Re: Nominative Loss. A strengthened theory?

--- In cybalist@yahoogroups.com, "Richard Wordingham"
<richard.wordingham@...> wrote:

> > > No, it would be an 'antigenitive', a construct state.
> >
> > Yes, I'm beginning to see that, though it feels *very* awkward.
> And
> > I do not really understand that a relative pronoun could be the
> > thing owned by anybody or anything.
>
> In this respect, apart from case (or whatever) the relative clause
> if not very different from 'which the wolf has',
>
> > This funny thing, "antigenitive", would require to be combined
> with
> > a genitive, would it not? In that case *wlkWos- (with pronouns,
> *tes-
> > ) would be a genitive. Is this right?
>
> In the construction I have in mind, *wlkWos would be the subject,
> and thus in the nominative; we have here a clause. It may be that
> this construction does not occur in any language; we seem to lack
> interested students of grammar.

But "the man's arm" is not sensibly "the man which the arm has", as
if the arm owned the man. I find this unacceptable unless supported
by very good and compelling arguments and not just chosen to salvage
a form which is strange on its own meirts already (the alleged loc.
*yo convetying a meaning that does not make sense even then).

> In the normal possessor + possession _phrase_, I don't think there
> is a fixed rule as to whether the possessor needs to be marked io
> the possession is marked. In Chickasaw, the possessor is

unmarked,

> and both the antigenitive and other case markings may be attached

to

> the possession. I think the same applies to Basque, but I could

be

> wrong.

Well, *wlkWos, whatever it is, is not unmarked in *wlkWos-yo.

> In Arabic, the possessor is in the genitive, and case markers are
> applied to the construct form. In Hebrew, there are no case
> markers, and where _possession_ is not indicated by the
> construction, the suffixes for pronominal possessors are added to
> the noun not in the construct.

Isn't that just rules pertaining to the use of the article and so of
no relevance here?

> Hungarian has a sort of anti-genitive; the construct is
>
> possessor[-GEN] + possession-HIS
>
> HIS = 3rd person possessive suffix.
> GEN = 'genitive' ( = 'dative') marker
>
> Note the optionality of the marking.

The optional marking is with the dative case ending. I do not see
the relevance. Surely there were no possessive endings sitting on
the possessum in PIE helping to show the construction.

> In the English analogues I gave, the possessor is nominative or
> accusative.
>
> It occurred to me that tatpurushas might be examples.

Of what?

> When the
> first element is a thematic stem, why does the thematic vowel
> surface as -o- rather than as -e-?

Certainly not because of a rule pertaining to word-final position.
Inside the word the thematic vowel should alternate and does so in
verbs and pronouns, but only in isolated remains in substantives
(the matter is unsettled for adjectives). In substantives there has
been a massive generalization of the variant *-o-, so it tells us
little that this form also appears in compounds, it is just the
productive allomorph.

Jens