These words are] the strongest-connected linguistically and also at the level of their pictorial representations - they're the only two sections that have these plants. Our analysis is the first one that actually links these sections only by their linguistic structure," Montemurro said.
This manuscript was housed in an Italian monastery when it was purchased by Wilfred Voynich, a book dealer, in 1912. In addition to the mysterious text, the book also contains many images of plants, astrological diagrams and nude nymphs.
The lengths of word strings containing high concentrations of these words, called the scale domain, was measured not just in texts from various human languages, but also the Fortran computer language and the DNA of yeast. The average string lengths was around 10 for the yeast, 300 for Fortran, 500-700 for modern languages and 800 for the Voynich document.
