deciphering

Voynich is an old acquaintance of this house: for years, we have been tracking (and gutting) each of the attempts to decipher the “most mysterious manuscript in the world.” They have all been unsuccessful and that includes, of course, the attempts to some of the sharpest minds of history. Now, however, we have a new idea. And, despite not solving absolutely anything, it sounds very good. What is the Voynich manuscript? Let’s start at the beginning: Between 1404 and 1438someone somewhere started writing a book in a language or code that no one has been able to decipher. A book that, since its rediscovery in 1912, has baffled everyone and especially cryptographers. Overall, this is an extraordinarily strange piece (full of illustrations of rare or non-existent plants, astrological symbols, strange creatures and naked women) about which we know only a handful of things. We know, for example, that it is a natural language (or a code related to a natural language) because complies with Zipf’s Lawan empirical regularity that only occurs in natural languages ​​and that describes the frequency of appearance of words. Invented languages ​​(especially languages ​​invented in the 15th century) do not comply. We have known this since the 60s, but little else. And people are still trying to figure it out? Yes, absolutely yes. The Voynichians are a group of people who are extremely passionate (and ‘insistent’) about their manuscript and, in fact, have members in almost every social strata in the wide world. An example is today’s protagonist. A few weeks ago, the magazine Cryptology public a job of Michael A. Greshko in which a new and very interesting idea was proposed. Greshko is a renowned science journalist, he is an editor at Science and has worked for media such as the New York Times, the Washington Post, Nature, Scientific American and National Geographic. He is someone who is risking part of his prestige on this, come on. And what does he propose? Greshko has exposed something called “Naibbe cipher”. Basically, it is an encryption system that allows languages ​​such as Italian or Latin to be transformed into a pseudo-writing that preserves properties of ‘voynichés’ (the ‘language’ of the manuscript). Respect, for example, things like glyph frequencies or word lengths. All this, with plausible cryptographic tools for the 15th century. And that’s precisely what’s interesting: Greshko doesn’t try to “read” the book; It attempts to demonstrate that, at that time and starting from a common language, a text similar to that of the manuscript could be constructed. How to make your own Voynich at home. According to the work of Cryptologiathe Naibbe method does things like break words into blocks (splits ‘gatto’ into ‘g’, ‘at’ and ‘to’), uses random systems (like dice or card rolls), and generates a homophonic cipher (ciphers specially designed to “counter the main deciphering tool for monoalphabetic substitutions, frequency analysis”). So, have we solved the problem? Not even close. As I said, Greshko has not deciphered the manuscript. He has simply looked for ways in which that manuscript could have been produced. For years, artificial intelligence algorithms have failed in the translation of the Voynich and, as the author explains, this may be because they do not know very well what to look for. Systems like Naibbe draw constructive possibilities that expand the options among which we can search. And in that sense, yes: Voynich is still much smarter than us. Although we don’t know for how long. Image | Gunnar Klack In Xataka | No, no “artificial intelligence” has deciphered the Voynich manuscript