PSAT Glitch

[Arnold Zwicky]

P-A-T:
 >the purpose of my response was not to support Keegan's position.
 >Rather, it was to ensure doubters that Keegan did not make this rule
 >up.  It does indeed appear in older grammar books.

so it turns out, to my great dismay.

yesterday, a colleague whose first language is not english told me
that she had been taught this proscription in her ESL classes, and
that it continues to haunt her.  every time she writes a possessive NP
in english, she worries about whether it's ok to use a pronoun to
refer back to it.  what a senseless waste of time and attention!

i have a large collection of grammar-and-usage books, most of them
college handbooks.  i don't recall this proscription being in any of
them, but i'll check.  it certainly is *not* in fowler, gowers,
garner, and so on - what i'd been taking as the top of the line in
prescriptivist usage manuals, and what i'd been naively taking to be
representative of the "tradition" in such things. (MWDEU is widely
castigated as being the work of evil descriptivist linguists.)  now it
turns out that there's a seamy underside to the tradition.  (maybe i
should have guessed that there was a dark side.  after all, fowler,
gowers, garner, etc. refuse to issue blanket condemnations of split
infinitives and stranded prepositions, so how much can we trust
*them*?)

(by the way, my colleague had been taught that possessive NPs, in
general, cannot serve as antecedents of pronouns.  in the discussion
here, we've gotten a more restricted formulation, namely that *NP's
possessives* are barred as antecedents.  the more general formulation
would permit possessive *pronouns* like "their" and "his" and "her"
as antecedents, since they're not NP's possessives.  so, on the more
restrictive formulation, "their writings made them famous" would be
ok, but "Toni Morrison's writings made her famous" wouldn't be, while
on the broader formulation, they'd both be out.)

i'd been aware of the tendency for high school students to have been
taught that grammar and usage are (almost entirely) a matter of
*avoiding error*; writing a "correct" essay is not getting graded down
on mistakes, and taking multiple-choice exams like the PSAT, some of
which are exquisitely important to students' futures, is even more a
matter of learning lists of prohibitions, most of which make no sense
to them.  (but then a significant part of their education requires
them to be compliant, to perform well on tasks that seem largely
senseless to them but show that they are Good Soldiers.)  i understand
that they've been told to take their audience into account and all
that, but when it comes down to the multiple-choice exams, it's just
grinding out the "rules".  they arrive in college classes, even
sometimes at stanford, viewing the task of writing as negotiating
their way through a minefield.  it's very distressing.

i'd assumed that multiple-choice exams would view all split
infinitives, stranded prepositions, and contracted auxiliaries (for
example) as ungrammatical, despite the opinions of fowler and other
top-line prescriptivists.  the principle at work is that If It's
Sometimes Inappropriate, It's Always Unacceptable.  it's sort of like
traffic laws, which apply even when their rationales don't apply: you
have to wait for a red light to turn green before crossing an
intersection, even in the middle of the night when you can see empty
streets for half a mile in every direction.  (my thanks to geoff
pullum for this analogy.)

this is a stupid attitude to take towards language use, but i'd known
it was there; after all, i've dealt with copy editors who insisted
that "since" and "while" cannot be used as logical, rather than
temporal, connectives, on the grounds that sometimes (it takes some
ingenuity to concoct the situations) an ambiguity would be possible.

what i hadn't realized is that there was a whole world of
proscriptions, the nature of which i hadn't appreciated, out there,
and that students were being judged on their conformity to them.  is
it really true that ETS's grammar/usage questions treat as
unacceptable any usage that anyone who's imaginably an "authority" has
branded as over the line?  if so, what's the point?  ("i know it hurts
terribly, and you don't understand why now, but it's for your own
good"?)

 >And that rule is not based, as Zwicky claims, on an assumption "that
 >english NP's expressions are derived, not inflected."  I am well
 >aware that English NPs occur in both these forms.  In the particular
 >example cited ("Toni Morrison's genius enables her to create novels
 >that arise from and express the injustices African Americans have
 >endured"), one has to take into account not just the semantic
 >features of the inflected NP, "Toni Morrison's," but also those of
 >the head noun, "genius," (and, often, the verb) in order to
 >determine the true antecedent.  Had the sentence been, "Toni
 >Morrison's sister left one of her books on the desk," then, the
 >antecedent would not be so clearcut.  The point of the rules , or
 >generalizations, proposed in these earlier grammar textbooks was, I
 >think, to give students a tool which they could used to apply in
 >most instances.

hey, give us a break.  we're trying to figure out the rationale for
the proscription from what you say - remember that most of your
readers have never even heard of this "rule" - and this isn't easy.
your earlier discussion, comparing the possessive examples to the
"noisy" example, suggested to me that anaphoric islands were the issue
(which is why derivational rather than inflectional morphology came
up; in my first bash at the rationale, i assumed that case inflection
was the issue, but then your analogy to "noisy" led me to believe this
was wrong).  now it seems that the rationale has to do with ambiguity:
if some occurrences of a pronoun with preceding NP[poss] + N could be
interpreted, out of context, as referring to the N *or* the NP, then
all such occurrences should be barred.  this is an instance of
reasoning from the principle (above) that If It's Sometimes
Inappropriate, It's Always Unacceptable.

but if we apply this principle to pronouns, we're screwed: as others
have pointed out, pronouns almost always have a number of possible
referents, in sentences taken out of context.  obeying IISIIAU would
mean giving up pronouns completely, except for "I" and some reflexives
and a few other instances.

i've been talking about this subject on your own ground, just taking
the proscription as given and exploring its rationale.  it's still a
mystery to me.  but the larger point is that i totally reject the idea
that people should invent proscriptions because they'd be good things
for a language to have.  on *those* grounds, why not insist that all
irregular verbs should be regularized?  that asyntactic idioms like
"by and large" and "trip the light fantastic" should be drummed out of
the language?  that "trick or treat" and "sink or swim" should be
replaced by the (vastly more rational) "treat or trick" and "swim or
sink"?  that "information" and "mail" should be usable as count nouns?
that "himself" and "themselves" should be replaced by "hisself" and
"theirselves"?  and so on.  why not fix all this stuff?  why not
attend to things that really make trouble for language learners, that
really *need* fixing?

but...  remember that you can't reply to me by saying, "but this is
what everybody (of significance) says!"  you can't do that because
everybody uses possessives as antecedents for pronouns.

arnold (zwicky at csli.stanford.edu)