position of subject in nonfinite clauses

[Jose-Luis Mendivil Giro]

On Mon, 16 Feb 1998, J. Clancy Clements wrote:

>I have another question which I can't figure out a functionalist-oriented
>response to.  In Spanish, nonfinite clauses (i.e. infinitival and
>gerundive clauses) cannot have a subject preverbally.  For example,
>
>Por decir estas cosas mi abuela...
>because-of say these things my grandmother
>"Because of my grandmother saying these things..."
>
>*Por mi abuela decir estas cosas...

And Jon Aske:

>I must say that this last starred clause doesn't sound all that
>bad to me, so let me hear from other speakers.  Maybe some other language is
>interefering with my Spanish here.

I fully agree with the asterisk.

Looking for a functional explanation, Jon suggests:

>Now, since the notion topic (as well as the notion focus) is not really
>relevant to non-asserted clauses, it is not surprising that the subject is
>not preverbal, since preverbal position is typically correlated  with topics
>in Spanish and postverbal position is associated with everything else.

But the issue is that such preverbal subjects are never allowed, asserted
or non-asserted:

(i) *Mi abuela decir estas cosas...

If we consider Aske's last example (ii) we see that _mi abuela_ is in
preverbal position, although _unasserted_:

(ii) Porque mi abuela dijo estas cosas, nos fuimos todos a casa.

So, despite things being much more complex, the reason may be formal, what
explains Clements' difficulties. I do not mean there is not a functional
explanation (there is always, soon or later, a functional explanation if
one looks for it); what I mean is that in this case, as has been
traditionally observed, the imposibility of a infinitival preverbal subject
relates to a formal aspect: agreement. So, subject-agreement (or in
generative terms, abstract nominative case assignation) is what licenses
the subject. If there is no agreement (as in Spanish infinitives) there is
not preverbal (licensed) suject.
If my starred example (i) is used in an embedded clause, the subject will
_raise_ to preverbal position:

(iii) He oido a mi abuela decir estas cosas...

This phenomenon is conceived of (in formalist contexts) as _exceptional
case marking_ (it is assumed that the objective (abstract?) case of the
main verb licenses the subject, that then raises). Then, the reason for the
postverbal position of the subject in (iv)

(iv) Por decir estas cosas mi abuela...

could be related to the unability of the non-agreeing infinitive to assign
nominative case (i.e., to license its subject).

Of course, what matters here is that I think that this (incompletely
presented and old) formal explanation is in no way incompatible with that
suggested by Jon, but that they are complementary.
In this way, formal restrictions ensure the _functional role_ an utterance
has to play, what does not mean that these formal restrictions should be
necessarily reductible to that discursive/perceptive function. In other
words, agreement could be functionally motivated in human languages, but
not (unless rather indirectly) the position of infinitive subjects.

Saludos cordiales, and, please, excuse my net-spanglish.

_______________________________
Dr. Jose-Luis Mendivil
Linguistica General
Universidad de Zaragoza (Spain)