The Voynich Manuscript, also known as the Voynich Codex, is perhaps one of humanity’s most enigmatic texts.
With its 204 illustrated pages and incomprehensible content, it earned the nickname “the book that no one can read”.
The mysterious manuscript was Carbon-14 dated to the early 15th century, which was confirmed by realizing that it was compiled by Dominican nuns as a reference source for Queen Maria of Castile.
Throughout its existence, the Voynich Codex has been the subject of intense study by many amateur and professional cryptographers, all of whom have failed to crack a single word.
A team of computer scientists from the University of Alberta, Canada, led by Greg Kondrak, used natural language processing algorithms to try to unravel, at least, the language in which the ancient text was written, which remained a mystery until today.
Based on previous studies, the scientists started from the most widespread hypothesis that the manuscript uses a defined alphabet and tried to create a system capable of identifying the grammar of this text.
Using samples from the Universal Declaration of Human Rights in more than 400 dialects and languages spread around the world, the system developed by the scientists concluded that the manuscript was most likely written in an ancient form of Hebrew, and then encoded by the author or authors.
The algorithm found that just over 80% of the words in the text existed in the ancient Hebrew language. Next, the scientists checked the grammar of a section of the manuscript.
“She made recommendations to the priest, man of the house, myself and the people,” reads the first sentence of the manuscript, roughly translated, if the algorithm is correct.
“It’s a strange sentence, but it makes sense,” Kondrak said.
The researchers are still looking for Hebrew language experts to validate the artificial intelligence discovery and try to decipher the rest of the cryptic text, but they say identifying the manuscript’s original language was the first step.
Discovering the original language of a manuscript may seem simple, but it is not.
The advantage of using an artificial intelligence algorithm is that researchers don’t need to gather hundreds of experts in every language on the planet to try to unravel the origin of words in the text. The algorithm does the work by itself and much faster.
The University of Alberta team also plans to bring this technology to the study of other ancient manuscripts.