I’ve been at this year’s CogSci conference in Boston this last week. I didn’t present—I started my new job after the submission deadline, having failed to get my act together sufficiently to simultaneously apply for jobs and submit to conferences I might be able to attend if I got those jobs.
On Friday evening the big event was Chomsky. Well, the scheduled event was a Governing Board Symposium with Noam Chomsky, David Poeppel, and Elissa Newport, but that reduces to Chomsky.
He’s looking old. He shuffled, bent, onto the stage and gave a long, somewhat rambling talk about everything you’d expect Chomsky to talk about at CogSci. Then he was asked a few questions, which he didn’t really address, but to which he gave long repetitive statements that repeated things he’d said earlier.
Several things struck me in particular. First, I was reminded of how odd his conception of language is. This is nothing new, but for readers unfamiliar with his views: he thinks communication is a peripheral exapted function of language, to which it is poorly designed. Language is primarily, he thinks, a means of structuring thought, for which it is pretty much perfectly designed. Leaving aside what he might mean by the last point, I think it’s worth stressing that by “means of structuring thought” he doesn’t mean thinking in words. By language he doesn’t really mean something like English, or Mandarin. He means something more formal: a generative engine for producing infinite resources from finite means. I’m not going to go into this much here—the fact that it’s rather counter-intuitive doesn’t make it wrong (or, indeed, right). But I want to point out something that isn’t pointed out enough: Chomsky doesn’t mean by language what most people, or even (I think) what most cognitive scientists, mean by language. And what he does mean is really weird.
My second thought is that you could characterise his view rather superficially as: human beings are born with a particular set of cognitive capacities; these are then exploited for communication. At this level of detail, his views are no different from those of most cognitive linguists and other non-Chomskyan language scientists who think that language is a means of communication that adapts to, or exploits (depending on your perspective), the structure of the human brain. The big difference is which element is referred to as “language”. The non-Chomskyans (in fact, most people) use it for the means of communication. Chomsky uses it for the set of cognitive capacities (or maybe only one of them).
The third thing that struck me is that he appears to have some very odd ideas about evolution. This, again, wasn’t news. However, I was particularly surprised by one thing he said: that what he calls language most likely comes from a small rewiring in the brain of one individual. This somehow spread through the population, but not by natural selection. Asked about this, he pointed out that we’re talking about something in the head of an individual, so how could it be under natural selection? I genuinely don’t know what to make of that.