Low animacy noun number mismatch (was Re: Epicene pronoun in english?)

Hallo!

On Tue, 9 Mar 2004 00:10:55 -0600,
"Thomas R. Wier" <trwier@...> wrote:

> From:    Philippe Caquant <herodote92@...>
> > When I learned Ancient Greek at school, we were
> > teached the following canonic example:
> >
> > *Ta zoa trekhei* (omega in zoa)
> >
> > meaning 'the animals run', or 'are running'.
> >
> > What's interesting is that 'zoa' in a neuter 3p, and
> > trekhei a 3s.
>
> Yes, it's not surprising cross-lingustically that there
> might be a number-mismatch for nouns low in animacy;  this
> is quite common.  However, in Greek this was a regular rule,
> but in no English dialect is it regular. [...]
ObConlang: In Old Albic, if the patient argument is inanimate,
the verb agreement marker is always singular, even if the argument
in question is plural:

Ndero     barasa            mbar.
man-AGT   build-3SG:P-3SG:A house-OBJ
"A man builds a house."

Ndero     barasa            mberim.
man-AGT   build-3SG:P-3SG:A house-PL-OBJ
"A man builds houses."

There is no such rule for agent arguments because such arguments
are always animate.

And on the original subject of epicene pronouns: Old Albic has such
a pronoun.  `He' is _so_, `she' is _se_, and the epicene pronoun
is _sa_.

Greetings,

Jörg.