Mr. Fournet:
I do not consider my approach to inter-proto-language comparison to be
"old(-fashioned)"; I think you mistake traditional methods for "old". Some
truths do not change much: Newton's apple still falls down from the tree, and
the best (so far) explanation for the phenomenon is the theory of gravity.
If one wishes to compare proto-languages, the method with the greatest
probability of success is to compare between languages/reconstructions that are
roughly contemporaneous though I do not rule out the possibility of
non-contemporaneous comparisons when one (or both) earlier forms are
unavailable.
As for Baxter, I think most would allow that in this area the
reconstructions of Sergei Starostin on Tower of Babel are worthy of, at least,
tentative credence.
In this case, the appropriate stage to compare with PIE is
Proto-Sino-Tibetan (not Sino-Caucasian as I mistakenly wrote).
His PST reconstruction for 'goose' is *nga:n(-s).
Though the initial *ng is the velar nasal, I am not taking any calamitous
leap to entertain the possibility that it was earlier *n + *g since anyone
familiar with Sino-Tibetan is bound to notice the pervasive prefixation in that
language. I believe these prefixes were a Sino-Caucasian (and Sino-Tibetan)
device for differentiating between related semantic applications of roots with a
more generalized meaning.
What a prefixed *n- might mean I do not know. But if Sino-Caucasian (or, at
least) Sino-Tibetan is anciently related to PIE, it might be a trace of the
individualizer -*n found in PIE but derived from an earlier proto-language
common to both.
This possibility is enhanced by the fact that *(n)ga:ns is so tantalizingly
similar to PIE *g^hans.
There is no mention of a *ngah in Starostin. I do not know if Baxter
proposes it but I choose to subscribe to Starostin's reconstruction. I suspect
that *ngah is from Fournet, a reconstruction, a verbum ex machina, to suit
Fournet flimsy argument. Tell me I am wrong.
However, the reconstruction *ngaens is of more interest in view of the long
vowel of *nga:ns. Also, as you know, I consider the *a of *g^hans to inevitably
derive from*a:, a certain trace of a 'laryngeal'.
Also, you mention that *ngaens is in the chain from *ngahans, presumably
where you got your silly *ngah by false division.
If we delete the prefix, and change the initial vowel from *a to *e,
perhaps accounting for the palatalization of *g to *y, we arrive at a from
*(n)gehans, which could easily lead to *(n)g(y)a:ns (you have something similar,
*ng_yans, whatever the underline is supposed to mean although you totally miss
the significance of the long vowel: a result of contraction and
compensation).
Also, I might remind you to take into consideration the palatal quality of
*g^hans; it makes a big difference.
*gehans would compare very interestingly with the form *gheHans which I
reconstruct to have preceded *g^ha(:)ns (from earlier *K?XE-HA-NA-SO).
A development of *K?XE to *g(y) (ST) and *g^h (PIE) is not too outré for
you, is it Fournet.
A *g(y) could account for the modern Beijing reading of yàn; and I withdraw
my previous objection to it arising from *nga:n(s) if we emend that form to
*(n)gya:n(s).
Now, so that you will not make such mistakes in the future, *(g)h- is not a
voiced velar spirant. Where did you ever get such a strange idea? Also, I am not
aware than anyone has reconstructed voiced aspirated stops for Sino-Tibetan (or
Chinese, for that matter). I am sure you will let me know if I am wrong.
I looks to me as if their "reinterpretation" and "adjustment" are better
viewed as _simple_ learner's mistakes.
Patrick Ryan
----- Original Message -----
Sent: Saturday, November 17, 2007 1:48
AM
Subject: Re: Re: [tied] Anser (was:
swallow vs. nighingale)
----- Original Message -----
Sent: Friday, November 16, 2007 9:24
PM
Subject: [Courrier indésirable] Re:
[tied] Anser (was: swallow vs. nighingale)
----- Original Message -----
Sent: Friday, November 16, 2007 11:21
AM
Subject: Re: [tied] Anser (was:
swallow vs. nighingale)
<snip>
> PIE *ghans is a compound word : gh_H2 + H2_ns.
> So
there is no "rural" or "irregular" situation.
> LAtin has the simple
word H2_ns. Most other languages have the compound.
What are
arguments for this hypothesis?
============ ======
A.F
The arguments are to be found outside PIE.
For example, Chinese has both :
e2 : goose < *ngah (= gh_H2)
yan4 : goose < *ngah +
ans- (=H2_ns)
Two synonyms : yan4 being a compound of ngah +
ans
I disagree with the traditional view holding the -a-
in *ghans < *ghH2H2ns
to be not of laryngeal origin.
I think this is wrong, (whatever Maitre Meillet
thought about this)
============ =======
Patrick Ryan wrote :
There is not a shred of credible evidence that yan4 is the result of
a compound of * ngah + *ans. For that matter, e2 from *ngah is
also so highly unlikely as to be incredible.
============ =========
A.F
OK. Let's get into Chinese reconstruction
:
yan4 : from Baxter (AD 500) *ngaens < (BC-1000)
*ngrans.
In the traditional view this -ae- vowel is
reinterpreted as being from *ra or *la,
I think this is much too simple, although sometimes
true.
I agree -ae- is from -ya-, which is sometimes from
*ra or *la.
In the case of *ngaens, it is from *nga(g)hans, the
voiced velar spirant -(g)h- became yod, so that *ngahans > ng_yans
> *ngaens.
This word was Two-syllabic with oxytonic stress as
is usual in Old Chinese.
e2 : from Baxter (AD 500) *nga < BC-1000
*ngaj
For the same reason *ngaj is from
*nga(g)h.
==
As you see, Dear M. Ryan, I am just reinterpreting
Baxter's reconstructions and making one and only minor and slight
adjustment : sometimes, yod is from a velar voiced spirant in Old
Chinese.
My reinterpretation is all the more probable as it
enables us to connect good old PIE with Old Chinese, as reconstructed by
sinologists.
This is sensical macro-comparison. Connecting
already existing reconstructions thru minor adjustments. It works. And it sheds a lot of new light on PIE.
I am afraid your old approach of PIE is bound to
blow up, sooner or later.
Get ready for rough weather.
Arnaud.
.
