Have we ruled out simple substitution unwisely?

[RadioFM]

Many people turn their attentions to cracking Labelese first, thinking they have a better shot at guessing a crib/cartouche. This is a good playground to try out different parsings of symbols. Words are short, a few anomalies here and there, and you can make educated guesses of what a certain label says, should the text be meaningful and related to nearby drawings.

IMHO the fact that nobody was able to make a dent on Labelese, even with such short words where the possibilities of parsing glyphs are limited, is one of the big culprits of cementing the idea that simple substitution is unlikely. Positional rigidity and other observations have only cemented it further.

Even if we were to correctly parse the characters, even if we had the right transliteration, there's some explaining to do about word statistics. There's some explaining to do about duplicate labels, quasi-reduplication...

Personally I'm not convinced our transliterations sit at the root of the problem, but honestly I've yet to run the numbers properly. So, thanks for bringing this up.