sources (re: PAP)

[Arnold Zwicky]

an exchange from way back, to which i'm finally replying.

Date:         Thu, 12 Jun 2003 20:14:37 -0400
From: "Baker, John" <JMB at STRADLEY.COM>
Subject:      Re: open letter to Andrea Lunsford 1
To: ADS-L at LISTSERV.UGA.EDU

PB>Arnold, I really like the open letter, and I appreciate your
  >posting it here.  Am I missing something with your initial
  >examples, though?  Both of them are at least potentially ambiguous,
  >which only provides fodder for the defenders of the PAP.  Detention
  >centers as well as detainees can be difficult to locate, and while
  >I believe that Carnegie Hall did not move or change its name in
  >1962, it could in principle have happened.  I note also that the
  >Carnegie Hall example does not violate the PAP, since the
  >antecedent of "it" is not "Philharmonic" but "home."  Actually, I
  >think these examples just support my point that the New York Times
  >is not a particularly good example of elite writing.

PB>With respect to your more substantive points: How would your
  >argument apply to, say, the double negative?  Once part of standard
  >English, it has been driven out of the language by a false
  >application of mathematical logic.  "I don't need no stinking
  >badge" is a stronger statement than its alternatives.  Still, there
  >is no question in my mind that the double negative is verboten in
  >formal written English.

this in response to a three-part message beginning with:

AZ>From the editorial page of the 6/3/03 New York Times...
  >In the first (of two) editorials, "The Abusive Detentions of
  > Sept. 11":
  >     Detention centers routinely blocked efforts by detainees'
  >     families and lawyers to locate them.
  >In the second, "A New Twist at Lincoln Center":
  >     After all, Carnegie Hall was the Philharmonic's home from
  >     1891 to 1962, when it moved to what was then called
  >     Philharmonic Hall at Lincoln Center.

---------------------------

in more or less reverse order:

multiple negation, or "negative concord", did indeed used to be
part of standard english, but now survives - though with tremendous
tenacity - only in nonstandard varieties, all over the world.  i am
very much not an expert on the historical syntax of english, but i
don't believe that multiple negation was driven out by "false
application of mathematical logic"; as i recall, the change in the
standard language was well underway before commentators began
justifying single negation, and deploring multiple negation, on the
basis of "logic".

actually, i don't know of any clear case where language change was in
fact driven by the pronouncements of usageists.  these pronouncements
do sometimes accord with changes, but after the fact: the usageists
defend variants that are already associated with the standard variety
and denounce variants that are already associated with nonstandard
varieties.  for the most part, the usageists are supporting the status
quo.  in the few cases where usageist pronouncements were essentially
made up out of whole cloth - split infinitives, stranded prepositions
- these pronouncements have never, so far as i know, succeeded in
altering the standard variety, only in making large numbers of writers
and speakers self-conscious and unsure of themselves.

as for what i'd say about multiple negation, it's pretty
straightforward.  just as some vocabulary items are associated with
nonstandard or informal varieties (and some with standard formal
written varieties), in essentially an arbitrary way, so some syntactic
constructions are so associated.  if some standard formal written
construction C isn't part of your native variety/varieties and you
propose to write in the standard formal variety, then you'll have to
learn to use C in formal writing instead of its counterparts in your
everyday variety, or accept the possibility that some people will
judge you badly, or flaunt your everyday variety in the hope of
gaining more acceptance for it.  (of course, your linguistic choices
are just part of a much bigger package that gives rise to such
judgments about the kind of person that you are.)

-------------

back to the PAP.  i'm grateful for the comments and criticisms i've
gotten.  somewhat defensively, i add that i see my ADS postings as a
kind of conversation, and they get relatively little editing (at most,
two or three cycles through, while my writing for publication
frequently goes through twenty to thirty).

i began the letter to andrea lunsford on june 3, and thought it would
be entertaining to use the two NYT editorials of that day for my
examples (though, goodness knows, i have plenty of others).  i was
then derailed by life events and finished the letter days later.  i
see now that the second example was not a good choice; it does have
two possible interpretations, and it's not especially important
which interpretation was intended.  (the interpretation i got on
first reading involved a violation of the PAP, so i just went with
it.  but it wasn't a particularly good example for my purposes.)

the first example - "Detention centers routinely blocked efforts by
detainees' families and lawyers to locate them." - i'd like to stick
with, though, despite its in-principle six-way (!) ambiguity in the
antecedent of "them": "detainees' lawyers", "detainees' families",
"detainees' families and lawyers", "detainees", "efforts", and
"detention centers".  even without context, it seems to me that there
is only one plausible antecedent, "detainees" (so that this is a PAP
violation), and i maintain that normal (non-linguist) readers would
select this interpretation without any problem.  when the context
of the sentence is added in - the editorial is about the detainees and
their treatment - there should be no doubt about the matter.

that's my claim about this particular example, and about many others
involving pronouns with possessive antecedents.  but what are the
facts here?  just how much potential ambiguity is there in pronoun
use, and how much of it affects the smooth course of reading?  how do
skilled (and well-intentioned) readers interpret (third person)
pronouns? how do skilled writers use them?

i'll set aside deictic uses of personal pronouns, and cataphora as
well.  i'll also set aside anaphoric linkages that are ruled out by
clashes in gender or number, or by structural considerations (like the
obligatoriness of reflexives in certain configurations, the phenomenon
that means that "her" in "Mary's mother admires her" cannot have
"Mary's mother" as its antecedent).

even so, a huge number of occurrences of (third person nonreflexive)
personal pronouns are potentially ambiguous, in the sense that there's
more than one structurally available antecedent for them in the
preceding linguistic context.  this has nothing to do with the
possibility of a possessive antecedent; it's the nature of pronouns.
aspects of the linguistic context, having to do with
topicality/salience, and background knowledge, and the reader's
estimate of the writer's goals normally select one antecedent as the
most likely one - and this probably happens without the reader ever
even entertaining other possible antecedents.

