How the Script Was Read

The decipherment is a rare case where the working can be followed. The tablets were unreadable and the language unknown, and the first real progress came from an American scholar who noticed sets of words sharing a stem but differing in their final signs, which showed that the language inflected and allowed a grid to be built relating signs that shared a consonant or a vowel without knowing what either was.

Michael Ventris, an English architect who had worked on the problem since boyhood, extended that grid and then made the guess that broke it open. On the assumption that certain frequently occurring words were place names on Crete, he assigned values that produced recognisable names, and the values so obtained generated Greek words elsewhere. Ventris had long expected the language to be something other than Greek and followed the evidence against his own expectation. Confirmation came from an independent direction shortly afterwards, when a newly excavated tablet listing vessels was found to match the readings against the drawings of the objects beside them. A philologist joined the work and the two published the decipherment together. Ventris died in a road accident a few years later.