Dave writes:

Also, in my esteemed opinion as an ‘A’ level Linguist, haha, spelling problems are solved by having phonetic spelling with existing glyphs ;)

(I don’t ordinarily like quoting private emails in public, but this one’s not too personal. I wouldn’t ever mention the devil orgy stuff.)

In my esteemed opinion as an obsessive dilettante, phonetic spelling can’t work when there are more phonemes than glyphs. You have to invoke digraphs, which are an ugly compromise. What we need is a new alphabet! Actually, we need a syllabary (a collection of glyphs, each of which is mapped exactly onto a single sound).

As long as we’re abandoning the practicality constraint, what would an optimal syllabary look like? If the number of phonemes in English is about 40-50, we need 40-50 distinct symbols. Most attempts at language design stick with the latin alphabet, maybe add a few accents or apostrophes, and end up with something nasty.

I propose that a syllabary needs to be designed by usability experts, not linguists. The similarity of the letters i and j, for example, is bad for usability, as is the fact that ‘lowercase ell’, ‘capital i’ and ‘one’ all look identical. I’m sure that a good designer, maybe teaming up with a semiotician, could design a set of 50 maximally distinct and usable characters. Once we have that, real language design can start.