Three transcriptions: v101, NEVA and NEVA spaced

To the main page

The v101 transcription, by Voynich scholar Glen Claston, is much the most accurate, noting many fine distinctions which previous transcriptions have missed. However, the transliteration it uses is not easy to read, unlike the older EVA transliteration devised by Gabriel Landini, which matches Voynich characters to similar looking Roman letters and is also designed to be pronounceable. I have proposed a compromise transliteration which I call NEVA, using diacritics and non-Roman letters to convert v101 into a transcription which looks like EVA while capturing the distinctions noted by GC. I have also produced a transcription which displays the NEVA transcription in a grid with a null character "." inserted where I believe the hypothesised grid may have contained no characters. Words which cannot be analysed as I propose are displayed as in NEVA.

The v101 transcription is designed to be displayed in the Voynich TrueType font, which reproduces the appearance of the original text. NEVA requires many rare characters which have only been available for use on web pages since the introduction of Unicode. This means that the NEVA and spaced transcriptions should be viewed in a font such as Charis SIL created by the Summer Institute of Linguistics to display rare characters of the sort used in the International Phonetic Alphabet.

I created these pages using the v101 transcription and two Python scripts, voynichDatabase.py and makeVoynichPages.py. A mapping from v101 to NEVA is given in the Word file v101_to_NEVA and the assignment of glyphs to positions in the hypothetical grid is given in the text file nevapositions.