Jon Aske Jon.Aske at SALEM.MASS.EDU
Wed Feb 18 03:41:10 UTC 1998

On Monday, February 16, 1998 5:07 PM, J. Clancy Clements (Kapil) wrote:

> This whole discussion started by my talking about ergative-type patterns
> in Spanish.  I agree that Spanish is not an ergative language.  I think,
> however, that there are suggestions of ergative-like patterns in the order
> of intrans. clause subjects and trans. clause objects.  Also, there seems
> to be a move toward ergative patterning in pronominalization.  In the big
> picture, Spanish may be moving toward such an ergative pattern, which at
> one point could become grammaticalized.  The grammaticalization of
> discourse phenomena is not that uncommon. (Cf. the grammaticalization of
> preposing/posposing rules in VO --> OV shift for example).


I hope I'm not being too persistent, but I am just trying to understand your
position and to find out what I may be missing:

I think that even if we wanted to call this weak ordering correlation
between S and O ergative (which I wouldn't want to, for reasons I already
tried to explain), as far as I know, no language has ever developed
grammatical ergativity from such a pattern.  (Also, as far as I know,
grammatical ergative coding is never manifested as a word order pattern, but
as case marking or verbal correferencing.  Perhaps I'm wrong.)

In other words, I have not heard before of a language grammaticalizing this
type of ordering pattern, so that it results in all, or most, S constituents
become postverbal--thus patterning with the O's--while the A's remain
preverbal.  I have not heard of grammatical(ized) ergativity arising that
way and I don't think it's possible either, for the simple reason that this
correlation has a clear functional basis (focus position) of a kind that
doesn't seem to be easily obscured or extended (reanalyzed).

Also, I still don't understand why the fact that human accusatives are coded
like datives in Spanish is a reflection of ergative patterning.



PS I am familiar with DuBois' claims that the absolutive is typically the
overt argument in narrative in some (grammatically ergative) languages, but
I have not been able to duplicate those results in any language that I
mostly work on (Basque, English, Spanish), so I remain skeptical about the
generality of this finding, and in particular about this property (being
coded by a full nominal) being the source of grammatical ergativity.

Jon Aske
Jon.Aske at - aske at
Department of Foreign Languages
Salem State College
Salem, Massachusetts 01970
Wise men talk because they have something to say; fools, because they have
to say something.  --Plato

More information about the Funknet mailing list