Re: [tied] Re: Short and long vowels

----- Original Message -----

From: elmeras2000

To: cybalist@yahoogroups.com

Sent: Friday, July 15, 2005 7:09 AM

Subject: [tied] Re: Short and long vowels

--- In cybalist@yahoogroups.com, "Patrick Ryan" <proto-
language@......> wrote:

>   The way I came across this problem initially was attempting to
understand why a word like *ayos-, 'metal', shows up with short <a>.
According to my hypothesis, /a/ was _not_ retained from Nostratic
unless it was lengthened. It could be lengthened by a 'laryngeal' or
by an aspirated stop (*ph, *th, *kh) or other aspirates when  they
gave up their aspiration. If Nostratic <a> were not lengthened, it
would fall into the regular Ablaut of *e/*o/*Ø. For PIE *ayos-, I
reconstruct Nostratic *hayés-.

Why not *H2éy-es-/*H2éy-os-?

 

***

Patrick:

 

Because I think that, of the four laryngeals in Nostratic (/?, h, ¿, H/), by pre-PIE, only two remained: /?/ and /h/, neither of which had 'coloring' properties.

 

In my hypothesis, PIE *a/*a: can only be there because Nostratic *a was lengthened, and so retained its vowel quality; Nostratic short *a (like short *e and *o) would have become the Ablaut vowel.

 

In order for PIE *a to have been retained, it must previously have been lengthened.

 

Therefore: *hayés- >  *a:yés- > *á:yos- > *áyos-.

 

***

>
>   ***
>
>
>   >   The problem I found with this is that where I expected Ca:C,
I
>   frequently found *CaC, etc.
>
>   The weak grade of /a:/ will be /a/ by regular development. You
must
>   mention the material if you want a reaction to your impression
of it.
>
>   ***
>   Patrick:
>
>   I am not trying to argue the merits of this idea, I simply
thought it might be helpful to explain the long/short variation
_you_ detected.
>
>   I believe *hayés- passed into pre-PIE as *a:yés-, and, with
transfer of stress-accent to the initial syllable: *á:yos-. Now
there may have been an **áyos- from another Nostratic root in PIE
but we do not know of one. Therefore, by economy of effort, *á:yos-
could be shortened to *áyos- without any loss of root integrity; and
was.

I do not see the necessity of any of these steps. IE s-stems are
generally held to be underlyingly root-stressed, the original
paradigm being apparently proterodynamic. That makes "*hayés-"
suspect. Why did you depart from precisely that form?

 

***

Patrick:

 

For two reasons:

 

1) I have identified, I think, two major sources of  -*s-stems in PIE: -*s, 'state or condition', and -*s, 'color adjectives'; I think this is from the latter;

 

2) I believe originally short PIE *o arises from *é from which the stress-accent has been removed.

 

3) the analytical development of this word, as I see it, is *há, 'water' + *ye, '-like' = *háye, 'water-like' = 'wet, reflective, bright' + *so, 'skin', forms color adjectives/nouns = *hayés(o), 'metal(lic), reflective substance'. > pre-PIE *a:yés-. Each new morpheme added onto the right shifts the stress-accent one place to the right; so, for polysyllables, basically a penultimate accent. But then, probably because of contact with a root-syllable-accenting language speaking group (Uralics? Basques?), root-syllable accentuation became the rule for PIE, with formerly accented *é becoming *o: > *á:yos-; with no competing **áyos- from another source > *áyos- by 'economy of effort' principle.

If you would want to provide an example of a -*s-root that you feel contravenes my hypotheses, I would be glad to _try_ to explain it.

 

The one that comes immediately to mind for me is *awes-. This is a little different accentually because, I believe, the -*s came from *-sa:, the earlier form of the 'condition or state' suffix. I hypothesize: *há-wa: > *há-wa + *sha = *a:wésa: (any short vowel, when stress-accented, becomes *é) > *a:wésa (final stress-unaccented vowel are shortened or lost, if short already) > (root-syllable accentuation leftward shift prevented by voweled final syllable) which is subsequently lost *a:wés-. PIE *(a/a:)us-o:s- is, I believe, a result of reduplication: *a:wos-á:wos-. I suspect that the earliest semantic reference of this word was to the dew that forms at dawn rather than the onset of sunlight itself.  There are many associations of this word with the planet Venus, which, our ancestors believed, was connected with moisture and water.

 

***

 

>
>   ***
>
>   >   After some thought, I applied what I think is a principle of
>   language development: economy of effort.
>   >
>   >   Where there was no homonymous *CaC, *Ca:C could be shortened
to
>   *CaC with no loss of root integrity; and frequently was so
>   shortened; when there was such a root, resistance to shortening
the
>   vowel was much stronger.
>
>   I would like to see your expose in concrete terms. While I
>   understand your suggestion for a principle quite well, it does
not
>   correspond to anything I have noticed. Therefore you need to
>   demonstrate it.
>
>   ***
>   Patrick:
>
>   On your two examples:
>
>   I also believe that *nas-, 'nose', should be reconstructed
with /a:/ but I believe that the reason it has /a: is that
a 'laryngeal' was involved: Nostratic *na?s-.

