Slow and not too bright

[Arnold Zwicky]

i was totally baffled by this discussion, until i found the entry
"pronoun with possessive antecedent" in MWDEU, where bernstein 1971 is
cited as referring to three handbooks (barzun 1985, simon 1980, and
ebbitt & ebbitt 1982) that declare "a pronoun cannot take as an
antecedent a noun in the possessive."

MWDEU's staff had not discovered where the rule originated, but
speculate that "it is likely to have been with one of those
18th-century appliers of logic to language."  that is, what we have
here is not a restriction that at some time governed speech and
writing in english, anybody's english, but a "rule" devised by
reasoning from first principles by language critics who proposed to
constrain the language in new ways.

the underlying idea seems to be that pronouns should agree in case
with their antecedents.  i'm not aware of *any* language where this
is true - in particular, it's not true of latin, the usual suspect
for proscriptions in english.  indeed, such a restriction would
seriously limit the utility of pronouns, *especially* in a language
with a rich case system.  it's really a very silly idea.

i was going to collect examples of pronouns diverging in case from
their antecedents, using today's New York Times as a corpus, but there
were just too many of them.  in fact, the very first example i came
across had a possessive antecedent and a nominative pronoun: "Now
it is Bill Frist's turn - but he should think twice before acting."
(norman ornstein's op-ed piece, p. A27)

the true generalization is that the possessive version of a NP
counts as an instance of that NP.  period.  and i don't think that
things were ever different in english.

arnold (zwicky at csli.stanford.edu)