Re: Niggahiita in IPA?

>> Buddhaghosa
>> appears to describe contemporary pronunciation, which according to
him
>> involves not articulating any other sound by suppressing or checking
the
>> organs of articulation (the karanaani) and producing the nasal
through the
>> nose without the mouth being open.

> Wouldn't this be the equivalent of the gutteral nasal?

No, it wouldn't.  The velar nasal ('gutteral' has a 19th century aura
in modern phonetic discussion) is a well defined articulation in which
the oral closure occurs at the velum (soft palate).  It is the final
consonant in the standard English pronunciation of -sing-, -sang- or
-sung-.  Buddhaghosa's dictum covers not only the velar nasal, but any
other nasal stop (-m- or any of the various -n-s) and does not
constitute a coherent phonetic description (how does one close the
mouth without using any of the organs of articulation?).  As already
noted, Buddhaghosa may not be aiming at a phonetic desription at all.

> So that one could use:"[n with leftward hook at right]" for the
niggahita in
> isolation?

You could do this (as IPA) only if you could show that niggahiita was
invariably pronounced as a velar nasal.  Buddhaghosa's testimony is
negative evidence here.  Also, there is the existence of another Pali
sound, the ka-vagga nasal, which is much more plausibly so interpreted.
  The grammarians may be frustratingly vague about phonetic detail, but
they unanimously regard these as two distinct sounds.  As Eisel Mazard
mentions, I objected to his usage of engma for Pali niggahiita (and
also to his use of the IPA palatal nasal symbol "[n with leftward hook
at left]" for the the Pali velar nasal) on precisely these grounds.  I
told him that the PTS standard dotted m and n was better.  He replied
with the same point he brings up in this discussion, that the use of
dots for other than retroflex consonants is phonetically unmotivated
and inconsistent.  Of course he is quite right about that.  But I think
he misses the point, because his alternative is equally unmotivated and
inconsistent.  The PTS standard is not preferable to his because it is
more phonetically motivated, but rather simply because it is the
standard.

> Would the sound represented by the IPA's superscript tilde cover the
anusvaara > as described in the article by Sharma?

Sharma's comments exclude nasalized vowels both on the grounds that
these involve both oral and nasal components (and cannot therefore be
'pure nasals') and because they lack oral 'stricture'.  In the portion
of his paper quoted, he tries to phonetically interpret some Sanskrit
phoneticians' descriptions of anusvaara (we are not given specific
references).  This is a mixture of articulatory phonetic observations,
logical contradiction and wishful thinking.  According to these
phoneticians, anusvaara is characterized by an absence of localized
oral stricture (i. e. not labial, dental, palatal. velar, etc.) but
nevertheless by non-localized oral stricture ('the mouth is simply
closed'). This is of course the same bind seen in Buddhaghosa above;
how can there be closure of the mouth without the mouth being closed in
some particular place?  The diagrams certainly do not show any such
thing.  (And one is tempted to ask what those diagrams are based on:
X-ray tracings or the author's intuition of his own articulation.)
Skeptics are instructed to study with 'native speakers' (who seem to be
pa.n.ditas trained in Taittiriiya Vedic recitation).

In thinking about IPA it is important to see that it is a set of
decisions on how to represent consistently the sounds of human
language, and not a theory which makes predictions which could prove
wrong.  Any idea that IPA works adequately for European languages but
might break down for Asian languages is thus founded on
misunderstanding.  If a sound is discovered for which no IPA symbol
exists, then we can (and must, according the spirit of IPA) immediately
create a new symbol for it.  IPA can be criticized as cumbersome, hard
to learn, etc., but it cannot be criticized as incapable of
representing the sounds of any human language.  IPA is a tool which
makes possible the comparison of the phonological aspect of languages
by removing orthographic differences; as such we (at least we
linguists) should cherish it.

It occurred to me in looking back at my earlier posting, that I had
overlooked an interesting point about the velar nasal (the ka-vagga
nasal, appropriately represented by the engma, or as -"n- by Velthuis).
  This sound, though listed in the inventories of both Sanskrit and Pali
grammarians, and accorded a distinct letter in the alphabets, is not
really a contrastively distinct consonant.  It is effectively limited
to clusters with following velar stops (k, kh, g, gh).  But grammarians
(both Sanskrit and Pali) use it distinctively in their metalinguistic
vocabulary.  As a simple example -ti"n- is Panini's acronym for the
active and middle agreement suffixes, but it is ill-formed as a
Sanskrit word.  Similarly, in Moggallaana, -"na.m- and -"naaka.m-
represent suffixes which show up in the declensions of second person
pronouns (II.236-7).  Again, these are ill-formed Pali words.

Finally, I have missed in our discussion to date any reference to I. Y.
Junghare 'Topics in Pâli Historical Phonology', Motilal Banarsidass,
1979.  I suspect some of us may not be familiar with this volume.  It
is a University of Texas dissertation written under the direction of E.
C. Polomé (S. M. Katre was also there at that time).  It uses the
framework of generative phonology popular then (under the influence of
Chomsky and Halle's 1968 'Sound Pattern of English').  Thus it will be
rough going especially for non-linguists.  But you venture into Pali
phonetics and/or phonology at some risk if you don't know what Junghare
has to say.  Her interpretation of the pronunciation of niggahiita
(rule 28, page 63) is very similiar to mine; one difference being that
she interprets it and the preceding vowel as coalescing into a short
nasalized vowel rather than a long half-nasalized vowel as I do.
Though I was not familiar with her work when I made my interpretation,
I really should have mentioned her as one of its sources.  In any case,
my interpretation is not as unprecedented as Eisel Mazard says.

George Bedell



*   *   *   *   *
George Bedell
120/2 Palm Springs Place
Mahidol Road, Chiang Mai 50000
THAILAND
+66 (0)53-241342
correspondence in Japanese may be addressed to
      gbedelljp@...

Send instant messages to your online friends http://uk.messenger.yahoo.com