Re: [tied] Bader's article on *-os(y)o

--- In cybalist@yahoogroups.com, "elmeras2000" <jer@...> wrote:

> No, the vowel could have been triggered after the accent adjustments
> were over. I derive both forms from earlier forms without vowels in
> their desinential segments. That makes them morphologically regular,
> and it makes them both "strong" paradigm forms.

I'm sure that that is possible; the question is, is it indeed what
happened?

> Very interesting! For the nominative plural this would not matter at
> all of course, it's all about the accusative. It's not a thing I

have

> observed, but perhaps one should be on the look-out.

I'm glad you think so. :) The only other language (to my knowledge)
that has case-number order is Classical Arabic with plural marker -na
following case endings.

Another problem if we follow Greenbergian Universals: the order of
affixes in your reconstruction *-z-c for nom. pl. would be backwards,
and should then be *-c-z (that is, plural + nominative).

> (You don't mean "not", I take it). We also have lengthening after
> stems in stops where the sibilant is retained. I therefore do not

find

> the assessment compensatory adequate, but that's a minor point.

Sorry, I had retyped that part and forgot to leave out the "not."
Would *re:gs 'king' (< **reHgs?) be an example of what you mention?
If not, what are some examples?

> The nominative marker is lost after lengthened vowel followed by /n,
> r, y/, perhaps /m/, probably not /l/, and certainly not /w/. You may
> work out your generalization on that.

I thought the nominative marker was retained in y-stems (i-stems)?

> You mean the form in *-es reflected by the nominative plurals of

most

> or all IE languages was not the nominative plural? It *is* possible,
> but what's the point then?

I think it's possible that *-es is a postfixed pronoun with some kind
of plural meaning. In any case, the *-es nom. pl. is from earlier **-
Vt (or, in my interpretation, pronominal **Vt), which would link it
to the plural suffix -t in Uralic. We could then posit a rule (as
Glen has done) whereby word-final /t/ ultimately becomes /s/ in some
prestage of IE, giving 2sg *-(e)s. This rule would have ceased
functioning before the adoption of 3sg *-(e)t (presumably from
demonstrative *to-).

> That would not explain (1) the o-form of the suffix of *H2ák^-mon-

es,

> (2) the presence of the unstressed short vowel in the *-mon-es, (3)
> the shortness of the surviving vowel in *-mon-. This is all

explained

> by the double sibilant ending. That also gives the nominative

plural a

> sensible marking, which I miss in your suggestion.

Hmm. The nom. sg. is *xákmo:n, later *ákmo:n, stem *(x)ákmon-
(thinking to myself). If the lengthened /o/ is due to compensatory
lengthening from loss of *-s, then the short unstressed /o/ is
explained. What is not explained is the difference between *(x)
ákmo:n (< *xákmons) and *xWnómn, if the latter is to be scanned *xWnó-
mn and not *xWnóm-n. So the question is, is the suffix of 'name' the
same as that of 'stone'?

> These numerals are adjectives, so it's no great wonder that they are
> inflected in concord with their head nouns.

Right. So where did the *-(w)or suffix come from in **kWetwor 'four'?

> They both are. Benveniste wanted to identify them.

Okay. Are they related (one and the same), or did they happen to
become homonyms due to sound changes?

- Rob