Ergativity

[J. Clancy Clements (Kapil)]

On Thu, 12 Feb 1998, Ricardo Maldonado wrote:

> Finally, I am not clear about what Clancy Clements  means by the relevance
> of  LEISMO to this specific topic and how that would constitute an argument
> for an ergative interpretation of intransitives. I would be very interested
> to know his reasoning on this topic.

Ricardo, Diego, Jon, and other funknetters,
Here's what I had in mind with leismo suggesting a Primary Object -
Secondary Object pattern.  I'll cite data from Klein-Andreu and then
offer a brief explanation.

LE/LO pronominalization of masc. direct objects

                Animates                Inanimates
                le   lo %le             le  lo  %le
St. Teresa      83   2  98%             22  3   88%

Professional
Women           67   6  92%             16   21 43%

Rural Speakers  36   0  100%            22   6  76%

"Frequency of le as a function of the referents' animacy in (1) writings
of St. Teresa; (2) speech of present-day Castilian professional women; (3)
speech of present-day Castilian rural speakers. Only count NPs were
considered."  From Klein-Andreu, Flora. 1992. Understanding standards. In
Garry W. David and Gregory K. Iverson (eds.) Explanation in historical
linguistics, 176-78.Amsterdam & Philadelphia: Benjamins.

The point is this: monotrans. DOs and ditrans. IOs are marked
increasingly by LE, in some cases (e.g. rural Castilian speakers)
exclusively by LE. This suggests that a Primary OBj. - Secondary Object
system is being adopted for masc.  In future, this may or may not extend
into the domain of the fem DO.  At present in rural speakers, we have the
following:

Primary Object:  LE (masc/fem. IO; masc. monotrans. DO) and LA (fem DO)
Secondary Object: LO (masc.) LA (fem.)

There is some evidence that LE is found with fem. DO.  In Baroja, for
example, we find the following (from Marcos Marin _Estud. sobre el
pronombre_, 1978, p.255):

LE D.O.
masc. person 131
fem person 6
animals 6
thing 1

Baroja was Basque and, consequently, there might be something going on
there (cf. Alazne Landa's dissertation on Basque Spanish for substrate
influence of Basque on Spanish of that region).

In his analysis of the magazine ABC as an example of modern Castilian,
Marcos Marin says there seem to be two pronominal systems in Castilian:
the etymological one (the DO-IO system), and a second system where the
masc. animate DO is being replaced by LE. He discovered, however, a third
system as well, in which, he says, LE has triumphed as the pronoun for the
masc. and "is invading dangerously but not overwhelmingly the domain of
the feminine..."  The feminine DO pronoun LA is, acc. to Marcos Marin,
gradually losing ground to LE.

In my estimation, the Castilian pronominal system seems to be moving
towards a PO - SO system, where all monotrans. animate (and gradually
inanimate) DOs are pronominalized by LE(S), all IOs are pronominalized by
LE(S), and all ditrans. DOs are pronominalized by LO(s)/LA(s).

Now, as for full objects NPs, *Luisa presento' Juan a sus padres* is
admittedly odd (though found acceptable by some native speakers I
consulted).  Another option *Luisa presento' a Juan a sus padres* was
preferred by another native speaker I consulted.  Still other native
speakers said they'd avoid the construction, which coincides with Victoria
Valzquez' findings, which Jon was nice enough to translate and share with
all.  The conclusion regarding full NP marking, then, as reflecting or
suggesting a PO - SO system instead of a DO - IO system, in inconclusive.
I find the pronominalization facts more suggestive of a PO - SO system.

Finally, please disregard the _visito a Paris_ example, which is, as
Ricardo pointed out, irrelevant for the discussion at hand.

Saludos,

Clancy