Re: [tied] All of creation in Six and Seven

> >> The suffix -at(u) is not only a feminine marker, but also a plural/collective.
> >
> >How do you get that? It is contrary to e.g Classical Arabic cases like
> >shajarun 'tree (collective/material)' vs. shajaratun '(one piece of) tree'
> >or H.ajarun 'stone, rock (collective)' vs H.ajaratun '(one individual)
> >rock' and other cases with animals.
>
> Correct, -at- is also a singulative marker in Arabic (and Hebrew). The
> usual function of -at-, besides marking the feminine, is to make abstract
> nouns.

The -at- for abstract nouns somehow sounds familiar but I can't really
see it as definite. In Ar. the plurals for some masc. nouns have -at-
and I guess those can be called collective. But those classes are very
limited and I don't know it ouside Arabic. Some infinitives have -at-
but that's clearly syllable-conditioned. What else are you referring
to?

> It's not a big step from abstract nouns to collectives (and vice
> versa),

Yes it is if the same language has -at- for the opposite denotation...
Unless you posit that the functions belong to different stages of
the language etc.

> >> The noun being counted is in the genitive plural, so "3 X's" in the
> >> masculine was:
> >>
> >> *Tala:T-at-u X-i:-n
> >>
> >> with the numeral in the nom. of the plural collective (-at-u) and the noun
> >> in the gen. of the (broken) plural (-i:-).
> >>
> >> The feminine we would have expected to be:
> >>
> >> **Tala:T-at-u X-a:t-i,
> >>
> >> but is in fact:
> >>
> >> **Tala:T-u X-a:t-i,
> >>
> >> with deletion of one instance of the suffix -a(:)t-. We can compare the
> >> deletion in the feminine plural of the plural definiteness marker -n.
> >
> >But this analysis leaves 11-19 unexplained because there we do have
> >similar (if not identical - just an intervening 3ashrun) opposition
> >between the masc. and fem. forms but the noun being counted is in the
> >acc. sg.. No construct or gen. pl. here (but I am not sure it goes back
> >to PSem of course).
>
> I was assuming that was analogical after the tens (20-90) which take the
> counted noun in the acc.sg.

Which is not a bad idea. That 3ashara and the tens (20-90) all end in -a
is enough to single out that class to take on the analogy.

Another more obscure problem is that the choice of masc. fem. form
in 3-10 really is gender-conditioned and not form-contitioned. In
Class. Arabic there are not so few masc. nouns. with -at- form
or that usually take -at- adjective agreement (i.e inanimate broken
plurals) and vice versa with feminines without -at-.

> I haven't given much thought to where that
> comes from.

I have no idea either but 'kam', the teens, and 20-90 "cannot" be put
in the construct state.

all the best,

Harald