April 26, 2006

The race to the bottom in science reporting

According to a recent LL post by Ben Zimmer, Lance Nathan feels that the recent "English Language Hits 1 Billion Words" headlines represent "new lows in linguistic reporting". Much as I respect Lance, I have to disagree. Having hit bottom, the folks at the Associated Press broke out the heavy excavation equipment and kept right on digging to come up with this:

While many animals can roar, sing, grunt or otherwise make noise, linguists have contended for years that the key to distinguishing language skills goes back to our elementary school teachers and basic grammar.

Sentences that contain an explanatory clause are something that humans can recognize, but not animals, researchers figured.

Two years ago, a top research team tried to get tamarin monkeys to recognize such phrasing, but they failed. The results were seen as upholding famed linguist Noam Chomsky's theory that "recursive grammar" is uniquely human and key to the facility to acquire language.

But after training, nine out of Gentner's 11 songbirds picked out the bird song with inserted warbling or rattling bird phrases about 90 percent of the time. Two continued to flunk grammar.

More later on the science behind those immortal phrases. (And really, I'm being somewhat unfair to the AP's science writers, as you'll see when we go over the press releases they were working from.)

[There's an excellent summary of the experiments, and some discussion of what they might mean, over at TstT. And some interesting discussion by Chris at Mixing Memory.]

[OK, you can read my promised (serious and far too long) commentary here (but it has music!). David Beaver has a shorter comment, from a slightly different perspective, here.

Why did the quoted AP story bother me so much? My main objection is the notion that this experiment has anything to do with "sentences that contain an explanatory clause". Even if we take Gentner's interpretation at face value, this asserts a kind of semantic content that the experimental materials clearly lacked. Secondarily, it's not a helpful or accurate account of what happened in the experiment to say that the starlings "picked out the bird song with inserted warbling or rattling bird phrases".

It's fair to respond that if I'm going to complain about the AP story, I ought to give an example of how to write about this work in a newswire-story style and at newswire-story length. I certainly haven't done this, since my discussion of the paper is way too long, way too technical and way too skeptical for a news story. Well, as an old colleague used to say, if I only had more time I could write less -- and perhaps write more simply and more charitably, too.]

Posted by Mark Liberman at April 26, 2006 10:02 PM