July 08, 2004

Caring less with stress

I thought I had already broken the ultimate taboo, but it turns out that there were some depths yet to be explored. Believe it or not, I'm about to defend Richard Lederer against Steven Pinker.

In a recent Language Log post, Eric Bakovic gives Lederer a hard time for asserting that people who say I could care less are being illogical.

Eric's first argument is that since people use this form just about as often as I couldn't care less, Norma Loquendi has spoken, and we need to listen. I agree with Eric up to this point. Eric offers some Google counts in support: the two phrases are about equally frequent. In an earlier posting here, Chris Potts listed I could care less as an example of the "handful of English constructions in which, quite surprisingly, one can add or remove a negation without change of meaning", and he took it for granted that this is just a usage to explain, not an error in logic.

Speaking for myself, I don't take this to be a point of principle, that whatever is said or written reasonably often must ipso facto be part of the language.. Especially where negation is involved, there are plenty of common mistakes, like fail to miss used to mean "miss", or no X is too Y to ignore used to mean "no X is so Y that one should ignore it." It can be a complicated question, conceptually and practically, to decide whether such cases are constructions, fixed expressions, idioms or whatever, as opposed to natural mistakes that people often make in using a psychologically difficult combination of elements and structures. This is partly a question of linguistic analysis, and partly a question of psychological interpretation, and partly a question of social norms. In any case, the categories of "mistake" and "English expression" are obviously overlapping, psychologically, historically and socially.

However, I don't think there's much question at all about could care less, which has clearly become a well-accepted colloquial expression in contemporary American English. This conclusion can claim the sanction of the OED, which gives sense 4 of care as

4. In negative and conditional construction: a. not to care passes from the notion of ‘not to trouble oneself’, to those of ‘not to mind, not to regard or pay any deference or attention, to pay no respect, be indifferent’.

and then among the various subtypes listed (e.g. care a button or a fig) comes eventually to the specific phrase in question,

(c) Colloq. phr. (I, etc.) couldn't care less: (I am, etc.) completely uninterested, utterly indifferent; freq. as phr. used attrib. Hence couldn't-care-less-ness.

for which the earliest citation is from 1946, and then gives an explicit listing to the unnegated form:

(d) U.S. colloq. phr. (I, etc.) could care less = sense (c) above, with omission of negative.

1966 Seattle Post-Intelligencer 1 Nov. 21/2 My husband is a lethargic, indecisive guy who drifts along from day to day. If a bill doesn't get paid he could care less.
1973 Washington Post 5 Jan. B1/1 A few crusty-souled Republican senators who could care less about symbolic rewards.
1978 J. CARROLL Mortal Friends III. iii. 281 ‘I hate sneaking past your servants in the morning.’ ‘They know, anyway. They could care less. Thornton mistreats them horribly.’

with a first citation a mere 20 years later. The OED does tell us that (I, etc.) could care less is a "colloq. phr." -- but so is (I, etc.) couldn't care less. The only difference is that (I, etc.) could care less is a "U.S. colloq. phr."

This is not much of a difference for Lederer to hang his hat on, but then English-language prescriptivism has always had a solid admixture of anti-Americanism.

OK, so far, so good. Eric, Chris, Google, the OED and I are all in agreement.

But Lederer is not off Eric's hook yet. Eric points out that Steven Pinker has written about could care less, both in The Language Instinct and in a 1994 New Republic article. The crucial Pinkerian passage is this:

A tin ear for stress and melody, and an obliviousness to the principles of discourse and rhetoric, are important tools of the trade for the language maven. Consider an alleged atrocity committed by today's youth: the expression I could care less. The teenagers are trying to express disdain, the adults note, in which case they should be saying I couldn't care less. If they could care less than they do, that means that they really do care, the opposite of what they are trying to say. But if these dudes would stop ragging on teenagers and scope out the construction, they would see that their argument is bogus. Listen to how the two versions are pronounced:

could ESS.

The melodies and stresses are completely different, and for a good reason. The second version is not illogical, it's sarcastic.

(By the way, could care less is hardly used only by "today's youth" -- the authors of the OED's 1966 and 1973 citations are presumably old and gray by now, if they are even still alive.)

As Eric says,

...this hypothesis has the added advantage of not insulting the intelligence of the half of the population that uses the allegedly incorrect form. (The only thing missing is independent evidence that the same intonational distinction holds of other sarcastic-nonsarcastic utterance pairs; Pinker also does not cite any sources for this claim, unlike many other claims made and discussed in the book.)

But unfortunately, that's not the only thing missing. Pinker doesn't provide any evidence that the claimed difference in stress and/or pitch is actually used to distinguish these phrases, or that it would have the asserted effect on interpretation if it did. And unfortunately for this otherwise neat hypothesis, I'm fairly confident that (a) the two phrases are not generally distinguished prosodically as Pinker asserts they are; and that (b) the cited prosodic difference would not as a general rule yield the asserted (sarcastic vs. non-sarcastic) difference in interpretation.

I promise to examine the examples in some conversational speech corpora to evaluate (a), and show you all the pitch tracks. [Update -- see here.] And I'll say more later about (b), and how to evaluate claims like this. For now you'll just have to take my word for it , or rather, take note that I disagree with Pinker's analysis. But I've put in some time working on the analysis and synthesis of English intonation, and I'm fairly confident that Pinker is stretching a bit here, as Tom Sawyer might have said.

Whatever the origin of I could care less -- and it's as likely to have to been confusion about negation as sarcasm -- by now, it's just an expression. And as Eric hints, grammar anti-mavens may also sometimes try to make us believe something false just by asserting it.

It wouldn't surprise me to learn that Lederer has never bothered to read Pinker's account of the alleged intonational disambiguous of could care less. But if he had read it, the kindest thing might have been just not to mention it, as indeed he didn't.

[More on this:

"'Could care less' occurs more" (7/13/2004)
"Negation by association" (7/13/2004)
"Speaking sarcastically" (7/13/2004)
"(Auto)biography of a blog thread" (7/16/2004)
"Most of the people in the world could care less (7/16/2004)
"Caring less all the time: A variant of the etymological fallacy, and some cautions about the pragmatics-phonetics connection (7/24/2004)
"The future of history of usage (4/16/2005)
"The care less train has left the station (6/20/2005)
"Caring more or less (6/29/2005)


Posted by Mark Liberman at July 8, 2004 10:25 PM