The single most crucial concept needed for me to explain to anyone what my academic specialism is all about, obviously, is the notion of
Or subscribe now to read with unlimited access for as low as $10/month.
A free account provides you access to a limited number of free articles each month, plus newsletters, job postings, salary data, and exclusive store discounts.
The single most crucial concept needed for me to explain to anyone what my academic specialism is all about, obviously, is the notion of language. And I sometimes feel a twinge of despair at the fact that the general public simply does not get that concept. Any kind of putative transmission of information, or any animal or device uttering a noise that almost sort of sounds like a word, is spoken of as language.
A paper entitled “Form and Function in Human Song,” by Samuel Mehr and Manvir Singh of Harvard, appears in Current Biology. (Danger sign! Why a biology journal, for a paper on psychology and ethnomusicology?) And The Economist (online January 25, print January 27) cannot resist discussing it in terms suggestive of language, even though the topic is reactions to song snippets. The paper is alleged to offer “evidence that music does indeed permit the communication of simple ideas between people even when they have no language in common.” It does nothing of the kind. The researchers took a large number of music performances from around the globe, established that the people who produced them classified them as either dance music, lullabies, healing songs, or love songs, and then asked a thousand volunteers worldwide to categorize them from random 14-second clips to see if they could match the creators’ reports about the intended functions.
To open the article, The Economist chooses Hans Christian Andersen’s remark that “where words fail, music speaks.” Music must speak in very muffled tones, because the subjects’ ability to classify music samples showed that healing songs turned out to be statistically indistinguishable from lullabies, and love songs could not be distinguished from either lullabies or dance music. (The latter two are of course distinguishable: The paper comments that “lullabies tend to be rhythmically and melodically simpler, slower, sung by one female person, and with low arousal relative to other forms of music.” Quite so. I don’t think we needed a biology journal to tell us that.)
The claim that information transmission was demonstrated in the music is patently ridiculous. You might just as well say that food permits the communication of simple ideas between people, given that they will (I predict with confidence) be able to classify food into broad categories like soup, steak, salad, and dessert. Even when they have no language in common.
An even worse case of perverting the notion of speaking a language appeared in the same week, and got far more coverage. Proceedings of the Royal Society B published a paper entitled “Imitation of novel conspecific and human speech sounds in the killer whale (Orcinus orca)” by José Z. Abramson, Maria Victoria Hernández-Lloreda, Lino García, Fernando Colmenares, Francisco Aboitiz, and Josep Call. It concerns the training of killer whales to imitate sounds, including the sound of human words.
“Killer whale learns to talk,” said the Daily Mail in an online headline.
“World’s first talking killer whale,” said the Daily Telegraph.
“A killer whale has been taught to talk human,” announced John Humphrys, a BBC radio news and politics broadcaster famous for his tough interviewing, his occasional grousing about “incorrect” English, and his $900K salary (soon to be partly reliquished in the wake of a gender pay-discrepancy scandal).
Again, nothing of the sort has been accomplished. Recordings of the animal trying to use its blowhole to mimic a few words can be found here. Any self-respecting parrot would be furious to hear this medley of squeals, squawks, and raspberries referred to as imitated word pronunciations.
But just suppose for a moment that an orca could be trained to imitate the sounds of isolated English words like “hello” or “bye-bye” for a fishy reward. Describing this as “talking” would still be a shocking untruth. Attempted mimicking of uncomprehended noises to win food rewards is not language!
On most academic subjects you simply cannot talk arrant nonsense or tell direct lies about simple, basic things on a BBC news magazine program and get away with it. Put out a press release asserting that cats are in fact reptiles from Venus, and you won’t get a respectful BBC news program interview, with commentary from a herpetologist and an astronomer. Absurd claims on most topics don’t make it out of the starting gate. But when the topic is supposed to be language, the loony theses gallop off down the course, cheered by thousands.
I was cheered to learn that comedians are harder to fool than science reporters. NPR’s Wait, Wait, Don’t Tell Me mocked the killer-whale story mercilessly (Mark Liberman supplies a recording on Language Log here).
