Quoting Garth Wallace <[log in to unmask]>:

> I've heard that antipassives are solely found in ergative languages. Is
> this true?

I don't know, but read on ...

> I ask because I'm currently working on a language that has strict word
> order (SOV, S-PO-SO-V for ditransitives), and a couple of voice
> suffixes: a passive that demotes the subject and promotes the object(s)
> (of transitives; intransitives end up ambient), and something I've been
> calling an "antipassive/applicative" that demotes the primary object
> (and promotes the secondary object to primary object position in
> ditransitives). I'm not sure how plausible this is. It's supposed to be
> like the difference between "He gave the dog a bone" and "He gave a bone
> (to the dog)" but marked with a suffix.

My conlang Altaii, which is syntactically accusative, has something I call
an "antipassive". It's described in some more detail at , but basically you have to add
the suffix -eiz- to the verbal root if you drop the object (the lang only
distinguishes one kind of object - any arguments beyond subject and object
must be introduced by prepositions). I'm not aware of any anadewism for this,
but I can't see why it shouldn't work.