On 11/14/2011 2:04 AM, George Corley wrote: > I've set down some of my thoughts regarding romanization into a blog post. > Hopefully this can start a good discussion going about it. > http://gacorley.squarespace.com/imported-20110129051548/2011/11/14/design-parameters-for-romanization.html > > I've had a number of different opinions on romanization over the years. One of the big reasons for the differences is the technical limitations of software. In the early days, when I did everything by hand on paper, I used more diacritics, ligatures, and even some non-Roman letters. Then when I started using computers in the 8-bit days, I had to deal with the limitations of ASCII or ISO 8859-1 at best. I thought it would be a good idea to have a single standard for Romanization, so back in the early 90's I came up with a system that used lots of digraphs and even a handful of trigraphs (e.g., dhl for [ɮ], hrr for [r̥]). These days I've gone back to using diacritics, even some unusual ones like ḽ [ɮ] in my current romanization for Jarda. I tend to prefer one symbol per phoneme (including letters with diacritics or ligatures) for many of my languages, although I've gone the opposite route, using digraphs and trigraphs with fewer or no diacritics, from time to time (as with early versions of Lindiga). I also have two different romanizations for Yasaro, but for a different reason. Originally (as used for instance on the Minza color page, http://www.prismnet.com/~hmiller/lang/Minza/Minza-colors.html), I had a more or less phonemic spelling, with "sáva" for "white" /sǎva/ and "ną̀ti" for "black" /nôñti/. The newer Yasaro romanization is a direct transcription of the Yasaro spelling: "savá", "nánte". The spelling represents an earlier version of the language, so a word like "krávga" (to eat) for instance is actually pronounced /čâ:ŋa/. The name "Yasaro" itself is actually spelled "jassavráň".