The Language Instinct: How the Mind Creates Language (Perennial Classics) (original 1994; edition 2000)

It's a hard one to treat fairly, this. Pinker is trying to do three things: 1) provide a general introduction to how language and the science of language work; 2) debunk some common myths, from the really dumb ones about splitting infinitives and suchlike; and 3) give a polemical screed against linguists and psychologists, and there are many, who argue that language in general is not necessarily a dedicated human capacity and that Chomskyan generative grammar in particular is not necessarily what actually generates linguistic productions.

The first two tasks are relatively uncontroversial, and he completes them with reasonable aplomb. There are different choices a non-Chomskyan would have made about how to talk about certain things, of course, notably grammar (although I pity the poor general readin' fool that tries to slog through that chapter for reasons that have nothing to do with my opinions about the content--or Pinker's writing, which is fine--but merely that modelling syntax is a mess and a half and seemingly always will be). And I'm not quite sure his ridicule of the "language mavens" is always proportional to their sins (big difference between hamless logophily stuff like silly fake etymologies and shopping actual class shibboleths like split infinitives).

But it's in what amounts to the same old nature–nurture debate that Pinker makes me turn a little green. Pretty much every linguist and cognitive scientist out there these days, as far as I know, thinks nature–nurture is at most 70–30 one way or the other. That debate is dead, or has at least advanced far, far beyond the stage it's presented at here, and with clearer vision Pinker might have realized that what he was actually doing was engaging in a bit of linguistic historiography on what was a powerful and perennial clash for a long long time. Instead, like a good Chomskyan, he constructs straw men as opponents, reducing the "language is learned" position's scope so that it only covers people who think language is 100% learned and leave 0% room for an innate linguistic module (which is nobody at all, not since, I dunno, Skinner in the fifties?) and then treats all his opponents like they fall into that tiny box.

It's a way to sell copies, I suppose. But it makes you like th Bill Bryson of actual linguists, Steven Pinker, with your reducing Whorfian linguistic relativity to George Orwell's Newspeak, your reduction of learning to vulgar induction (pretty sure everyone in that camp thinks kids learn language mostly by pattern-finding and hypothesizing and trying things out), your trading of tired myths like the poverty of the stimulus and the idea that you can read a billion English (or whatever) sentences and never read the same one twice. Sometimes they are just little factual inaccuracies because you are trying to throw your discipline a coming-out party and want everyone to have a good time and want things to seem more exciting than they are, like when you overstate the scope of the McGurk effect. Sometimes, though, you're making choices that skew things in a more fundamental way, and you are most certainly not stupid, so I think it has to be intentional, building the broader perception of the field in your (camp's) image.

I think I won't even get into the generative grammar stuff, except to say that clearly if you're trying to make arguments about what kind of sentences we find syntactically appropriate even if semantically odd, "colourless green ideas sleep furiously" doesn't license you to take structurally very strange (but interpretable and plausibly grammatical) sentences as examples, because there is nothing weird about its grammar. I also want to say that obviously a general-learning theory that works is not rule- but token-based, and that the weirder and more different your sentences get form normal speech the less relevant it is whether you can make a tree that works for them, since obviously real speakers find them problematic, even if not in the same way as word salad. "Oooh, technically, this is a sentence!" Right, and you make the tree more and more complicated to deal with it, which is the sickness of syntax. "Just move this and this and this and the theory can handle it!" Even when it's quite evident that the human being cannot handle a sentence like that (e.g., with more than a few layers of embedded phrases). Working memory, jerk. It seems so clear that grammar is a few basic parameters and then convention and probabilism gradually laying in patterns over free variation within those parameters. (In this regard I wonder about Chomsky's "minimalism," which I don't know much about. It may be that it handles the rulehappiness of the old ways.)

It gets more innocuous after that. The chapter on language acquisition is good, although Pinker obviously has his biases as hinted above. I like the evolutionary explanation for the critical period--why spend limited genetic resources giving the old human abilities when statistically speaking the human is more likely to be young than old (since in the aggregate more of us are dead when we're e.g. 50 than 20, etc.), and when 90% (or whatever) of all people born will benefit from super language power at age 5 but only some smaller proportion at 35? And the the chapter on proto-language hypotheses (Nostratic, Proto-World, etc.) gives an interesting look into that freaky world (though it has nothing to do with language being an instinct--and there is another dumb error here where he says that the Indo-European people must have dominated everything from Ireland to India, obviously thinking of groups of them spreading out and hten staying in place and developing their different languages, when clearly migration and language differentiation happened simultaneously and in weird back-and-forth never-gonna-be-completely-traceable ways).

