Possessive Antecedent Again

[Arnold M. Zwicky]

On Apr 23, 2004, at 6:11 AM, Geoff Nathan wrote:

> I have no intention of stirring up this debate again, but I read a
> sentence
> this morning where there was an alarming ambiguity that I realized was
> induced by this construction and it brought me up short.  It was on
> the CNN
> website
> (http://www.cnn.com/2004/WORLD/europe/04/22/amorous.rhino.reut/
> index.html)
> and dealt with a rhinoceros attempting to mate with a car.  In the
> sentence
> below there is a serious problem with the coreferent of 'he' in the
> 'when'
> clause:
>
>> Sharka, a two-ton white rhino, got amorous with Dave Alsop's car when
>> he
>> stopped with three friends to take pictures of the animal mating with
>> his
>> partner Trixie at the West Midland Safari Park.

as a staunch defender of possessive antecedents, i have to point out
that the problem here (if there is one) really has nothing to do with
possessives.  the problem is that there are two NPs, "Sharka" and "Dave
Alsop", that establish discourse referents to which the later pronoun
"he" (and, for that matter, the later definite NP anaphor "the animal"
and the pronoun "his") can refer.  things are just the same with
"...got amorous with a car belonging to Dave Alsop", which doesn't have
a possessive in it.

in fact, if you *really* believe in the Possessive Antecedent
Proscription (as no one actually does in practice, despite their
pronouncements, but let that pass), then the referent of the "he" is
unambiguously *not* to dave alsop, but to sharka the rhino.  this is
not the intended meaning, or the one any well-intentioned reader would
get, but to a true believer in the PAP this is the only one possible,
so that the problem with the sentence isn't ambiguity, but
ungrammaticality (on the intended reading) -- not two interpretations,
but one wrong one.

> For your amusement on this Friday near the end of the American
> semester.
> Geoff

well, here in american quarter-land, we're just halfway through spring
quarter.

arnold (zwicky at csli.stanford.edu)