Re: [tied] The "Mother" Problem

--- In cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@...>
wrote:

> No, PIE *ph2té:r contained only one phonological vowel (as opposed
> to *mah2tV:r).

I understand that. I'm just saying that *pxté:r seems in accordance
with the apparent phonological properties of PIE, while *méxte:r and
*bhréxte:r do not.

> I don't know of any such. But the existence of something like
> *mama, *mamma, *ama or *amma for 'mom' is almost guaranteed by the
> general tendencies of baby-talk (not to mention the existence of
> such items in historically known IE languages). For 'dad' we find
> words of the same structure but with *t or *p instead of *m.

Do you think there's a rhyme or reason to that baby-talk?

> The forms with unstressed *e: do look relatively young and
> analogous to such regular cases as *ph2té:r, *dHugh2té:r, etc. I'm
> inclined to reconstruct *bHráh2to:r and *máh2to:r for PIE, with
> branch-specific analogical levellings of stress and vowel quality
> in post-PIE times.

Yes. However, that begs the question: what caused the root- and
ending-accented alternations in the first place?

> > I also hypothesize that the suffix *-xter is a reanalysis of the
> > word *pxté:r from **p(e)x-tér- to **p(e)-xtér-. Once the meaning
> > changed from "protector" or "provider" to "father", it could be
> > extended to other kinship terms.
>
> Yes, that's a common view.

Do you agree?

> Alternating with *-tor-s > *-to:r if the stress was on the verb
> root. "Originally" is a dangerous word. The reconstructable forms
> of the suffix are PIE, but of course one may speculate ad libitum
> about its pre-PIE prehistory. A variant with *l rather than *r is
> also marginally attested -- cf. the *-tlo-/*-tro- alternation in
> the (related?) suffix of PIE nomina instrumenti.

Again, what caused the distinction between unaccented *-tor-s and
accented *-tér-s? It seems like the former could not have been added
during the stage when the zero-grade was created.

> Many people have proposed such a connection, but I'd like to see a
> convincing explanation of the *bHrá- part with its full vowel and
> retracted stress. If PIE-speaker had intended simply to emphasise
> the etymological link between *bHer- and 'brother' by applying some
> kind of vr.ddhi reinforcement, the word would surely have ended up
> as *bHérh2to:r. The vowel was inserted in the wrong place, which is
> strange.

I agree. The only real possibilities, given the nature of IE
phonology as we know it, are *bhr.xté:r and *bhérxto:r. The latter
seems to be a standard agent noun "bearer, carrier". Perhaps the
reanalysis of *pxté:r from *p(e)x-tér- to *p(e)-xtér- also resulted
in the suffix having a fixed form. Then this suffix was free to be
attached to a supposed *ma- and also *bher-. However, if I
understand correctly, having a zero-grade form like *m.xté:r would
cause the /m/ to be syllabic, which maybe caused problems with the
transparency of the word. Or the vowel was retained due to the
primary element being of "baby-talk" origin (i.e. *mama or something
similar).

- Rob