Draft transliteration scheme on the Web

[Vidhyanath Rao]

> Among the many transliterations of this letter, scholars now seem to use
> three in particular: l_macr-b, r_dia-b, z_dot-b.   As well as the fricative
> aspect, the Malayalam and Tamil pronunciations sound something like 'r' and
> 'l', respectively.    Hence z_dot-b has three advantages:   (1) dot-b
> matching retroflexes, (2) z suggesting a fricative, (3) no problem about
> the r/l aspect.

For those who know Tamil/Malayalam, the actual symbol makes little
difference; humans can get used to almost anything. I no longer
blanch when I see z for "s.  But for others,
z-underdot may be too confusing: There have been people who thought
that the sound is a voiced retroflex >sibilant<, a sound that is rare,
and not found in modern Dravidian languages, but is reconstructed for
proto-IA. I have never heard it in any natural language, but it is not
hard to voice .s, and the result does not sound anything like the sound
in question.

I doubt that there is any easy answer. Any choice will confuse those
who have never heard it.