Re: THEORY: Non-nom Subj & Nom Obj -- Quirky OVS Word Order Or Quirky Case?

Thanks for writing, Henrik.

--- In conlang@yahoogroups.com, Henrik Theiling <theiling@A...> wrote:
> tomhchappell <tomhchappell@Y...> writes:
> > [snip]
> > I hope that didn't delay your answer, if you have one.
>
> It didn't -- I had no good one.
The answer you just gave is a good one, to me.

> [snip]
> It's still a funny phenomenon and I'd like to understand it better,
Amen.

> especially would like to know whether
> there is a formal definition of
> 'subject' (maybe with special emphasis on German and Icelandic).  I
> tend to avoid that term, since I don't know exactly what it means.
>
> Anyway, using some tests, you can check that the status of the
> nominative object is different from a nominative subject, namely by
> checking whether it can be referred back to in a coordinated clause
> from an ellipsis.
> IIRC, it was Markus who mentioned this a few days ago.
I remember Markus's answer as being helpful, informative, and
evocative of further thought and study; but if he specifically
included the answer you are giving here, he must have implied it or
encapsulated it, or I lost it amid the other information he provided.

> Usually:
>
> a) Ich trinke Bier und []  esse Wurst
>    NOM        ACC      NOM      ACC.
>    I   drink  beer and [] eat  sausage
>    'I'm drinking beer and [I] am eating sausage.'
>
> (Gap marked with [].)
>
> In the second clause, the (nominative) subject is the same as in the
> first clause and can be dropped.  Order does not matter, but
> subjecthood + case does (as we saw earlier, subjecthood alone is not
> enough in German):
>
> b) Bier trinke ich und []  esse Wurst.
>    ACC         NOM     NOM       ACC
>    'It's beer I'm drinking and [I] eat sausage.'
>
> As an important test later, I'd note here that ellipsis on objects
> is ungrammatical:
>
> a') *Ich trinke Bier und []   sieht der Mann.
>      NOM        ACC      ACC        NOM
>     *'I trink beer and the man sees [beer]'.
>
> b') *Bier trinke ich und []  sieht der Mann.
>      ACC         NOM     ACC       NOM
>     *'Beer is what I drink and the man sees [it]'.
>
> or  *Bier trinke ich und der Mann sieht [].
>      ACC         NOM     NOM            ACC
>      -"-
>
> This also works with datives, btw:
>
> a2) Ich entsage dem Wein und []  trinke Bier.
>     NOM         DAT          NOM        ACC.
>     I   abdicate the wine and [] drink beer.
>     'I abdicate the wine and [I] drink beer (instead).'
'Abdicate'?!?  Dang, Germans take soft liquor kind of seriously, huh?

> b2) Dem Wein entsage ich und [] trinke Bier.
>     DAT              NOM               ACC
>
> a2') *Ich entsage dem Wein und der Mann spricht []  zu.
>       NOM         DAT          NOM              DAT
>      *'I abdicate the wine and the man does justice to [the wine].'
'Justice', too, huh?

>      Again, dative ellipsis does not work here.
>
> Again,  word order doesn't matter in the second
> clause either:
>
> a2'') *Ich entsage dem Wein und [] spricht der Mann zu.
>        NOM         DAT          DAT        NOM
>
> b2'') *Dem Wein entsage ich und [] spricht der Mann zu.
>        DAT              NOM     DAT        NOM
>
> And with genitives:
>
> a3) Ich harre des Mannes und [] friere.
>     NOM       GEN
>     I   wait  the man and [] am_cold.
>     'I'm waiting for the man and [I] am cold.'
>
> b3) Des Mannes harre ich und [] friere.
>     GEN              NOM
>
> a3') <give me a second verb that takes a genitive object and
>       I'll give you a sentence in return.>
Sorry, can't think of one right off the bat.
When I do think of one, it probably won't be in German.

> That's the 'normal' cases.
> --------------------------------------------------------------------
> Now let's use the quirky case examples and check whether they behave
> the same:
>
> c) Mir schmeckt das Bier.
>    DAT          NOM
>    I   like     the beer.
>
> Technically, the case assignment is like the first clause in b2).
> But
> still, it's different: the difference is that you *cannot* drop 'the
> beer' in a coordinated clause that would have it is nominative:
>
> d) *Mir schmeckt   das Bier und []  fällt um.
>     DAT            NOM          NOM
>     I   like[food] the beer and     falls___
>    *'I like the beer and it falls.'
>
> That's ungrammatical.  So 'Bier' is probably does not have the same
> function in the first clause, since then you'd probably predict that
> sentence to be grammatical.  This is in constrast to c).
>
> You must use a pronoun to make it grammatical:
>
> d2) Mir schmeckt   das Bier und es fällt um.
>     DAT like[food]             'it'
>
> But interestingly, you can use a gap for 'mir':
>
> d') Mir schmeckt   das Bier und gefällt der Raum.
>     DAT            NOM                  NOM
>     I   like[food] the beer and like    the room.
>
> This is in contrast to a2') etc.  Again, it shows that this is
> different.
>
> But OTOH, for some strange reason, if 'beer' is in front, it *does*
> become grammatical with an ellipsis, too:
>
> e)  Das Bier schmeckt mir und []  fällt um.
>     NOM               DAT     NOM
>
> Now this is funny. :-)
>
>
> Let's check whether it is animacy that matters:
>
> f)  Der Bus überfährt   den Mann und []  bremst.
>     NOM                 ACC          NOM
>     The bus drives_over the man  and     brakes.
>
>     OK.
>
> g)  Den Mann überfährt der Bus und []  bremst.
>     ACC                NOM         NOM
>
>     OK.
>
> So animacy probably doesn't do much here in German.
>
> Whatever all this tells you about German.
>
> *I* interpret it as a clear indication that
> it's an instance of quirky
> case of the subject and not quirky word order.  The rule being in
> German that
> the *subject* can be elided in a coordinated clause if it
> is in the same case, but regardless of what case exactly it is in.
> OTOH, non-subject arguments cannot be elided, again, regardless of
> the case, nominative included *unless* they are in initial position.
> Funny rules, but that's what the examples show.
>
> Further, for a subject,
> the sentence initial position seems to be the
> default, unmarked order,
> so you could mistake it has quirky word order
> for subjects in dative case,
As, perhaps, I apparently did?

> but it still behaves differently.
>
> And I understand that the case equality constraint does not exist
> in Icelandic:
> the subject can be elided from a coordinated clause even
> if in a different case.
>
> HTH??
That helped.
(I assume HTH stands for "(I) Hope That Help(s/ed)".)
Thank you.

Now, this is my fourth posting today, so I am not going to start a
new sub-thread, but it seems worth-while to point out that we can (to
a degree) concentrate on the

[QUIRKIEST CASES]

These would be clauses with both a Quirky Subject and a Quirky Object
in the same clause.

There are four sub-kinds,
two for accusative/nominative alignments and two for
ergative/absolutive alignments.

1) Non-Nominative Subject With Nominative Object
2) Non-Accusative Object With Accusative Subject
3) Non-Ergative Subject With Ergative Object
4) Non-Absolutive Object With Absolutive Subject

---

And, of course, there is the question of the "Hyper-Quirky";

5) Are there natlang examples of Accusative Subjects with Nominative
Objects in the same clause?
6) Are there natlang examples of Absolutive Subjects with Ergative
Objects in the same clause?

----------

Thanks for your help.

Tom H.C. in MI

Replies