January 28, 2006

The parts of speech

Luckily, the late Peter Ladefoged had a really good sense of humor, and as he sits in the faculty club at the University of Heaven and reads the AP story about his death, I'm sure the club booms with his rich laughter. The first sentence of what the Associated Press put on the wires (which appears in, for example, the San Jose Mercury News today) says:

LOS ANGELES - Peter Ladefoged, a UCLA linguistics professor emeritus who made it his life's work to record the parts of speech used in human languages, has died.

But (although it is easy to see how this howler might arise) the parts of speech are not in fact the parts of your body that you use for speech. More journalistic ignorance of even the most absolutely basic notions of linguistic science. Sigh.

"Parts of speech" is an old-fashioned name for lexical categories — classes to which words with similar grammatical properties belong, e.g., noun, verb, adjective, adverb, preposition. Categories of this sort are in fact the key element that lifts grammatical description up to a level of abstraction where you are not talking about speech, you are talking about higher-level units to which various grammatically equivalent small stretches of speech can be treated as belonging. The job of a phonetician is to describe in minute detail the speech sounds themselves; so classifying into parts of speech (lexical categories) is exactly what the phonetician must never do, in his capacity as phonetician. And if you will pardon my being a little irritable (I'm sorry, but a friend of mine recently died), I do think we have a right to expect better from the AP than this. What they have done is like writing up Einstein's demise as the passing of a chemist. It is rank ignorance. Journalists just don't know anything about the language sciences, but instead of asking they just write nonsense. Peter deserved better than to have his passing commemorated with an embarrassing goof that any of the students that he taught in UCLA's excellent Department of Linguistics could have fact-checked.

Posted by Geoffrey K. Pullum at January 28, 2006 05:02 PM