Re: PIE Stop System

From: Abdullah Konushevci
Message: 25936
Date: 2003-09-22

--- In cybalist@yahoogroups.com, "elmeras2000" <jer@...> wrote:
> --- In cybalist@yahoogroups.com, "Abdullah Konushevci"
> <a_konushevci@...> wrote:
>
> I hate to criticize anything as courageous as a survey of Albanian
> soundlaws, but this was too provocative for my resistance. I guess
> it is the idea with the list to get feedback. Well, here is some:
>
> > 1. Evolution /*e/ > /je/
> > We could say that generally PIE /*e/ was diphthongized in
Albanian
> > language in /ie/: *bher > bie/bjer `to bring', *degWh- > djeg `to
> > burn', *gWher- > zjarr < zjerm/zjarm `fire', but *n-gWhre-sco- >
> > ngroh `to warm', *pekW- > pjek `to burn, to ripen', *perd >
> > pjerdh `to fart'.
>
> ngroh is a "stative inchoative" of the Latin type in /-e:sco:/
> (made:sco: 'am getting wet'), so here the -o- is regular from /e:/.

[AK]
I agree completly, but it was just a lapsus calami. As I have mention
many time, long PIE /*e:/ yields in Albanian regulary /o/. In
Albanian version, it is exactly *ngWhre:sco:-.

> > 1.1. Evolution /*e/ > /ja/
> > Bisedes evolution /*e/ > /ie/ its also common evolution /*e/
> > /ja/,
> > especialy before the nasals and liquids: *esmi > (unë) jam `(I)
> am',
> > *mel > mjaltë, but also *dek^s- > i djathtë `right', *H3ep > jap
> `to
> > produce'.
>
> /ja/ is the regular product of *e before (some) old clusters.
> Jap 'give' has optional j-, prs. jap jep jep or ap ep ep, so the
> vowel is Alb. /a/.
>
> > 1.2. Evolution /*e/ > /e/
> > If verbs are prefixed and extended in PIE causative –eyo (in
> Albanian
> > usualy –enyo), than we have /*e/ > /e/: *n-degW-eyo > ndez `to
> > light, to kindle, to fire' (cf. above *degW- > djeg `to burn'),
*n-
> > gWher-enyo > nxej `to heat, to warm up' (cf. above *gWher- >
> zjarr).
>
> The causative-iterative has *-o- in IE, thus *dhogWh-éye- > *dazi-
>
> n-dez with -e- by umlaut of *-a- from *o. - Nxej is a contamination
> of nxeh 'warm' and ziej 'boil' which probably belong with Greek
> zéo:, OHG jesan; the root is *yes-, but the morphology is hard to
> make out; not a good basis for a soundlaw.

[AK]
OK. I was not aware that the causative-iterative has o-grade form.

> > 1.3. Evolution /*e/ > /i/
> > If /*e/ is followed by consonantal cluster, the outcome is e-CC>i-
> CC:
> > *g^herzd(h)-, metathetic variant *g^hrezd(h) > drithë `cereals,
> bread
> > grains', *pezd- > pith, pidhi `vagine', *bhend- > bind `to
> convince,
> > to persuade'.
>
> drithë corresponds to Greek kri:thé:, so the immediate preform
> has /i:/; there was a relation of paradigmatic alternation between
> these words and OHG gersta, Lat. hordeum, but that is not relevant
> here. - In general terms, /i/ is the Alb. reflex of *e before
> secondary clusters, but there are quite many unexpected forms.
>
> >
> > 2. Evolution /*-en/ > /-ân/ in Gegë dialect and /-ën/ in Toskë one
> > *en esti > (g.) âshtë < ânshtë, (t.) është `is'; *g^enH-ter >
(g.)
> > dhândër, (t.) dhëndër `bridegroom, son-in-law', *po:r-ent-or >
> > perandor `imperator' (only in Gegë dialect) and (g.) perendi <
> > perandi (i-Umlaut) and (t.) perëndi `God'.
>
> Since Geg distinguishes ê and â, âshtë is rather from *an-esti "ist
> anwesend". That also applies to dhândër; perëndí etc. has no
> phonemic vowel between /r/ and /n/ (variant perndí), the form with -
> a- being a readjustment to a Vulgar Latin preform *imperantía (sic,
> cf. the definite form perëndía); the retained /e/ of the first
> syllable must reflect a secondary accent; the accent on *-ía
> reflects Greek /-ía:/ (the type Italí, -ía).

[AK]
I agree with one remark that not only you, but also Piotr and others,
write wrongly Geg instead of Gegë (also Tosk instead of Toskë).

> > 2.1. Evolution /*-em/ > (g.) /-âm/, (t.) /-ëm/
> > *tem-th > (g.) tâmth, (t.) tëmth anatomy `temple', *nem-a > (g.)
> > nâmë, (t.) nëmë `curse, malediction'.
>
> There is no precise etymology for tëmth, tâmth, certainly not with
> secured *e. - namë "ist lautlich = gr. nómo-s" (Meyer), except of
> course that the Alb. word is feminine, i.e. from */nom-a:/.
> >
> > 2.2. /-er/ > /ër/ in both dialects
> > *g^enH-ter- > dhëndër `bridegroom', *derder- > dërdëllit <
> > dërdërit `to murmur', *ma:ter > motër `sister'
>
> All three examples have unaccented [ë] which may reflect *any*
> vowel, or even zero.
>
> > 3. /se-/ > /gja-/ in both dialects
> > *serp- > gjarpër `snake', *s(w)eks- > gjashtë `six', *seuHk- >
> > gjak `blood', *sem- > n-gjamë `similar'
>
> 'Snake' and 'six' are okay, except that the /a/ is part of a
> diphthongization of *e to /ja/ (before old clusters) which is here
> obscured by the preceding /gj/. - Gjak is most often understood as
> *sokWos, identical with OCS sokU and Gk. opós 'juice', a fine
> metaphor for blood. - ngjamë or just gjamë is the ptc. og ngjas or
> just gjas, i.e. the -m- is the participial morpheme and not part of
> the root; to boot, gj- is here from gl-, Buzuk glame; the prefixed
> nasal is perhaps a reduced *som-.
> >
> > 4. Diphthongs
> > /*ei/ > /i/: *gWei-tyo- > nxis < n-zis `to encourage, to urge',
> *leig-
> > > lidh `to tie, to bind', *g^heim- > dimër `winter', peik^- >
> > pikë `spot'
> > /*eu/ > /e/: *skeud- > hedh `to throw', *dheub-to- > det `sea'.
>
> The two last are fine (except that the suffix should be *dheub-eto-
,
> the old sg. corresponding to the coll. seen in Goth. diupitha, Eng.
> depth). - I do not see why any of the examples of ei > i really
must
> have full grade. Which ones cannot simply reflect *-i-?
>
> Jens

Konushevci