October 18, 2007

Neanderthals may have had headline writing gene

Much of the blame for the public's poor understanding of science must go to a little studied but culturally pivotal genre: news report headlines. Short snappy headlines provide the lazy reader with just enough information to totally misconstrue a story.

Take the evidence reported in today's NYT that Neanderthals may have possessed a FoxP2 gene like that of modern humans. It's a striking result, if it holds up, though it might not. But, as is clear from the article, presence of a modern human FoxP2 gene doesn't establish one way or another whether Neandertals had modern human speech abilities. The thing is, the nature of the FoxP2 gene establishes neither sufficient nor necessary conditions for speech: see the Language Log coverage of FoxP2 in bats and chimps.

The NYT article, reporting on a complex set of controversial hypotheses and results, doesn't do too bad a job. The headline, however, "Neanderthals May Have Had Gene for Speech", is potentially misleading. Even Nature's coverage, titled "Modern speech gene found in Neanderthals" would tend to confuse those who didn't already know something about genetics.

The very fact that the headline writers use the bare singulars "gene for speech" and "speech gene" suggests, through omission, that this is not just one of many genes linked to speech, but possibly *the* speech gene. But the more important fact (as the LL pieces make abundantly clear) is that FoxP2 doesn't deserve to be called a speech gene at all. At best, it seems to be a vocal articulation gene, although it may well have other as yet unidentified roles. And vocal articulation, even very precisely managed vocal articulation, is not speech.

To speak you need to have language. Otherwise all your twitterings, squawks, grunts and growls are just twitterings, squawks, grunts and growls. FoxP2 has nothing to do with language per se. For example, it is presumably totally irrelevant to speakers of human sign languages. At best, FoxP2 may have been an important factor in the evolution of language... but I feel safe in stating that there must have been many other such factors, both social and genetic.

I understand that "Neanderthal's may have had vocal articulation gene" is not such a sexy headline. But please, please can we all stop labeling FoxP2 a/the "speech gene"?

Posted by David Beaver at October 18, 2007 03:06 PM