The Voynich Manuscript: Data Science Takes on the World's Most Famous Undeciphered Document

2025/03/18

The term "data science" has become a household word. Against the backdrop of big data, data analysis utilizing machine learning and AI is flourishing. However, there may be an image that it has little relevance to the humanities, which often deal with subjects that are difficult to quantify. Wanting to demonstrate that potential, I accepted the editor's proposal without hesitation.

The Voynich Manuscript is a parchment manuscript estimated to have been created in Central Europe between the late 15th and 16th centuries. Despite attempts from diverse fields such as the occult, codicology, linguistics, cryptography, and informatics, it remains undeciphered more than 100 years after its discovery. Some research results even suggest it is meaningless gibberish.

It was over 10 years ago that we analyzed this manuscript using text data clustering, but I believe it remains a valid approach today. Gratifyingly, we obtained results showing that the content has a consistent structure and is theoretically decipherable. A complex cipher would not possess such characteristics, and a simple cipher would have been cracked long ago, so I believe it is written in an unknown language. However, since it has not yet been deciphered, the title is "Takes on" rather than "Deciphers."

Nowadays, one might want to pin their hopes on AI, but since there are no other documents with the same characters and the amount of data is too small, it is unlikely to lead to an immediate decipherment. On the other hand, technology is advancing by leaps and bounds, making this an exciting era where we can look forward to the emergence of new methods.

Looking back at the research history of the Voynich Manuscript, the achievements of amateur researchers are prominent alongside those of so-called professional researchers. Therefore, I also introduced trends in scientific research by ordinary citizens (citizen science).

Additionally, Chapter 6 includes a three-way conversation with Hiroshi Aramata, a Keio University alumni well-known as a naturalist, fantasy literature author, and translator. You will be able to enjoy a perspective different from that of the first five chapters.

I hope this book will serve as a reference for those interested in tackling humanities research topics from the perspective of data science.

The Voynich Manuscript: Data Science Takes on the World's Most Famous Undeciphered Document

Mari Agata (Co-author)

Seikaisha Shinsho

192 pages, 1,485 yen (tax included)

*Affiliations and titles are as of the time of publication.