Re: [tied] Dissimilation of gW/kWVw to gVw/kVw

From: Patrick Ryan
Message: 37264
Date: 2005-04-20

 
----- Original Message -----
Sent: Tuesday, April 19, 2005 6:22 PM
Subject: Re: [tied] Dissimilation of gW/kWVw to gVw/kVw


On Tue, 19 Apr 2005 12:34:05 -0500, Patrick Ryan
<proto-language@...> wrote:

 <snip>
 
<PCR>
Very interesting.
 
 

One important case of lengthening occurs in posttonic
(svarita) position (unless the preceding morpheme is heavy),
which explains the vowel alternations in a paradigm like
(svarita lengthening notated as <â> instead of <a:>):

nom. **xák-man-z  > **xák-mân-z  > *h2ák^mo:n
acc. **xák-man-m  > **xák-mân-m  > *h2ák^monm.
gen. **xak-mán-as > **xak-mán-âs > *h2(a)k^ménos
PCR>Do you think the following developments are plausible?
 
from *?á-(n)ke, 'top-prick' = 'sharp'
*Ha(:)(n)k^-mó to *Ha(:)k^-món
 
nom. *Ha(:)k^-món-z  > *Ha(: )^-mó:n > *á(:)k^-mo:n
acc. *Ha(:)k^-món-m.  > *Ha(:)k^-mó-nm.  > *á(:)k^monm.
gen. *Ha(:)k^-me-nós > **Ha(:)k^-mé-no-s > *a(:)k^ménos
 

"Lexical" long vowels are very much rarer than short vowels,
but their presence explains Ablaut patterns like:
 
o~e
*pá:d-z  > *pó:ds  > *pódz  > *pó:d
*pá:d-m  > *pó:dm  > *pódm.
*pa:d-ás > *pad&'s > *péd&s > *pedés
from *pa-?á-d(o), 'flat/swollen-(stative)-lump' = 'foot' (cf. Egyptian ptpt, 'tread')
*peHód-
 
*peHód-z  > *pó:dz > *pó:d
*peHód-m.  > *pHód-m. >*pódm.
*peHod-és > *peHdés > *pedés 
 

e:~0
*kí:rd    > *k^é:rd   >  *k^é:r(d)
*k^i:rd-és > *k^6rdés > *k^r.dés

 
from *khé-ra, 'deer-tree=antler=top=foremost+*do, 'lump=organ' (cf. Egyptian H3t.y, 'heart') [H=dotted h]
*k^e(:)ród-
 
*k^e(:)ród-z > *k^é(:)rdz- > *k^e:r
*k^e(:)rodés > *k^e(:)rdés > *k^r.dés
 
 
 
 
 
o~0
*pú:nt-ah2-z  > *pó:nt&h2s   > *pónt&:h2s > *póntoh2s
*pú:nt-ah2-m  > *pó:nt&h2m   > *pónt&h2m  > *pónth2m.
*pu:nt-áh2-âs > *p&nt&'h2o:s > *p&nt&h2ós > *pn.th2ós
from *pá-na, 'flat-(indivdualizer)=path'+*tha, 'elongate'
*penótha(:)-
 
*penótha(:)-s > *póntha:s-
*penótha(:)-m > *póntha:m-
*penothá(:)s > *penthá(:)s > pn.thá:s
 
 

<PCR>
I think you can see how the example above could apply to these as well.
 
 

> I assume no long vowels until laryngeals begin disappearing.
>
>I notice you seem to be suggesting that laryngeals develop out of voiceless stops
>in some positions. Would the other voiceless stops in the same positions also generate
>*H1? or are you suggesting a tripartite process: *k, in the same process, for example
>to *H2, and *p to *H3?

I think that *p > *f > *h3 before a stop or /n/ (as in
*pnóbh- > *h3nobh- "navel").  There is no final *-p, as far
as I know, but had it existed, it would also have given
*-h3.

Final *k and *k^ give *h2.  It's also possible that *k/*k^ >
*h2 in contact with another stop (or /n/), e.g. a word like
*h2osth2- "bone" may come from *xa:stk-. But the contact has
to be pre-zero-grade (post-zero-grade *tk etc. give the
so-called PIE "thorns").

Final *-kW (from earlier **-ku) gives *-h3, but final **-ki
gives *-h1.  I explain the dual nominative (preserved in
thematic nouns) as from *-iku > *-ih3 (thematic *-á-iku >
*-é-h3, with regular loss of *i after a stressed thematic
vowel, as in the feminine *-á-ih2 > *-é-h2), while the
oblique is based on *-iki > *-ih1.

