paucity of Na-Dene roots with initial *ts- or *s-

[Geoffrey Caveney]

I have noticed, in looking for roots shared by Proto-Athabaskan, Eyak and
Tlingit, a striking paucity of such roots with initial *ts-, *ts'-, or *s-.
I would be interested to know if others on this list can point to roots I
am missing, or if other Athabaskan and Tlingit linguists have noticed and
commented on this fact as well.

By contrast, Jeff Leer reconstructed a sizable number of roots exhibiting
the correspondences Tlingit /k/ or /sh/ : PA */ts/, Tlingit /k'/ : PA
*/ts'/, Tlingit /g/ : PA */z/, and Tlingit /x/ : PA */s/ in his article
several years ago on recent advances in Athabaskan-Eyak-Tlingit comparison.
(The Eyak reflexes are similar to the Proto-Athabaskan ones.) These are the
phonemes that Leer reconstructs as "palatal velars" *ky, *k'y, *gy, *xy in
Proto-Na-Dene. Among the roots with these initial phonemes are many basic
nouns that belong to the core lexicon: *kye/i(:)ng' 'head', *kya(:)y 'rock,
stone', *kye/i 'blood', *kyi()tL' 'ashes', *kyo:ng 'see', *kyo()x 'be
dry', *kyi()ng 'sing/song', *kyi(:)y 'limb, knot', *k'yey
'straight', *-k'yin()d 'to fall over', *k'yi()tL' 'nape of neck,
occiput', *gyu: 'good, pleasant'.

But where are the Proto-Na-Dene roots for basic nouns in the core lexicon
with initial *ts- or *ts'- or *s-? There may be *sEnt' 'liver', but without
a Tlingit cognate Leer notes that it may as well have been *xyEnt'. There
is PA *ts'Eq 'finger' and Tlingit *tL'iG *'finger' with Tlingit
L-assimilation. And that's about it! Leer notes the forms PA *tsEny'
'flesh, meat' and Tlingit *dli'y* 'flesh, meat' but in the end he rejects
the comparison due to the lack of correspondence in aspiration.

I would appreciate any comments that members of this list may have about
this observation.

Geoffrey Caveney
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/athapbasckan-l/attachments/20140620/2cc83dc7/attachment.htm>