But then in the last chapter he's ushing his thesis again in this cowardly way where he deploys Fodor to say "I hate relativism" and some smuggy smug grad student to sneer at the caricaturized version of the standard social sciences model (NOBODY except far rightists thinks human cultures can vary freely and without limit forever, my god. You say this has been the model "since the 1920s" but I think you mean it was the model in the 1920s). And of course he implies by proxy that relativism is more totalitarian since the blank slate (which is what you need at the beginning to have relativism if you are going to take biological determinism off the table, since either our differences are rooted in cognition or in culture) is the dictator's dream. Where to start? First--of course there is a fascist relativism where groups are qualitatively, biologically different and other. And of course a totally blank slate, which nobody thinks is what a person is, can be abused. But there is also a fluid, polyphilic difference-of-tendencies rooted in the various manifestations of culture that leads us to multiculturalism and good things. Pinker seems to be blaming relativism for both the biological and cultural variants, which is silly considering no one relativist can hold both positions.

And alongside Pinker's "we're all one fam" nativist universalism there is clearly a totalitarian universalism where there is one single biological human nature and those who fall outside it are illegit. (In practice, of course, Pinker's universal human nature is more or less neoliberal--he is keen to reject Chomsky's progressive politics.)

And the list of universals across cultures he gives from the work of Donald Brown is interesting and inspiring and takes up several pages (everything from gossip to tool use to extended families some form of privacy urge for sex to a "natural biology" where we recognize the differences between organisms as qualitatively different from those between other objects--so humans have no problem with a wheelchair being furniture and a vehicle but a big problem with a mule being a donkey and a horse--no, it's a mule, crossculturally!)

This would make a great beach read if you felt no compunctions about skipping the sentence trees and if you didn't have a lot invested in the universalism/relativism and naturalism/arbitrarism debates, which I currently seem to. I'll be less judgy when my MA thesis is done, pinks. ( )

As the title of this book might suggest, Steven Pinker, following in the footsteps of Noam Chomsky, contends that humans are born with an innate instinct for language. Not with language itself, obviously, but with mechanisms in our brains that make it easy for us to learn language and that account for commonalities of structure that exist across all languages, despite their obvious variability. Some of Pinker's arguments and conclusions are stronger than others, but the general idea seems pretty sound to me, although I know there's still some controversy over it, two decades later.

Pinker goes into a lot of detail about how languages are structured and how our brains process that structure. I found this detail quite interesting, but rather slow going, despite the fact that Pinker's prose is very accessible to the layman and is broken up here and there with moments of humor or the occasional whimsical quotation. Those who are just looking for a general overview of the subject might find those chapters, which make up about half the book, to be a bit much, but if you're at all interested in the nitty-gritty details of how the human brain constructs sentences, it's well worth reading. ( )

This is indeed quite an amazing book. The writing style is simple, and Stevenmanages to handle this considerably complex subject with a great deal of dexterity.
Each chapter is complete in itself, and I would recommend that each chapter be read on a separate day. This allows you to think about what has been written, before proceeding further.
It is not a book for the casual reader, nor for the dilettante.
It is a book that you must return to after a while. ( )

The incredibly dry chapters on the structure of English grammar marred my enjoyment of what was otherwise a fascinating insight into how people think and communicate. That approximate third of the book presented a real struggle for me to keep my interest. However the rest made up for it. I loved the chapters on the evolution of language and the mind. ( )

A highly readable account on how language is an inherent characteristic of the human species, which I found a bit unpleasant to read at times. Pinker is such a good writer that I feel a little inadequate in responding to his book, but that aside, I thought it was an erudite book on a complex topic, like all Pinker’s books. It is also a bit controversial, as Pinker skewers many a layman’s misguided ideas about language, its origins, and its uniqueness to humanity. And not only a layman’s ideas; Pinker takes everyone from the social scientists to what he calls the ‘language mavens’ (editors and other arbiters of prescriptive grammar) to task for promulgating false ideas about language. I found Pinker’s more polemical chapters a bit uncongenial, mostly because they attack some of my own subconscious ideas about language. I didn’t realise that I felt as strongly about prescriptive grammar until Pinker attacked it and its proponents. I don’t mind Pinker’s attacks on some of the more archaic rules of grammar (such as split infinitives and ending sentences on prepositions, and so forth) but I did find his fulminating a bit tiresome at times, especially when he sets up some straw men that he can easily knockdown. A quibble, really, but still.

Pinker is on much firmer, and to me more interesting, ground when he explains the psychological and evolutionary origins of language. This is simply brilliant and lucid exposition, and I enjoyed it immeasurably. Pinker’s explanation of how language evolves in children, and how this seems to argue for a ‘language instinct’ in humans (Chomsky’s Universal Grammar) is masterful. I also enjoyed his withering refutations of the assertions of those primatologists who claim to have taught chimpanzees sign language, and the more absurd claims of some anthropologists (such as the infamous ‘100 different words for snow’ claimed for the Eskimos).

My one problem with the book is that it came out in 1994, so how up to date it is, in an ever-changing field, is problematic. I wish Pinker would update the book, but maybe he’s too busy writing books about the decline in violence (The Better Angels of Our Nature, which I intend to read next year), and whatnot.

Highly recommended, but not one to swallow hook, line, and sinker. ( )