Karnataka/KannaDa (was: Karave caste and Kurus)

[Robert Zydenbos]

Am 11 Jan 2001, um 1:03 schrieb N. Ganesan:

>  >I cannot think of any example right now; on the other hand, I do not
>  >know whether we should consider the difference e / a important
>  >here.
>
> At least in Tamil, it is very important.

Do you know an example? Because what you mention is the
following concerns something else -

> -run- where n is dental  assimilates to
> -n2n2- where n2 is alveolar. Eg., karunATu > kan2n2ATu
> if we apply Tamil sandhi.

One question to be considered is whether Tamil sandhi should be
applied to Kannada. :-) Kannada (where a distinction between n
and n2 is nowhere recorded, unlike in Tamil) goes further in
consonant assimilation in compounds, as we can see, e.g., in
words in which the first element is he+[consonant], meaning 'large'.
(The Kannada adjective hiriya is the cognate of Tamil periya; hiridu
< OKa piridu 'what is large / old / great / senior'.) The r is obviously
lexically important, as we can see in older literature: per-maram
'large tree', but later it becomes hem-mara; cf. also heb-bera.lu
'large finger = thumb', hed-daari 'big road = highway', hej-jiga.ne
'large leech'.

If the tapped r in the above examples can be assimilated to the
following consonant to become just about anything (even a labial
nasal, as in hemmara), then an assimilation of m > n is only to be
expected. Cf. also hindu 'what is behind' (OKa pindu, cognate of
Tamil pin2), which produces compounds like him-me.t.tu- 'to turn
back, retreat', him-ma.di 'heel' etc. Here we see that an originally
alveolar and a labial nasal can be assimilated; and from all the
above examples we see that the preceding consonant is
assimilated to the following.

So I still believe that R. Narasimhacharya's suggestion Kannada <
kam + naa.du deserves serious consideration. It may be a pity that
apparently there are no other short syllables with the structure
(C)Vm- that form such compounds; but I do not see any principle
why the suggestion should be rejected.



Robert Zydenbos
Institut für Indologie und Iranistik
Universität München