*t medially before a stop or /n/ gives *h1 (*méh1-not-),
[except that *k takes precedence: *tk > *th2].  In final
position *-t (< **-ta) gives *-h1 only when preceded by a
stressed *é (as in the ins.sg. *-ét > *-éh1), but not in
Anatolian (Hitt. ins.sg. -ét).  Otherwise final *-t is
maintained (as in the 3sg. verbal ending *-t), or weakened
to *-d (as in the pronominal neuter sg. e.g. *to-d).  I'm
not sure whether the abl.sg. (the proterodynamic variant of
the intrumental, with "svarita lengthening" of the vowel) is
to be reconstructed as *-ot or *-od.

Original **-ti gives *-y and **-tu > *-sW (Armenian -k`,
elsewhere -s).  That explains the pronominal/thematic
plurals, from (nominative) **-átu > *-ésW, **-a-atu >
*-o(:)sW, versus oblique **-áti > *-éy, **-a-ati > *-oy-.
For **-ti > *-y also cf. the numeral *trey-es(W) "3" <
**tilati-, Semitic *c^&lác^ > *t_alá:t_.
 
<PCR>
I guess we will habe to agree to disagree here.

>Do you believe that *H1, *H2, and *h3 had phonemic status, for instance, initially?

I assume you mean "initially before a vowel".  Yes, *h2 and
*h3 certainly.  *h1 is more doubtful, as there is no way to
distinguish between *h1V- and *V-.
 
<PCR>
No word begins with a vowel in Afrasian. And, accordingly, at least at the earliest date, PIE should have been exclusively CVC.

>Without the Hittite evidence, I deduce laryngeals for pre-PIE through Egyptian cognates showing j (pre-PIE *Ha/*He) or h (pre-PIE *Ho) with *H deriving from even earlier *?,*h, *¿, and *H (dotted *h).
>
>In "month", I believe like you, apparently, that it is a derivation from *me:-, "measure", with you presumably preferring derivation from the extended form *met-.
>
>I believe that *me-, in turn, is derived from *me + *?a, 'stative', so that *me:- is "measured". Further derivation with *na, 'individualizer', produces *me:n-, which, I think can be seen in Egyptian mjn, 'today' ('measured thing'), and, of course *me:n-, 'month'. Of course, the indefatigable IE's had to add more onto that!
>
>Whereupon, the collective *-t(o) was added, producing *me:'net-, with Ablaut *me'not-; whereupon nominative *-s(o) was added, producing *me:'nots-. The cluster was eliminated, lengthening the preceding vowel: *me:no:t-. *me:s- is probably *me:-  +  *s(o), nominative; *me:ns-, *me:n- +*s(o), nominative. *me:nes-, genitive, is simply *me:n- + *se-, genitive (*me:'nes-); *me:neses-, simply *me:'nes- + *so, nominative, +*se, genitive: *me:ne'ses-.
>
>But to get back to your example, does not underived *me:- suggest an *e quality in your first derivation: **mátnu:-t-?

I think the e: is from *eh1 (and in turn the *h1 from *t).
The root "measure" has variants *met- and *meh1- (which I
assume are originally from *met-V vs. *met-C-).
 
<PCR>
An awful lot of final *-t's that do not become *H1 or amything else.
 


There is external evidence in the Kartvelian word for
"moon", OGeo. ttue-, twite-, tue-, Mod. Geo. tve-, Megr.
tuta-, Laz mtuta, which I would derive from the oblique base
*mat-núta > *m&t-túta > *(m)t(t)uta.
<PCR>
I will have to pass comment on that. I know next to nothing about Kartvelian.
 

>And, in connection with the vowel-glides, do you assume as I do, that PIE *k^- is a result of pre-PIE *ke?

That would be *ki in my scheme. No.  The PIE *k^-series is
many times more frequent than the *k-series, which leads me
to believe that it stands for unmarked /k/ (/g/, /gh/).
That would make the *k-series marked back velars (uvulars)
in pre-PIE (*q, *G, *Gh).  There *is* a connection with
vowel quality, as follows:

**-ká- > *-k^é-
**-kí- > *-k^é-
**-kú- > *-kWé-

**-qá- > *-ká-
**-qí- > *-ké-
**-qú- > *-kWá-
 
<PCR>
For whatever it might be worth in this context, I have found that Afrasian back velars (Arabic q; Egyptian q, g) are the result of earlier velar nasals.
 


So after velars/uvulars, it's _almost_ possible to
reconstruct the original vowel quality, which is impossible
after other consonants.  For instance, *t:

**-tá- > *-té-
**-tí- > *-té-
**-tú- > *-té-

(but when *t follows:

**-át- > *-ét-
**-ít- > *-ét-
**-út- > *-és-
)

<PCR>
In Arabic, yes. But in Egyptian, g tells you the velar nasal was +*o; unfortunately, q only tells you it was +*a or +*e. However, a Sumerian i, a, u will specify the vowel quality; e is ambiguous, either ai or ei; but also a poor transcription for real i.
 
 
Thank you for taking the time to share your thoughtd with me. I may not always agree with you but I certainly admire your Gründlichkeit.

=======================
Miguel Carrasquer Vidal
mcv@...

Patrick