Re: split ergativity

Quoting "Thomas R. Wier" <trwier@...>:

> Doug Dee wrote:
> > Dixon says (p. 14) "However -- and this is a most interesting and
> > significant fact -- no language has thus far been reported that is
> > fully ergative, at both morphological and syntactic levels."  So,
> > although it may not be a question of "splits," Dixon supports
> > Andreas's point that no language is entirely ergative.  Dixon lists
> > Basque among "languages with some ergative morphology but an
> > entirely accusative syntax." (p. 175)
>
> For some reason, I never got this email in original form.  I think
> a lot of this discussion is wrong-headed, because it assumes that
> languages conform to some kind of platonic 'type', rather than
> being subject to a number of independent competing forces.  All
> languages tend to show both accusative and ergative characteristics,
> and differ simply to what extent they grammaticalize that fact.
But if accusative is easily the most common pattern, and mainly ergative
languages are more prone exhibit accusative traits than mainly accusative ones
are to exhibit ergative ones, it's hard to avoid the suspicion that the forces
clamouring for accusativity are typically the strongest. Founder effects and
pure chance could certainly account for the imbalances observed, but that does
hardly seem like the most likely explanation, does it?

                                                           Andreas