Does this mean: Nostratic *-a?- > PIE *-a:- without laryngeal in
PIE? I would be ready to accept that, provided there is some
evidence showing it.

***

Patrick:

 

Well, my reasoning here will probably be a little strange but based on other hypotheses, any long vowel after *n can _only_ come as a result of a 'laryngeal' after the vowel; any aspirated Nostratic *nh* became *l in PIE (in Sumerian also; though it shows up as <n> in Egyptian) so *nV: has to reflect a following 'laryngeal' or deletion of some other consonantal phoneme. Though you may not agree, for my purposes, pre-PIE /?/ is also a 'laryngeal'. I do not believe, as Bomhard does, that long vowels were present in Nostratic. Any long vowel in PIE, therefore, is a result of a short Nostratic vowel + preceding lost aspiration or a following 'laryngeal' or compensated consonant.

 

I speculate that the process involved is basically 'any hemimora is retained': if the *C is removed from any VC* closed segment, the *V is lengthened by one hemimora: *VV = *V:; I consider *CV and *V: one full mora.

 

This particular root is one that derives from (Northern?) Nostratic but does not seem to be present in either Egyptian (as a representative of AA), or Sumerian so direct proof is not possible.

 

But any PIE root that has a voiced initial plosive, *b, *bh, *d, *dh, *g, *gh which shows *CV: is a result of Nostratic *CVL(aryngeal). In the long discussion we had over the significance of *CV?- roots, which you showed rather conclusively, could not be 'statives' for PIE as I had postulated, what we actually see is *CV:-. I still think they had a stative significance in Nostratic, and are the result of *CV + *?a, a 'stative' suffix, but this was obviously lost by PIE; and we can consider PIE *CV: a lexicalized root without stative significance. 

 

>   I am afraid you will like my explanation of *sal- even less.
Here, I believe the pre-Nostratic form was *sHala-, i.e. with
aspirated /s/: Nostratic *sa:l-.

Does this mean: Nost. aspiration + /a/ > PIE /a:/ without
laryngeals? I can accept that too if it is supported by evidence.

***

Patrick:

 

Yes, if we mean 'lost' aspiration.

 

As for evidence, this may be the hypothesis that you will have the most difficult time accepting because it contravenes current PIE theory so much. I believe that voiceless aspirated plosives must be reconstructed for PIE: *ph, *th, and *kh. Aside from the sproadic evidence in Old Indian, they can be detected by the lengthened vowel in PIE but, with the lengthening effect of suffixed -*?a, this is a very tricky proposition. It is really only through Egyptian that an aspirated voiceless plosive can be confirmed. If we have a PIE *ta:, we can only know that it is from *t(h)a rather than *ta? if an Egyptian cognate shows <D> rather than <d>. And then there is the semantic 'evidence'; *tha means something slightly different than *ta.

 

 

***

 

 

>   Not having to contend with **nas- or **sal- from a different
root, they could be shortened without sacrificing root integrity,
i.e. unambiguous semantic reference.

But *were* they shortened? It seems to me they just alternate a:/a
by normal rules.

***

Patrick:

 

Well, of course, once we have a short /a/, it might be drawn into the regular Ablaut series /e-o-Ø/.

 

PIE *mel-, 'beat', for example, I believe derives from Nostratic *male:- (Egyptian mn, '*mace'); it forms no exceptions to regular Ablaut.

 

My impression is that /a:/ will not reduce farther than /a/, however.

 

On the other hand, we have PIE *ansu-, 'spirit', which, I believe, derives from Nostratic *ngo-só-wa. This produced *n(g)séu, and finally with the root-syllable stress-accentuation, *'n(g)sou-, to which a prothetic *a- was added: *ánsu-. This prothetic vowel _could_ disappear with regular further shifts: *Nsú- (cognate: Egyptian gs, 'friendly, (by the) side').

 

*** 
>   Before I offer an opinion on "*ste:w- -> *sté:w-/*stéw-", could
you confirm for me if this is the root meaning 'thicken'?

It isn't: *stew- 'bekannt sein, preisen' (Kümmel in LIV,
although 'bekannt sein' is plainly just the mediopassive). It's one
of the most classical examples of Narten ablaut.

***

Patrick:

 

Yes, I would have to agree on all counts. I would hypothesize that the Nostratic form was *dá-wa, seen in PIE *deu-, 'honor, praise' (cognate, Egyptian dw[3], 'praise'); with *s-mobile, *ste/e:u-, 'praise well/loudly. this long vowel is a result of purely internal PIE processes  

I do not expect that you will be able to agree with much of what I have hypothesized but thank you for the opportunity to expostulate it.