handbook weirdness (was Re: PSAT Glitch)

[Arnold Zwicky]

i continue to scan grammar/usage handbooks, and i continue
to find weird stuff.  i was astonished to see that jacques
barzun's version of the Possessive Antecedent Proscription
in Simple & Direct applies only to *proper names*: pp. 95-6 in
the 4th ed. (2001)...

  The proper linking of pronouns with anteceden nouns includes
  one commandment that may seem superfluous because it is
  difficult to remember: there can be no logical link between
  a proper name in the possessive case and a personal pronoun:
  "Wellington's victory at Waterloo made him the greatest name
  in Europe" is all askew, because there is in fact no person
  named for the _him_ to refer to.  _Wellington's_ is not a
  noun but an adjective; it corresponds to "the _Wellingtonian_
  (victory)" and the only subject word is that same _victory_,
  with which _him_ obviously doesn't go.  Since many sentences
  follow this pattern of possessive + noun + pronoun, watch
  the connections and take care either to free the person from
  the possessive or to change the pronoun and make it refer to
  what is actually named.

meanwhile, an e-mail correspondent suggested a college text that he
found useful.  here, very slightly edited, is what i wrote to this
correspondent after i checked out the handbook:

----------------
i've just checked out the 4th edition (2003) of martha kolln's
Rhetorical Grammar (which, in general, looks to me like a good book),
and it turns out to have the most dogmatic proscription of any i've
read - because *all* of her examples are of the type that native
speakers find unexceptional and that careful professional writers use
all the time.  that is, instead of giving an overkill "rule" based on
some genuinely clunky examples, well, ones that are clunky out of
context, she goes for the generally impeccable examples of the "Mary's
father adores all of his children, but she is the apple of his eye"
type.

(so: no examples that are potentially ambiguous out of context; no
examples of possessive antecedents inside adverbials; not even any
examples of relative clauses extraposed from possessives, which are
things i'd certainly avoid in careful writing.)

the section is entitled "The Vague Antecedent" (p. 262).  it starts
with examples of good pronoun use.  then...

"As you can clearly see, the pronoun stands in for the entire noun
phrase, not just for the headword--and certainly not for a modifier of
the headword.  Now look at the following sentences with that principle
in mind:

  The neighbor's front porch is covered with trash, but _he_
    refuses to clean it up.
  The neighbor's dog gets into my garbage every week, but _he_
    refuses to do anything about it.
  My sister's boyfriend works for a meat-packing plant.  _She's_
    a vegetarian.
  It's hard to keep track of the Administration's stand on
    immigration.  _They_ say something different every week.
  Last summer I didn't get to a single baseball game, even though
    _it's_ my favorite sport.

Notice what has happened.  The subject of the second clause in each
case is a pronoun.  But its antecedent is not a complete noun phrase;
it's only a noun modifier.  The problem is not with communication:
The reader will understand these sentences.  And in a conversation we
might not even notice anything amiss.  But there is a problem of
fuzziness that could easily cause a blip in the reader's
comprehension.  As writers we have the obligation to consider the
reader's expectations, to get rid of the fuzziness caused by vague
antecedents."

[end of quotation]
-----------
[amz to correspondent]

oh dear.  the antecedents in every case but the last clearly *are*
complete noun phrases - noun phrases used as determiners, which is
what possessives are.  not only are they noun phrases, but they are
clearly *referential* noun phrases.  there is absolutely nothing
wrong with them structurally.  (the vegetarian sister example isn't
great out of context, because the first sentence is pretty clearly
*about* the boyfriend, not the sister.  but that can be jiggled by
manipulating the context: "At least one person in my family has a
conflict between belief and emotional attachment.  My sister's
boyfriend works for a meat-packing plant, yet she's a vegetarian.")

the final example is a N+N compound case, but "baseball" is a mass
noun, hence can be referential as it stands.  i have no problem with
this example, in fact.

i'm afraid that kolln is Blinded By The Rules.  the subtitle of the
book ("Grammatical Choices, Rhetorical Effects") suggests that the
author would be especially sensitive to information structure and to
the context of sentences, but the discussion of possessive antecedents
is totally deaf to these considerations.

[end of e-mail message to correspondent]

arnold (zwicky at csli.stanford.edu)