Another Newt "not"
« previous post | next post »
Once again on the Newt negation watch… In last night's Republican debate in Iowa, Gingrich defended his previous support of an individual mandate for health care insurance. He explained that he held this stance back in 1993, when he was combating so-called "Hillarycare":
I frankly was floundering, trying to find a way to make sure that people who could afford it were paying their hospital bills, while still leaving an out for libertarians to not buy insurance. (video)
From a linguistic standpoint, there wasn't anything too remarkable about this statement, though it's a nice example of a well-split infinitive ("to not buy insurance" is far more preferable here than "not to buy insurance"). It also provides an example of Gingrich's Southern-style stress shift, putting the word stress on the first syllable of insurance ("INsurance" rather than "inSURance") — just as he said "NOMinee" rather than "nomiNEE" in the clip we looked at earlier this month. (More on this initial-stress pattern in Mark Liberman's post, "Thanksgiving variation.")
But in today's Associated Press debate analysis by Charles Babington, Gingrich's quote got changed a bit:
I frankly was floundering, trying to find a way to make sure that people who could afford it were paying their hospital bills, while still leaving an out so libertarians could not buy insurance.
This mistranscription adds an extra layer of ambiguity to Gingrich's statement. It could be read as equivalent to what Gingrich actually said: he was looking for an out [OED: "a means of escape or avoidance; a way out"] that would make it possible for libertarians to not buy insurance. But this version could also be construed to mean that the out would make it impossible for libertarians to buy insurance.
Laurence R. Horn covered this type of ambiguity in "Some Aspects of Negation" (1978), giving the example, "A good Christian can not attend church (and still be saved)," which has both "not possible" and "possible not" readings. Changing can not to cannot, or contracting it to can't, forces the "not possible" interpretation. Horn observes,
The same distinction holds between could not and couldn't, although no orthographic convention permits *couldnot to disambiguate the uncontracted form as cannot does.
In speech, we usually rely on prosody to disambiguate the "not possible" and "possible not" interpretations: "so libertarians could NOT buy insurance" matches Gingrich's original "for libertarians to NOT buy insurance," while "COULD not buy insurance" would be the stress pattern associated with the "not possible" reading.
All of this ambiguity could have been avoided, of course, if the AP had correctly transcribed what was actually said. I'm sure Newt would have a thing or two to say about the reliability of mass-media political coverage.
Tom Recht said,
December 11, 2011 @ 5:20 pm
A colleague, on hearing that a mutual friend had applied for the same fellowship she had applied for, recently said to me: "I hope he doesn't get it and I don't get it."
What she meant was not "I hope that [[he doesn't get it] and [I don't get it]]", but "I hope that [not [he gets it and I don't get it]]". She was morphosyntactically negating only the first of the two coordinated clauses even though the negation applied to the entire coordination – grammatically impossible, you might think, but immediately intelligible in context.
This is a different type of ambiguity from that in the post, of course; I wonder if it's been observed before (and what a generative syntactician would make of it).
The Ridger said,
December 11, 2011 @ 5:26 pm
@Tom Recht: It took me quite a while to see how her meaning could be gotten from what you typed, but once I had, and I said the sentence out loud, the meaning was obvious. I find it fascinating that sentence intonation can not only resolve apparent ambiguities, but actually make things that are, as you say, grammatically impossible BE in fact immediately intelligible in context.
Eric P Smith said,
December 11, 2011 @ 5:33 pm
Perhaps Newt has especial difficulty because his name is so like the word for 'not' in many languages: Dutch 'niet', German 'nicht', Russian 'не' [nʲe].
Eric P Smith said,
December 11, 2011 @ 5:36 pm
@Tom Recht: surely "I hope that [not [he gets it OR I don't get it]]"? So your colleague was even more mixed up than you thought.
Tom Recht said,
December 11, 2011 @ 6:49 pm
@Eric P Smith: no, she did mean "I hope that [not [he gets it and I don't get it]]" – it was unlikely that both would get it, so his doing so would lessen her chances. Thinking about it some more, this isn't actually just a plain-vanilla coordination; that "and" contains an implied "therefore".
Eric P Smith said,
December 11, 2011 @ 7:49 pm
@Tom Recht: Yes, I’m with you now. My error was not an error of Boolean algebra, but a misunderstanding of what she meant. I had thought she meant, “I hope I get it and he doesn't get it”, but now I see that she meant “I hope that he doesn't get it at my expense.” That is typical of me. I make errors of Boolean algebra very seldom, but I am always misunderstanding what people mean, especially females!
un malpaso said,
December 11, 2011 @ 8:27 pm
I say INsurance too. I didn't even know it was Southern-style. INteresting! (I am from Atlanta, so my speech is a mix of Southernisms and standard middle American… I always thought the inSUrance/inteRESting pronounciation was pure RP British.)
However, I usually say nomiNEE, or at least put a slightly greater stress on the final syllable than the first. Otherwise, it rhymes with "hominy," another Southern favorite.
Ran Ari-Gur said,
December 11, 2011 @ 9:06 pm
@Tom Recht: Neal Whitman has written about the "wide-scoping operator" phenomenon (such as your colleague's "-n't" that somehow negates two whole coordinated clauses) a few times at his excellent blog, Literal-Minded; he has a whole category dedicated to it, comprising dozens of examples he's come across.