here are a couple of pretty much randomly selected examples from
calvin trillin's book of humorous essays, Too Soon to Tell (NY: Farrar
Straus Giroux, 1995):

p. 69: "No, I don't think he's much like the late King Farouk of
Egypt, although if he doesn't start going light on the pasta he may
get there one of these days."

it takes some work to realize that the second and third occurrences of
"he" are each potentially two-ways ambiguous, having either the first
"he" or the (nearer) "King Farouk of Egypt" as antecedent (not to
mention some other NP in preceding context); so there are, in
principle, four readings for this sentence with respect to the
antecedents of pronouns.  but "King Farouk" is just out of it: the
first "he" names the individual (trillin himself, as it turns out)
that this bit of discourse is about; and the fact that the adjective
"late" means that King Farouk is dead pretty much rules out the
possibility that we're talking about the king as someone who will
change his diet (in the future) and/or possibly get fat (in the
future).  *potential* ambiguities, yes, but not ones that any sensible
reader would entertain.

p. 87: "The other day I managed to get a glimpse of the future and the
past at the same time: I saw an item reporting that Barry Goldwater
had gone back to being publicly pro-choice and another item reporting
that the most recently transcribed tapes from Richard Nixon's Oval
Office reveal him once more as a vindictive, unscrupulous paranoid."

(this is the very beginning of the essay.)  what's the antecedent for
"him"?  there are, in fact, two possibilities, "Richard Nixon" (though
this would constitute a PAP violation) and "Barry Goldwater" (though
this NP is more distant).  suppose you knew nothing of modern american
political history; then both possibilities would be live options,
depending on whether you take this passage to be about barry goldwater
or to be about two parallel situations, one involving barry goldwater,
the other richard nixon.  (and if you truly love the PAP, you'll go
with goldwater all the way.) but trillin assumes that his readers know
this history, so they'll never entertain the possibility that
goldwater appears on the tapes from nixon's office and reveals himself
there to be a vindictive, unscrupulous paranoid.

i claim that the original detainees sentence is just as clear as these
examples from trillin (and endless numbers of other examples i could
have supplied).

but why are we looking at examples to see what might be *wrong* with
them, what might be erroneous or infelicitous (due to ambiguity)?  why
this fault-finding stance?

this is the stance of the usageists.

"prescriptive grammar" is a misnomer, and i'm reluctant to use the
label.  the basic impulse is to root out error and infelicity, to find
fault, to deplore, deprecate, and disapprove (often in a spirit of
genuine helpfulness, i don't deny).  you collect the mistakes, sort
them into types, and issue proscriptions, which might be stated in
overtly negative terms (don't end a sentence with a preposition) or in
overtly positive terms (the "and" introducing the last member in a
series must be preceded by a comma), but are in either case designed
to pick out bits of language that have somehow gone wrong.  the
*prescription* consists in recommending replacements for these.

usage manuals are, for the most part, giant appendices probi - and,
like the appendices probi of yore, they are valuable testaments to
what lots and lots of people - even *most* people* - are doing.

geoff nunberg has cited william cobbett, who proposed to give advice
to those who would better themselves, by showing them what to avoid.
cobbett provided bad examples in high places, and it's worth asking
why he did this.  (several different instincts could lead to this sort
of elite-baiting.)  [i hope you'll now see that this discussion is
relevant to john baker's worries about the NYT as a source of
examples.]  my understanding - which could well be wrong, i haven't
read cobbett in twenty years - is that he saw standards as lying
outside social privilege, with failures to meet those standards as
distributed across the spectrum of social rank.  so his critiques
of usage were egalitarian, targeting government ministers and plowboys
(or ploughboys) equally.

h.w. fowler, the 20th-century usageist hero of english, similarly
began by providing advice.  here's his letter of 19 december 1904 to
an editor at the clarendon press ("we" here is hwf plus his brother
frank, francis arthur fowler):

  There is a vast number of writers nowadays who have something
  to say & know how to make it lively or picturesque, but being
  uneducated cannot write a page with a blunder or cacaphony or
  piece of verbiage or false pathos or clumsiness or avoidable
  dulness [hwf's spelling].  The book we are thinking of would
  consist chiefly of classified examples in terrorem, with a few
  rules on common solecisms.  The examples would not be artificial,
  but taken partly from the standard authors, & partly from papers
  like the Times and the Spectator.  It might possibly, we think,
  be mildly entertaining as well as serviceable.

this is sterner stuff than cobbett, and distinctly patrician.
fowler's criticism of the elite writers is, essentially, that they
should know better, they should provide a better model for the
uneducated, so their shame is greater than that of the merely
uneducated.  [but i might have misread him.  i'm not a social or
intellectual historian.]  in any case, the fowlers' approach to
usage set off a giant scavenger hunt for error in high places;
people mailed in wonderfully awful examples, some of which appeared
in the next fowler book or edition. [the process is well documented
in burchfield's chapter on the fowlers in his Unlocking the English
Language, and of course in morris's biography of hwf, The Warden
of English.]

the legacy of these traditions, whether they are egalitarian or
patrician, is that usageists - and, by extension, people in general -
are still searching for error, rather than trying to figure out how
writers, readers, speakers, and hearers manage their (mostly
successful) interchanges.  john baker earnestly wonders whether the
NYT is an appropriate source of examples, just because he can find
potential ambiguities in these examples.  on this criterion, no source
would do; jane austen, calvin trillin, john mcphee, anyone you can
name can be impugned.  yes, everyone slips every so often.  but most
of these writers do very well almost all of the time, and we should
take them seriously.