Dear all,
I would recommend that all of us, in our discussions about transliteration and standards, bear a few things in mind.
1. It is 2023. If you are typing on a computer, any transliteration system you use is going to be interconvertible with any other. (With some exceptions: see below.) I assume most Indologists still don't know this, but going from ISO-15919 to IAST and back is a totally trivial process, and there are many tools available to do so (including
Aksharamukha, and the
Sanscript libraries for python, javascript, etc.). If you are happy with typing in IAST, or HK, or SLP1, nobody will take that away from you.
2. There is a totally separate question of what transliteration system is to be employed in publications. Web-based publications in principle allow for Indic-language text to be displayed in any arbitrary system. On my
Sanskrit course people can choose between viewing the text in ISO-15919 and Devanagari. But most publications are still based on a "paper-like" model, where the final result is fixed, and therefore a choice needs to be made. The primary reasoning behind this choice is
inertia, i.e., whatever system of transliteration the author or publisher has used in the past. This is not necessarily a bad reason!
3. There is one very simple reason to prefer ISO-15919. If you are working with languages other than Sanskrit, you simply cannot use IAST without ambiguity. Is ṛ a flap or a vowel? Is ḷ a retroflex lateral or a vowel? For distinctions that are not made in Sanskrit, one has to add new signs anyway, such as ṟ, ḻ, ṉ, ə, and so on. And the vowels present particular difficulties. Either you retain "e" and "o" as long vowels, and mark the short versions with "ĕ" and "ŏ," or you adopt a consist system of representing vowel length (no macron means short, macron means long, or something like that) which contravenes IAST. Since IAST is not a standard, any additions or modifications are per se arbitrary, and for that reason I have seen a number of very different solutions for, e.g., solving the vowel length issue while retaining some version of the IAST system. For these reasons, I have been persuaded a long time ago that ISO-15919 is to be preferred for representing Indic languages in transliteration, especially if one's research includes languages other than Sanskrit.
4. To ask ISO-15919 to include ṛ and ṃ as "permitted variants" would be a violation of the principle of non-ambiguity, since those graphemes are already reserved for transcribing "ड़" and "ੰ" (0A70, Gurmukhi tippi). If you want to use those signs in their traditional IAST values, use IAST.
5. To reiterate what Jan has already said, it is one thing to define a standard, and another to implement it. Font support for ISO-15919 is not as robust for IAST, in part because IAST makes use of a smaller range of diacritics and combinations. But I have added the relevant diacritic combinations to the fonts that I use, and a wider uptake of ISO-15919 would probably lead to wider font support.
Andrew