December 16, 2006

Intonation contours and polonium poisoning

As Language Log has repeatedly pointed out, journalists simply don't have the vocabulary to talk about linguistic phenomena, but they just go ahead and write about it anyway, ignoring the need for the requisite concepts and terminology. The New Yorker this week (December 18, 2006; 60-69) has a feature article about an Australian social scientist, David Kilcullen, who has studied counterinsurgency warfare. He speaks, as many younger people do these days, in a way that often uses rising intonational contour on declarative clauses, a habit that often suggests to people that the speaker is seeking confirmation that the hearer has understood what he is saying. The writer of the article, George Packer, wants to describe this style of speech (which has been discussed carefully and critically a number of times on Language Log under the heading uptalk), but simply doesn't have the (very simple) technical vocabulary; so what we get is this claim about Kilcullen:

He has a talent for making everything sound like common sense by turning disturbing explanations into brisk, cheerful questions: "America is very, very good at big, short conventional wars? It's not very good at small, long wars? But it's even worse at big, long wars? And that's what we've got."

But Kilcullen is not using questions here. "Question" is a semantic term. Questions are defined by the property of having corresponding sets of statements expressing their sets of possible answers. For Are you cold? there are two statements in the answer set (they could be expressed as Yes, I'm cold and No, I'm not cold; don't confuse answers with responses, by the way: your response might be "As if you cared!", but that would not be an answer). For Who was your favorite Beatle? the answer set includes at least four fully appropriate statements (more when you count silly answers; it is no part of the semantics of questions to rule out silly or misguided answers, like "Eric Clapton", even though they cannot be correct). Notice, by the way, that questions do not all have rising intonation: many of them have falling intonation (Who was your favorite Beatle? normally would).

The sentences with question marks on the end in the above quote are not questions at all. They are statements. The question marks on the ends are supposed to indicate to us (by a reasonable compromise convention) that Kilcullen made them using the rising "uptalk" intonation. It is well known to be common among Australian speakers. But that doesn't mean he makes things "sound like common sense by turning ... explanations into ... questions." Is America good at big, short conventional wars? would be a question, and it would not sound like an explanation of anything.

Is it so hard to grasp this? I don't think so. It seems less complex than basic polonium chemistry, for example. Yet it seems to me that if George Packer were writing about poisoning by polonium-210, he would have headed for a reference source. He wouldn't just have blundered about. Nor would he have used 19th-century terms (as people so often do with language), talking about Litvinenko having been fed a noxious earth, or a poisonous herb. He would have found out what polonium is and what is known about it and what the right terminology is: it's an element, a metal; it's extremely rare in the terrestrial environment; the atomic number is 84; it's radioactive, and it decays by alpha particle emission; it's 250 billion times as poisonous (weight for weight) as hydrocyanic acid; and so on. He would have checked this because it is important and chemists will call you on it and you don't want to get stuff wrong.

But when the topic is language, nobody looks anything up or calls anyone. Calling a declarative clause with rising intonation a "question" is like calling polonium "thallium". Yet Packer feels he doesn't need to spend an extra minute or two to get the vocabulary right. When you're talking about language, it seems, you don't need to do that. No one will challenge anything. Except perhaps on Language Log, and that's just the bloggers in their pajamas.

Posted by Geoffrey K. Pullum at December 16, 2006 12:03 AM