On Mon, 13 Nov 2000 07:35:50 GMT, "Glen Gordon"
<
glengordon01@...> wrote:
>Miguel:
>>Labialized /pw/ is quite easy to make, in fact. Try it. Maybe you
>>should first get your facts straight, and then accuse other people of
>>being amateurish.
>
>You conveniently skipped the next question... How rare is this and is it
>honestly necessary?
You conveniently lie. I quote:
>There was no *pw. There can't be any *pw sound. The sound *p is inheirantly
>labial to begin with! How does one honestly distinguish *p and *pW? Any
>examples of this in a HUMAN language? And, in the end, how necessary is it
>to posit this?
There's a difference between "how rare is this" and "there can't be
any *pw sound. [..snip ignorance..] Any examples of this in a HUMAN
language?"
The fact is that we find in Germanic a substantial number of cases
where PIE *kw is reflected as *f/*b. For instance: "four", "liver",
"wolf", "-lve". BTW, does anyone know where I might find an
exhaustive list of these? I have never seen a good explanation of
this, except vague phrases along the line of "*kw > *p is a natural
development" (it is, but that does not explain why _some_ words do
this in Germanic and others don't).
On the other hand, if we once had *pw, which turned *kw elsewhere
(given the markedness of *pw, a natural thing to occur), but was
retained as *p in Germanic, that would make a good explanation. In
some cases, this can be made plausible by related forms which also
show *p in other IE lgs. (e.g. Lat. lupus "wolf" ~ vulpes "fox" for
*wlpw-), in other cases the only evidence I have at the moment are
comparisons with extra-IE forms, such as *pwetwor- ~ PAA *p.wat.-
"four" or *l^a:pwnt- "liver" ~ Uralic *lapde- "spleen".
When I first started thinking about the possibility that, besides *k,
*k^, *kw (etc.) [and possibly *x^, *x, *xw], the other pPIE consonants
may also have had palatalized and labialized variants, I was not
thinking of these Germanic cases of *p ~ *kw alternation, nor of the
famous problem of Armenian plural -k`. What I had in mind was the PIE
verbal system, where we find *m ~ *w alternations in the 1st person,
and *t ~ *s alternations in the second. I was also thinking of the
irregular o-grade presents as in Latin molere "to grind", where a root
*mwel- would explain Latin -o- as well as Greek zero-grade mule:,
besides delabialized e.g. Slavic mel-jo~. Also the *t ~ *s
alternation in the perfect active participle and in some other words
(*h2aus- ~ *h2aut- "dawn", etc.), and maybe the curious Greek *pt-
words besides *p- elsewhere (Greek *pj -> pt in the j-stems). But
there are more phenomena that suddenly begin to look understandable if
we assume palatalized and labialized variants of the consonants: the
*n/*i stems like *poti-/*potn- (< *potn^-), or the verb *nem- ~ *yem-,
some *l/*i alternations such as in the "liver" word (< *l^a:pwnt-),
and maybe even the Caland system (*n -> -r ~ -n-; *nw > -u ~ -m-; *n^
> -i ~ -n-).
=======================
Miguel Carrasquer Vidal
mcv@...