> "
> What my work suggests to me is that "dogs" were an important
part of our _earliest_ ancestors' lives. And built into the earliest
language is a clear distinction between "wolves" (FHA; Nostratic
*wa:-), and "predators" in general; and "dogs" (KHE; Nostratic *k^A-
).
>
> The word Torsten cites for "dog" is _obviously_ not very
early! "*kWVn/r-" (where in Heaven's name does the -*/r come
from???; and why *W rather than *w???).
How do you distinguish 'obviously' and '_obviously_'?
> The root on which it is based is obviously **k^eH- (Nostratic
>**k^A?-), 'to be a dog" + *-w, 'to wag the tail like a dog' (PIE
>*k^eHw-) + *n(A), 'a (tail-)wagger' (*k^won-). This (PIE *k^eHw-)
>is most probably also the basis for "howl" (PIE **k^eHwl-) rather
>than *ul (owls do not howl!!!) whereas dogs, when they are not
>wagging their tails or sleeping, frequently do.
Here's Pulleyblanks proposal for the connections of the "dog" word
with Old Chinese and PIE:
(the digraph u¨ is used to render a consonant corresponding to the
vowel ü, as in French huit /u¨í/; correponding to rounded kW and
palatal k´ (kY) there is a palatal-rounded kU¨)
Chinese characters are everywehere replaced with .. .
"
5. DOG
Chinese .. quan EMC khwen` Tib. khyi , Burm. khwè; for other TB
forms see Benedict (1972:44), who reconstructs TB *kwij or *kw&j.
Benedict explains the final *-n of the Chinese form as
a 'collective' suffix, also found in min .. EMC mjin 'people.'
IE *k´won-, *k´un- 'dog,' Skt. svá:, svan-, Gk. kuon, kunós, Goth.
hunds, Eng. hound,
etc.
Chinese has another word for 'dog,' gou .. EMC k&w` < *káu¨? which
is not usually connected with quan .. . As now reconstructed,
however, we can see that both are based on the same root with the
consonantal skeleton *k-u¨). We can reconstruct quan .. as *khu¨&´n?
or *khu¨&´?j?. The aspiration may be accounted for by the fact that
*khu¨-originated as a cluster rather than the single segment *kU¨-.
Quan .. does not occur as a rhyme in early texts so there is
uncertainty about its Old Chinese final. It is sometimes placed in
the *-an group, apparently on the grounds of a supposed phonetic
relationship to rán .., the upper part of which the Shuowen treats
as a phonetic speller (Dong Tonghe 1948, Wang Li 1958: 65). This
does not occur separately in texts. It is said by the Shuowen to be
composed of 'dog' + 'meat,' read like rán .. and meaning 'dog meat'
which scarcely justifies the assumption that 'dog' is a phonetic
speller. Karlgren may be correct in placing quan it in the *-&n
rhyme group. It is phonetic in quan .. EMC kwen', "watering
channels,' also written .., with chuan .. EMC tchwian playing both a
phonetic and a semantic role. Though chuan .. lowered to -an for
unexplained reasons in Middle Chinese, it rhymed in *-&n in the
Shijing and other words in its phonetic series belong in this group.
This analysis suggests that *-n may indeed be a suffixed element,
though the alleged 'collective' meaning is not apparent.
The phonetic correspondence between the Sino-Tibetan and Indo-
European forms is extremely close. If it is not a matter of pure
coincidence, it can only be explained by (a) borrowing in one
direction or the other, (b) inheritance from a time of shared
linguistic unity. Since the word is widespread in both language
families, borrowing would have to have occurred at a very early
period. Since the domestication of the dog goes back to Upper
Palaeolithic times, however, borrowing seems rather improbable.
"
The later determination at 14,000 years makes it probable again.
This looks almost like Patrick's analysis ;-)
The 'collective ending" -n is known from a few dozen examples in
Sino-Tibetan (Matisoff). A suspiciously large number of them have
previously been suggested as PIE cognates, which makes me suspect
they are originally from some language west of (but close to) China.
Torsten