Much like the mathematical precision of music, it should come as no surprise that the evolution of human languages follows mathematical patterns, patterns that can even be seen in languages that haven’t been deciphered, including the text that appears in the mysterious 15th century work known as the Voynich manuscript.

Written in the 15th century, the 240-page Voynich manuscript has confounded researchers for decades: composed in an unknown language with accompanying illustrations, the codex is named after antiquarian Wilfrid Voynich, who purchased the book in 1912. Although Voynich’s attempts to have the manuscript analyzed by scholars were unsuccessful, a 2018 attempt to decipher the text using AI found that the majority of the words may be ancient Hebrew, although no definitive conclusion has been reached regarding this. The strange book eventually found its way into Yale University’s Beinecke Rare Book & Manuscript Library, catalogued under call number MS 408.

While the meaning of the individual words found in the text remain a mystery, the fact that the overall work is composed of a structured language is apparent, allowing researchers to perform a statistical analysis of the cipher; to that end it has been discovered that whatever language is represented in the manuscript, it follows a relationship of word patterns seen across the majority of human languages.

Although the phenomenon of statistical patterns in natural language had been observed earlier in history, the concept of a mathematical correlation regarding the frequency of the use of certain words in a given language was popularized in 1932 by linguist George Kingsley Zipf, who observed that “the frequency of a word seems to be a power law function of its frequency rank,” meaning that each word’s place on the list has a direct correlation to how often it is actually used, according to a 2016 study that applied what is now known as “Zipf’s law” to 50 different languages.

As an example, the most frequently-used word in English–the word “the”–is used twice as often as the second-most used word (“of”), and appears three times as often as the third-most used word (“and”), and so on. This pattern also holds for the majority of human languages, including ones that have yet to be deciphered.

Although the reason as to why languages follow the pattern seen in Zipf’s law is currently unknown, it has been established that the text of the Voynich manuscript follows the same pattern, indicating that it is written in a legitimate language and not a hoax simply composed of a collection of random symbols.

“Precise features of Zipf’s law in languages do not emerge in simple random sequences and generally require interplay between multiplicative and additive processes,” according to a 2013 study. “Moreover, Zipf’s law was discovered centuries after the accepted date of creation of the Voynich text,” with the velum the book is made of carbon dated to sometime between 1404 and 1438. “Thus, proposed solutions like the use of sixteenth-century cipher methods, although not impossible, can hardly account for the presence of Zipf’s law in the Voynich text.”

“Interestingly, the network of relationships that we obtained showed that related words share similar morphological patterns, either in their prefixes or suffixes,” the study added. “This fact suggests that any underlying code or language in the Voynich manuscript has a strong connection between morphology and semantics, recalling scripts where–as in the cases of Chinese and hierographical Ancient Egyptian–the graphical form of words directly derives from their meaning.”

While this revelation doesn’t reveal what the writing in the Voynich manuscript says, it does indicate that it is indeed written in a bona-fide language that could be deciphered, provided researchers can determine what language is involved.