In a recent Guardian article Harry Ritchie asks the following question: ‘Why do we persist in thinking that Standard English is right, when it is spoken by only 15% of the population?’ It’s an interesting question. My observation is that, actually, fewer and fewer people do in fact persist in thinking this. But I’m not blogging in order to discuss the question. Rather, I’d like to draw attention to Ritchie’s truly mind-boggling answer to the question he poses: he blames Noam Chomsky.
I’ve read the article repeatedly now, and I have to say I’m still no closer to finding any real justification for Ritchie’s claim. Still, in the interests of cool-headed analysis as opposed to the vitriolic dismissal my emotions are screaming at me to write, let’s examine his case. As far as I can see, it rests on two related claims. Let’s examine them in turn:
Firstly, Chomsky (and his disciples, including Pinker) have been so convinced that language is partly innately-specified that the whole discipline has been (in Ritchie’s words) ‘hunting unicorns’ and ignoring matters such as language death, social and political factors in language use and environmental factors in language acquisition. Secondly (and I quote again), ‘Recent evidence from neurology, genetics and linguistics all points to there being no innate programming. Children learn language just as they learn other skills, by experience’.
Turning to the first claim, we can only take it seriously if it can be shown that Chomsky’s work has somehow prevented others from exploring language death, social and political factors in language use and environmental factors.[1] Has Ritchie ever visited a library devoted to language and linguistics? Over the past twenty years, there have been billions of words devoted to these areas of study. Does Ritchie really believe that if Chomsky hadn’t come along, there would have been more? Actually, I believe there would have been less. Without Chomsky there would be fewer, not more, departments of Linguistics.
Turning to the second, it is simply not true that all the evidence points against Chomsky’s claims of innateness. There is certainly some evidence from work on neural-nets in Connectionist frameworks that ‘minds’ can learn aspects of language from experience. But these rely on the highly controversial claim that neural-nets somehow model cognition. Equally, they conveniently ignore many of the linguistic data Chomsky had in mind that are patently unlearnable. Besides which, there is at least as much evidence in support of innateness claims as there is against it. (For those interested in Chomsky’s views, rather than take them from Ritchie I’d recommend the highly accessible introduction in ‘Problems of Knowledge: the Managua Lectures’, published, I think, in 1978.)
Ritchie also criticizes Pinker’s excellent book The Language Instinct. It has, he claims, ‘…a very specific agenda – to support Noam Chomsky’s theories about our language skills being innate’. He goes on: ‘other areas of linguistics are glimpsed, if at all, fuzzily in the background.’ From here, Ritchie moves on to the kind of linguistics he’s interested in and discusses how the language of the Proto-Indo-Europeans laid the foundations upon which many modern languages have been built.
Interested readers might notice that this topic, far from being ignored in The Language Instinct, actually receives a whole chapter. Has Ritchie read it, I wonder?
And how any of this justifies Ritchie’s blaming Chomsky for the view that most people (if they, in fact, do) believe that Standard English is the correct version of English, I just don’t know. I welcome an explanation. It appears to me to be some convoluted argument resting on the two claims above. Since the two claims are fallacious, however, it simply can’t go through.
The study of language and linguistics is in far better shape now than it was fifty years ago. Why? Because Chomsky.
[1] We’ll ignore for the time being that (a) by far the lion’s share of Chomsky’s work is about politics, language and propaganda, and (b) that he has never denied the role of environmental factors in language acquisition.
Ritchie ‘s article is very poor indeed in almost every technical aspect. Nevertheless, regardless of who is to blame, the truth is that lay people still think that those who speak substandard varieties are dumb or uneducated and that there is a “correct” English (or Spanish or…). That’s a very important point and it’s great that such a thing can be read in a general newspaper, because this idea is as scandalous today as the idea that men descended from apes in Darwin’s time. The example of the job candidate is clear enough about the terrible social consequences this has. That’s a point, and it doesn’t matter who is to blame about it: we should work to erradicate this pernicious prejudices about language.
Moreover,I would have liked to see in your review an examination of Ritchie’s claim that “there is evidence from neurology and genetics that language cannot be innate”. I don’t know the issue enough to express and opinion, but I was shocked by something claimed by Paul Churchland (not necessary to explain who he is): that in the whole human genome there are not enough bits to codify universal grammar. Is it true or false, I don’t know, but I suspect Churchland is wrong in saying so, because, as Dawkins explains, the genome is not a blueprint but a recipe, and god knows what embriology can make out of it.
Last, let me ask you something. I’m investigating the empirical testing of linguistic theories, and you seem to be a specialist in generative linguistics. So let me pose you a question: how do generative linguists empirically test a given (generative) grammar? I mean, Chomsky says a given generative grammar is a theory of a given language * (equating a language to an (infinite) set of sentences). So, well, say I want to test empirically a grammar of L some linguist proposes. How do I proceed? I’ve read some generative linguistics (not much, but something) and all I found was that I must do something like asking a speaker of L. But, alas, that’s too general and I would like to see the very experimental protocols.
* (and at the same time, a representation of the knowledge of that language a speaker has.)
Sorry for my poor (sub-standard!) English. English is my second language, my native language is Spanish.
Yours faithfully,
Pablo Usabiaga
Universidad Nacional del Sur
Argentina
Pablo, thanks for your reply, and please don’t apologise for your English! I’d be reluctant to call what I wrote a ‘review’. I write reviews for journals etc and you’re absolutely right that any serious review would provide references in support of the innateness claims. Mine was simply a brief response to one of the shoddiest pieces of pseudo-academic writing I have ever seen. I’m also not really an expert in generative linguistics (though my degree at UCL was very much a generative one. I’d recommend this book – http://www.amazon.co.uk/Chomsky-Ideas-Ideals-Neil-Smith/dp/0521546885 – by my teacher, and now friend, Neil Smith. Incidentally, if you’re interested in matters linguistic, why not become and honorary member of our Brighton Linguistics Facebook group?
Thanks for your interest,
Tim
Thank you very much, Tim! I’ll look for the group in facebook and join in.
Pablo
Great article! I do take Pinker’s main argument to be able the innateness of language, even though he does include an entire chapter on language change, but I don’t see why this is a criticism. Ritchie may disagree with this thesis, but books have theses. No book could ever discuss every aspect of language. But, as you say, there is certainly room for diverse topics and views in a library. Finally, anyone who knows anything about linguistics knows that connectionism has never been able to produce an artificial system that learns language in a natural-language-like way (e.g., making the types of mistakes that children make). Rumelhart & McClelland’s “past tense” debacle is the iconic example.