Font-related problem for linguists

[Peter Kahrel]

Martin and I seem to be agreed that the best (if not the only) way to 
solve this is to add a character to a Unicode range. Here's why. Suppose 
you have some roman text in, say, English, in which a word or a phrase 
is used from language B, using the open a. In the roman text, all the 
a's look the same. But when you italicise the text, you want the English 
a's to look like italic script a's, and the a's in the bits from 
language B like slanted a's.

This cannot be solved at the font level. At a practical level, 
alternate glyphs are not an option because they're not available in word 
processors. Besides, even in typesetters like InDesign, treating the 
different a's as alternate glyphs is not an option because you don't 
want to go about marking all those alternate glyphs. Besides, treating 
the a's as alternate glyphs is wrong because they're not alternates in 
the first place, just as e and schwa aren't alternates.

Finally, the only way to ensure the smooth transition from word 
processor to typesetting software is via dedicated Unicode points. Any 
other solution will require typesetters manually to change some 
characters, and I guess you all know what that can lead to.

Peter