No Subject

[And Rosta <cbmvax!uunet!ucl.ac.uk!ucleaar>]

Dan:
> Here are extracts of a discussion between And and Don, with my comments:
> >>And Rosta writes:
> >I omitted to mention _whom_. In my idiolect _who_ takes the form of
> >_whom_ when it is the complement of a preposition & when it follows
> >the preposition. This odd fact doesn't justify introducing dative
> >case into English grammar.
> eg "To whom did you give the newspaper?" but "Who did you give the newspaper
> to?"

Yes. (Furthermore, I, like most people, piedpipe only in formal written
style, so _to whom_ would be very rarely heard.)
[Piedpiping = bringing the prepositional head of a _wh_ word along with the
_wh_ word when it is placed at the start of the clause.]

> >>Hmm. In the idiolect (?) that I learned, at home and in school, _whom_ is 
> >>_who_ in the direct object, as well as when it is the object of a preposition, 
> >>just like every "non-standard" pronoun form. This makes it 
> >>"prepositional-dative-accusative" in my book, just as "me," "him," "her," 
> >>"us" and "them." I think you'll find that the analogous forms in other 
> >>Germanic languages are generally shown as accusatives (e.g. "mich" in 
> >>German, "mig" in Danish), even when (as in Danish) no noun accusative 
> >>ending exists.
> 
> It appears to me that Don is taking the traditional view and And is taking
> a more modern view espoused by some present-day theoreticians, probably
> supporters of GPSG. The point of this view, as I understand it, is that
> "me, us, him," etc. are used in all syntactic contexts except subject and 
> possessor, eg. "Who's there?" "me." or "it's me". So we might say that these 
> forms are the default forms. But even within this view, it seems sensible
> to use some word like "nominative" to describe the case of the "special" forms
>      "he, she, I", etc.
> unless one claims that it makes no sense to say that English has a nominative
> but no accusative.

I agree to this extent. Personal pronouns have a morphosyntactic feature
[+/-'subjective']; these pronouns are [-subjective] unless they are
subjects. 
 
> >_'s_ is a determiner. It certainly isn't an affix - cf. _The king
> >of England's son_. Rather, it is a clitic. It takes a common noun
> >as its complement, and, exceptionally among determiners, it has a
> >kind of subject: in the example, _son_ is the complement of _'s_
> >and _the (king)_ is the 'subject' of _'s_. Not all syntacticians
> >will agree with me than _'s_ is a determiner - some take it to
> >be a postposition - but I don't think it gets taken as an argument
> >for genitive case (though one never knows with syntactic theory).
> 
> >>I believe you'll find that a genitive case (often called "possessive" in 
> >>English) functions in precisely the same way, except that it doesn't 
> >>require a _common_ noun as a complement ("Susy's Joe, not Mary's").
> 
> And's basic point here is that "'s" in English is different from "s" in other
> Germanic languages in an essential way: it can go with an entire phrase
> rather than just the preceding word. In And's example, "the king of England's
> son", "'s" does not go with England but rather with "the King of England".
> I don't understand the view that "'s" is a determiner. I am one of those who
> consider it to be a postposition.

A singular noun given a count reading must be complement of a determiner:
   *Friend left.
    The friend left.
    Sophy's friend left.
_'S_ patterns distributionally with determiners rather than adpositions.
   *Friend of Sophy left.

There may well be good arguments against this that I'm ignorant of.

---
And