Nik Taylor wrote: > I'd say it depends on the language. For instance, in Japanese, [P] > (bilabial fricative) is an allophone of /h/ before /m\/ (unrounded back > high vowel). Nevertheless, in the most common transliteration, one > writes <fu>, and not <hu> (which is fortunate, since recent borrowing > have begun to use [P] for [f] in other places), thus <fuji>, and not > <huji>. However, sometimes it's best to use a phonemic transcription. > For instance, in my conlang, /t/ is pronounced [tS] before /i/ and /j/, > however, it is always written <t>, this is especially good with the > gender-prefixes. Gender 1 (female-rational) uses the prefix t(i)-. > Before a non-glide consonant, it is ti- ([tSi]), before a glide or vowel > it is t-, thus [ta], [tu], [tSi], [tl], [tSj], and [p] (slight > complication: *tw, *dw, and *nw are impossible, they become p, b, and > m). To write it phonetically would require additional rules: > 1. Chi- before non-glide consonants > 2. T- before a, u, and l > 3. P- replacing w. > 4. Ch- before i and y > > Whereas now one only needs to write > 1. T(i)- > 2. P- replacing w. This is the exact reason given in a book on Japanese I have. It's so much easier to explain different endings in Japanese when the consonant is always written the same, even if pronounced differently. However, I very much dislike the look of this system. Who wants to see 'Fuji' written Huzi, and 'chotto' as tyotto? (Or is it -ou?) Although, to be fair, I have heard Japanese speakers pronounce Fuji as Huji anyway.