Jerry Friedman said,
December 11, 2011 @ 10:56 pm
@un malpaso: "INsurance" is heard a lot more places than the traditional South, I believe, but "inSURance" is the only pronunciation in the AHD 4th edition. It gives "INteresting" as the first pronunciation and "interESTing" as the third. I think that if you listen to, say, NPR, you'll mostly hear the announcers say "inSURance" and "INteresting".
To my limited knowledge, COMponent is another one of these, while POlice and UMbrella are more restricted.
Jerry Friedman said,
December 11, 2011 @ 11:07 pm
Same pronunciations in AHD5. I didn't know its definitions and pronunciations were on line. Incidentally, I meant "INt'resting" for the first pronunciation.
Bob Violence said,
December 11, 2011 @ 11:57 pm
From Marnie (1964):
"You're not from Los Angeles, Miss Edgar. 'InSURance' is only pronounced 'INsurance' in the South."
This came as a surprise to me on my first viewing since I never recalled hearing the "Southern" variant.
Chris said,
December 12, 2011 @ 3:57 am
"to not buy insurance" is far more preferable here than "not to buy insurance"
Interesting to see these Language Log prescriptivists lecturing those of us who prefer not to split infinitives (or should I have said "to not split"?).
[(myl) You clearly haven't been paying attention.]
David Bloom said,
December 12, 2011 @ 8:48 am
@The Ridger: Intonational elements are morphemes! We won't have an adequate grammatical description until we figure out how to make it explicit. Love how your last sentence illustrates the point with the capped "BE", without which the sentence would mean not something different but just a little less.
Katherine said,
December 12, 2011 @ 8:51 am
"InterESTing" British RP? Never. Intr'sting, sometimes; INteresting. Not InterESTng.
Never heard COMponent in 54 years on this planet, either (aye-ther? EE-thur?) even in Texas and deepest darkest West Virginia – and certainly never in any region of the UK (26 years and counting).
Confused,
Tunbridge Wells
Theophylact said,
December 12, 2011 @ 9:58 am
As Tom Lehrer sang,
Trimegistus said,
December 12, 2011 @ 10:45 am
It's kind of predictable how the language use of this week's leading Republican contender immediately comes under attack. Like clockwork.
[(myl) The only clockwork-like activity here is your reflexive partisan misunderstanding. As J.W. Brewer observes below, "what [came] under attack in this post was the Liberal Media's misreporting of the candidate's actual (and perfectly acceptable except to anti-splitter prescriptivists) language usage".
Looking back over recent discussions here of leading Republican figures, in fact, I see that most of them are more like defenses than like attacks (e.g. here and here and here). Please wake up and try looking at the world as it is, rather than as your prejudices lead you to expect it to be.]
ENKI-][ said,
December 12, 2011 @ 11:18 am
It's particularly amusing since libertarians are the one group who would be angry about a legislation forcing their hand, regardless of whether it forced them to do something or it forced them not to do the same.
J. W. Brewer said,
December 12, 2011 @ 11:34 am
Wait, I thought what had come under attack in this post was the Liberal Media's misreporting of the candidate's actual (and perfectly acceptable except to anti-splitter prescriptivists) language usage?
Barbara Partee said,
December 12, 2011 @ 2:24 pm
That special kind of negated coordinate sentence has been studied — it's really interesting. It includes puzzling examples like how to express "CAN'T [you go to the movies and I stay home with the kids]", which often comes out as "You can't go to the movies and me stay home with the kids", with accusative "me" as subject of the second sentence. The first paper devoted to that construction that I know of is Siegel, Muffy E.A. 1987. Compositionality, Case, and the Scope of Auxiliaries. Linguistics and Philosophy 10:53-76.
I don't know if there's more recent work on the topic.
rukymoss said,
December 12, 2011 @ 2:52 pm
In west-central Wisconsin, I frequently hear "INsurance", "UMbrella", but it seems to be limited to those people with a small-town or rural origin, with education usually HS or a year or two of tech school. As for British pronunciations, I can usually figure them out, but "intesTINE-al" and "u-RI-nary" had me puzzled till context made them clear ( these pronunciations were from an English physician teaching physiology to pharmacy students in Pittsburgh.)
Neal Whitman said,
December 12, 2011 @ 10:37 pm
@Ran, @Barbarak @Tom Recht:
Thanks for the link, Ran. Barbara: Yes, in addition to the blog posts that Ran linked to, I presented a paper on this topic at LSA 2010, citing Siegel's work and that of Kyle Johnson, and presenting a categorial grammar analysis of their sentences as well as utterances such as "Do you want to do the dishes and I'll feed the cats?"
marc said,
December 13, 2011 @ 11:21 am
The Libertarian viewpoint is the less coercion the better. Legislation designed to allow them to not buy insurance would therefore be a good thing in their eyes. There's no great paradox here, just common sense.
George said,
December 14, 2011 @ 2:33 am
@ rukymoss
Your English physician notwithstanding, "intesTINE-al" and "u-RI-nary" are certainly not widespread (to say the least) British pronunciations. If anything, I'd be a great deal less surprised (but still somewhat surprised) to hear them coming from a North American.
Michael Briggs said,
December 18, 2011 @ 11:03 pm
My ears tingle when I hear sports announcers say OF-fense and OF-fensive, DE-fense and DE-fensive. And I thought about complaining to a flight attendant recently when the captain announced that we were commencing our DE-scent into Portland.