Hallo!

Henrik Theiling wrote:

> Hi!
> 
> Jörg Rhiemeier <[log in to unmask]> writes:
> > I briefly mentioned yesterday that Old Albic has a fluid-S pivot.
> >...
> > which can be both.  Fortunately, whether it is active or stative
> > can be told by the agreement markers it takes. ...
> 
> Interesting and nice system.  Do you know how fluid-S natlangs
> actually handle this?  Did you take it from a natlang?

I didn't take it from a particular natlang, but designed it when
I found that there are languages with non-nominative pivots.
Note that the pivot of a language often behaves differently
from the case marking; there are plenty of languages with
ergative case marking, but a nominative pivot.  I wanted to
create a language that is purely and thoroughly fluid-S
- including a fluid-S pivot.

> In Tyl Sjok (also fluid-S), the system is taken from Chinese, which is
> accusative, but that doesn't matter: which argument is referred to is
> purely determined by semantics

So `The child threw the ball and [] sang' can only mean that the
child sang because balls (at least, ordinary balls) can't sing...

>       and may become ambiguous.  It is not
> grammatically defined.

So a sentence like `The child threw the ball and [] fell' is
ambiguous, because both children and balls can fall?

>       So the system is quite different from Old
> Albic.

If I remember correctly, Tyl Sjok is isolating, with no kind of
person-number marking on the verb which makes it clear whether the
verb is active or stative in Old Albic.  In the first two of my
examples, the person-number marker is the only thing that's
different.  How is a sentence such as `A man wrote a letter
and [] came' resolved in Tyl Sjok?

> > This can be taken even further.  If the second verb is transitive,
> > there can be _two_ gaps coreferential with the two arguments of the
> > first verb:
> 
> HAHA!  That's great. :-)

Thanks!

> Anyway, I'd usually expect this to be handled with verb coordination,
> but still it's funny.

What do you mean by "verb coordination"?

> My first conlang Fukhian had a double-reference relative clause,
> BTW. :-)
> 
> > (3) Agratara              ndero   gratath    a   aracara.
> >     AOR-write-3SG:P-3SG:A man-AGT letter-OBJ and AOR-rip-3SG:P-3SG:A
> >     `A man wrote a letter and [the man] ripped [the letter].'
> 
> Yeah, that's fun! :-)
> 
> > To make things more complex, there are also switch-reference
> > pronouns.  The agentive case of the switch-reference pronoun, _ra_,
> > is coreferential to a patient in the preceding clause, while its
> > objective case, _ram_, is coreferential to an agent.
> 
> Nice!  Natlangs like to mark this on verbs, but Qthyn|gai, which also
> has this, also uses special pronouns.

I am not even sure whether "switch reference" is the correct term
for it.  The switch reference morphemes in Amerindian languages
mostly have to do with subordinate clauses, it seems, though
I haven't really understood those switch reference systems yet.

> I like the system in Old Albic.

Thank you!  I am pleased by your liking it.

>       My languages' grammars are usually
> less strict -- much has to be inferred from context.  Most markers are
> optional and the default is to use context.  Old Albic feels quite
> different.

The explicitness of the morphology of Old Albic may create the
impression of a very strictly regulated language, but it frees
the language in other ways.  For example, word order is almost
totally free - VSO, SVO, SOV, VOS, OVS and OSV are all equally
possible (with VSO being the "unmarked" order), adjectives can
be moved away from the nouns they modify, etc.  The fluid-S
argument marking on verbs also makes the fluid-S pivot possible
by eliminating ambiguities as in the example sentence `A man
wrote a letter and [] came'.  I simply like langauges of the
"classical" style (like Latin, Greek, Sanskrit or Quenya),
with a rich inflectional morphology and great freedom in the
realm of syntax.

Greetings,

Jörg.