implosive pronunciation of t

John Peterson jpeterso at UNI-OSNABRUECK.DE
Fri Jan 25 09:05:04 UTC 2002

Michael Witzel wrote:

> However, I simply use the traditional Avestan/South Asian terms. There is
> some infighting going on whether Munda (or Khmer for that matter) has
> ingressives. I don't think so. They are of the quality of English "yepp"
> for "yes", thus plosives without the release (explosion), i.e.

I don't believe that Munda has implosives, either:
I think we have to differentiate here somewhat, at least in the case of
Munda. In northern Munda languages such as Santali, we have a
post-glottalized stop of the type "yepp" in English. These are non-released,
voiceless stops produced (almost) simultaneously with a glottal  stop. This
is also found in the Kiranti languages. The glottal stop is only audible
when the next word (or syllable) is pronounced, if it is heard at all.

The situation in Kharia (and perhaps other south Munda languages) is
entirely different, although perhaps historically of the same origin. Here
we have PRE-glottalized, VOICED stops, as in the copula "ai?j" (the "j" is
actually a voiced palatal stop (upside down "f", for  those familiar with
IPA), not an affricate). The stop itself is also not released here, and the
effect is quite different from that of post-glottalized stops. When the
pre-glottalized stop is released in Kharia, the air flows inward, and what
one hears is a homorganic nasal, in other words "ai?(j)ny". This is termed
(unfortunately) "implosive" by Pinnow and many others, but is in fact quite
different from canonical implosives, which are not pre-glottalized. This is
probably also the reason that e.g.  Banerjee (1894), in his grammar of
Kharia, often gives forms such as "ainy" for the copula, as the homorganic
nasal is more easily heard than the non-released stop.


John Peterson
FB 7, Sprachwissenschaft
D-49069 Osnabrueck
Tel: (+49) 541-969-4252
Fax: (+49) 541-969-4256
Email: jpeterso at

More information about the INDOLOGY mailing list