For a long time, there have been attempts to convert the symbolic characters (graphemes and glyphs) of the Voynich manuscript into text with Latin characters, in order to make it also accessible for machine translation.
The EVA (European Voynich Alphabet) transcription system was developed until 1998, based on a previous version, INTERLN.EVT.
Known letters that are legible in the manuscript were simply transferred “as is”: a,o,i,y,s,e (but not definitively). These letters, as well as ‘l’, but not the “ii” combination, remain valid even in our “Arabic Solution”. For all the other combinations of symbols and letters, the developers adopted values which seemed to make sense to them.
Since then, the EVA transcription has been evaluated according to all the rules of linguistic science, to see if a language that really existed (at the time) could be identified from it – without success. Is it likely that a language written in the middle ages would have so many “eee” (in EVA 275 times) or “ii” (in EVA 1363 times)combinations? Did those who wrote it really have so much spare ink at the time?
Nevertheless, EVA turned out to be a very valuable piece of work, thanks to a team who operated unbelievably meticulously and laboriously, led in particular by Prof. Jorge Stolfi, which also included:
Denis V. Mardle
Takeshi Takahashi (last but not least !)
The true benefit of EVA is that other solutions can be extrapolated electronically. This is how, by running EVA through my transcription software eva2arab.exe, I was able to produce a rough draft of the entire manuscript in romanized Arabic, which was the basis for our translation. The transliteration software bases its conversion on this map .
Unfortunately, these days no-one is still interested in updating EVA, although that would seem essential, given the new interpretations of some of the characters in the manuscript.