Inscriptions and Dravidian sound changes "y" > "c" and "y" >

[Palaniappa]

In a message dated 98-05-25 20:43:53 EDT, narayana at HD1.VSNL.NET.IN writes:

<< This is not correct. It is the illiterate person who is afraid that
 he is pronouncing a word wrongly that tries to hypercorrect.

 For example a street vendor selling limes
 calling 'darbhakAyalu' instead of 'dabbakayalu'.

 Sometimes even the learned fall into the hypercorrection trap but
 more often it is the unlearned and illiterate.
 >>

I cannot comment on the Telugu forms. But as far as Tamil is concerned, the
type of hypercorrections made by an illiterate Tamil person is very different
from the type of hyperstandard forms suggested by Dr. Krishnamurti. In
Tamilnadu, upper caste dialects often use Sanskrit words with voiced sounds in
the word-initial positions, and sounds such as S, z, and J. So, a person with
hyper-corrective behavior would sometimes use these sounds in places where the
original does not have them. For instance, pUri, the well-known food item, is
called by a person I know as "bUri". Similarly the name "padmA" is changed by
some into "batma". A friend's wife's name is "Shuseela". I have come across
all these in college-educated persons. All these I can accept as possible even
from an illiterate person because these are based on features readily
observable in literate/upper class persons by a illiterate/lower class
persons.

But the knowledge of the loss of initial c/s etc. based on words like Ayiram,
ENi, etc., is simply impossible to expect from ordinary people. I have been
interested in the study of language for a long time. I did not know about this
phenomenon until I read Burrow's work. The posting by Selvakumar  reflects
that too. The average illiterate person is not a Burrow or Emeneau. Just out
of curiosity, how and when many Tamil members on this list knew the
derivations of Ayiram and ENi, and what those member's Sanskrit/linguistic
knowledge background is

In any case, my point is simple. Until an objective linguistic reason is
presented (without any circular reasoning), which explains why *-y->-k-/-g-
and *-y->-j- are acceptable as natural sound changes while *-y->-c-/-s- is
not, we have to accept *-y>-c-/-s- also as natural.

Regards
S. Palaniappan