Re: Help: Zhyler ECM/Raising Verbs (Longish)

Quoting David Peterson <ThatBlueCat@...>:

> Andreas wrote:
>
> << This is one of the ways to say it in Tairezazh: Ta zrón shu sha raig e'
>  zaves "I want that you eat the bread". There's no overt accusative marker
> on
>  the sentence _sha raig e' zaves_ "you eat the bread".>>
>
> This is the comment that got my wheels spinning on my eventual answer.
Happy to've been of help!

> Anyway, what you wrote in the parentheses is what gave me the rest of my
> idea, so here it is:
>
> In order to express desire of x for y to act z, you do the following:
>
> (1) X is the subject of the verb "to see" which has a volitive suffix.  So,
> /eat-VOL.-1sg./.  This verb is placed sentence-finally.
>
> (2) The action is expressed normally.  So, if it's "you eat food", it's
> /food-ACC. eat-2sg./, with the sole exception that the verb is always in the
> irrealis.  So: /food-ACC. eat-IRR.-2sg./.
>
> (3) Last, the *entire* VP is nominalized by adding a class XIV nominal suffix
> to the end of the verb.  So what you have is a verb that means something
> like, "for you to eat the food".  (I was having problems with this because I
> didn't know how a noun should mark its verb-like arguments.  In this case,
> though, the argument structure is set, and the nominal suffix is a clitic
> attached at the phrasal level, so there's no need to worry about it.)
>
> (4) Finally, the VP in (3), being the direct object of the verb in (1), is
> marked with the accusative case.
>
> So, to say, "I want you to eat the food", you say:
>
> uspan-ar us-wM-l-an-ar mat-po-m
> /food-ACC. eat-IRR.-2sg.-xiv-ACC. see-VOL.-1sg./
>
> And there you have it.
Neat, I'd say.

I'm a bit daunted by a class _XIV_ nominal suffix. How many are there, and
what differentiates them?

I guess I'd better figure out how to express this in Meghean, BTW.
Nominalizations might be the solution there to, I guess. Subject genitive
seems inescable with identical subject and possessive pronouns -
_rachoar_ "thou atest", _rachoaras_ "thy eating". I don't have a verb "to
want, to desire" yet, but spelling "want" Mezanoidically, you'd have ?_seoant
rachoaraso_ "I-want thy-eating-ACC"="I want you to eat". But what of any
object of "eat"? Object genitive is out of the question, I'm afraid.

The simple solution is to say that verbal nouns in -as can still take objects -
 _seoant rachoaraso bhado_ "I want your eating the bread". Very similar to the
Tairezan construction. It also offers the absolutely delicious question of how
to handle an object pronominal marker - does it go before or behind the
nominalizer? _Rachoarethas_ "thou-atest-it-ing" or _rachoaraseth_ "thy-eating-
it"? I'm afraid the later makes more sense.

(Where, by the way, did Meghean words acquire this tendency to explode in
length?)

With nominal objects, one could also perpetrate an ad hoc compound - _seoant
rabhad-coaraso_ "I want your bread-eating". But we'd still need something for
pronouns, and I'm not gonna allow pronominal roots in such compounds; Meghean
is sposta to be a beautiful language, not a corruption of "migraine". Or do we
need something for pronouns? Swedish doesn't use verbal nouns in this
particular context, but you can supply nominal but not pronominal objects as
the leading element in compounds. You'd simply need to bring along the actual
noun again instead of playing lazy with pronouns.

I guess the decision will be made when I for some other reason decide that -as
nominalizations can take objects - then there'll be absolutely no reason not
to go with the first model.

Or one could institute some sort of denominal adjective from expressly for the
objects of nominalized verbs. The beauty of that would be that you could slap
an extra accusative ending onto it in agreement with any on the verbal noun.
How about deriving the adj not from the uninflected noun but from that
accusative thereof, allowing for two layers of accusative endings? I think the
word is _magel_ ...

(That's Meghean for "evil" - I think you can figure out where I got that
particular shape from. If not, the archives should be helpful.)

Supposing this adj would be formed in -r, we'd then net

Seoant rachoaraso bhadoro "I-want thy-eating-ACC the-bread-ACC-ADJ-ACC"

Anyone thinking this is a good idea?

                                                        Andreas