Inter-Script Identity

Pronunciation not determinative ___ YA and YYA __ Visarga and TAM aaytam

1. Pronunciation not determinative ______ (Contents)

When thinking about similar Indian scripts, most letters are recognisably 'the same' in each script. If we were dealing with transcription (of sounds), 'same' would mean phonetically. In the context of transliteration, however, 'same' has to mean linguistically the same, in a general sense.
A simple example is the letter, short a. In Sanskrit, Hindi, and many other languages this is pronounced like the vowel in 'hut', but in Assamese, Bengali, and Oriya it is pronounced like a short /o/, as in 'ball'. In spite of this, there is no doubt that this letter is the same one in each language. The reason is simply that many Sanskrit words occur in each language, with the same spelling: a.m;sa, 'part', agni, 'fire', etc.
This example shows that pronunciation can be a bad guide to inter-script identity.

2. YA and YYA ______ (Contents)

The Assamese, Bengali, and Oriya alphabets have a letter YA (using the notation of Unicode) at the beginning of the semivowels and another letter YYA, called 'antastha a' in Bengali. YYA is written as a modification of YA. ISCII [1991, Annex-A], identifies YYA with Devanagari ya, and introduces Devanagari ya with an underdot to represent YA. I believe this is the wrong way round, if we are thinking of transliteration rather than transcription.
In those three languages (the Eastern Indian languages, say), it is YYA that is pronounced like Devanagari ya, while YA (when alone) is pronounced like Devanagari ja. Annex-F of ISCII gives the rule that the unpronounced inherent a at the end of words is not to be 'transliterated' in North Indian scripts. These facts show that in ISCII, inter-script identity is based on phonetics. The general linguistic facts, however, lead to a different conclusion:
No word in the Eastern Indian languages begins with the letter YYA. On the other hand, many Sanskrit words beginning with the letter YA are found in these languages, such as yadi, 'if', and yuddha, 'war'.
The pronunciation of YA and JA has become identical in the Eastern Indian languages by a sound change (palatalization). Hence the better transliteration is one such as

                   Y = y          YY = .y

3. Visarga (.h) and Tamil aaytam (_k) ______ (Contents)

ISCII [1991, Annex-A] takes these two to be identical. Bishop Caldwell [1875, p.13] took them to be different. Their use is as follows. The Tamil Lexicon, s.v., tells us that Tamil words can have

                  short initial vowel + _k + hard consonant.

Agesthialingam [1977, p.1] says he will treat aaytam as an alternant of both v and l.

In Sanskrit there were originally six voiceless fricatives [Macdonnell, 1916, pp.13.18]:

       velar     jihvâmûlîya   
       palatal   ;s                
       retroflex .s                
       dental    s                 
       labial    upadhmânîya
                 visarjanîya (now called visarga)    

Visarga now replaces both jihvâmûlîya and upadhmânîya, and may occur at the end of a word, either alone or in a compound. So in spite of aaytam being a voiceless velar fricative (or sometimes labial), its linguistics is quite different from that of visarga. One may note in passing that Skr du.hkam > Tam tukkam, not *tu_kkam.

So I believe that aaytam and visarga are not equivalent for transcription.