Geoffrey Sampson (1) (1944–)

I was lucky to come across this book in the humanities bookstore of my local university. Every chapter has given me a mini "Eureka!" moment, which is very rare for a proceedings-style book.

In linguistics, it has long been taken as an axiom that all languages are equally complex. The chapters of this book all challenge this assumption. From showing that the idea itself arose in the late fifties, without any real quantitative or qualitative research behind it to back it up, to comparing two show more closely related languages (Elfdalian and Standard Swedish) and showing that one is definitely more complex than the other, to showing that languages (creoles, that is) grow more complex and can become so by overtly and analytically marking information that no other languages mark and not through growing inflections, to showing that literate registers grow out of formal registers, that in turn are more complex than informal registers, to discussing whether it makes sense at all to talk about "overall complexity" for any language... It is very tempting to whisper "paradigm shift".

This is certainly the most important book I have found so far this year, and possibly for several years to come. show less

There's nothing better than seeing an overconfident favourite getting a proper seeing to from an unfancied underdog.

All the same, when best-selling MIT and Harvard-credentialised psycho-linguist Steven Pinker's book "the Language Instinct" - a work feted far and wide and rarely challenged in polite circles - is subjected to critical treatment by an curmudgeonly British professor from an unfashionable second tier university in the home counties, it is a hopeful chap indeed who thinks an show more upset might be on the cards.

Pinker, after all, has the weight of Noam Chomsky (self styled most important intellect on the planet) behind him, and rates consistently favourable mentions from the literary review sections of important newspapers and that peculiar clique of populist science writers (Dan Dennett, Alan Sokal and Richard Dawkins among others).

The best you could say for Sampson, on the other hand, is that he lacks profile: His tenure is at the University of Sussex - yes, there is one - and the profile he does have isn't the sort most people would want: as far back as 1977, Christopher Hitchens described him as "an academic nonentity who made various other incautious allegations [about Noam Chomsky's political views] and who later ... strolled into the propellers and was distributed into such fine particles that he has never been heard from again." Ouch.

That's all ancient history, though, and the pleasant surprise is that over the last thirty years the plucky little Britisher has made a full recovery from his encounter with the propellers and is in fine enough fettle to give said global linguistic superstar a good old-fashioned intellectual walloping. Even read alone, Pinker's book is built on a wobbly edifice, but with Sampson's expert guide, it looks positively idiotic. Sampson is systematic: he sets up each of Pinker's arguments (such as they are), represents them fairly (I read Pinker's original concurrently to check) and then, like a gentleman cricketer on the village green dispatches each of them deftly to the boundary through extra-cover.

I'm really not sure why Geoffrey Sampson's book hasn't received more attention: possibly the author's history (he seems to made a number of "incautious" political statements over his life and doesn't seem to be the recanting type), but also because it swims bravely against an intellectual tide: Sampson is - though I don't think he expressly says it - a relativist:

"What the language learner is trying to bring his tacit theory into correspondence with is not some single, consistent grammar inhering in a collective national psyche, the sort of mystic entity that a sociologist such as Emile Durkheim would call a "social fact". Rather, he is trying to reconstruct a system underlying the usage of various speakers to whom he is exposed, and these speakers will almost certainly be working at any given time with non-identical tacit theories of their own - so that there will not be any wholly coherent and irrefutable grammar available to be formulated"

Advocating relativism, as I think Sampson coherently and convincingly does, has the misfortune to be about as incautious as criticising Noam Chomsky these days, so perhaps Sampson's card is marked and that's that. All the same, the passage cited above is beautifully put, and by itself is more persuasive than Steven Pinker's whole book.

All the same, who's laughing now? Probably not G. Sampson esq., as he strolls from the wicket at stumps, having carried his bat valiantly, but not having managed to save the innings. But up on the grassy bank, this cricket connoisseur stand to applaud this stylish, defiant knock.

Well batted, sir. show less

I enjoy looking at both side of an argument; as an introductory linguistics course I am taking has highlighted this book as a good way of looking at the other side of the Chomskyan claims I swiftly ordered my copy, fully expecting several evenings of intellectual challenge and brain stretching. The book has fully disappointed me. It is a pamphlet rather than a scientific work.

I'll propose that the first paragraph of the book is an acid test for the potential reader. It reads as follows: show more "the English language, and other languages, are institutions like country dancing or the game of cricket: cultural creations that individuals may learn during their lifetimes, if they happen to be born into the appropriate cultures, but to which no one is innately predisposed".
If you're struck by the fact you hardly know anybody who doesn't speak a language (I know nobody having chosen to not learn one during her lifetime), but probably know several people who don't play cricket you'll be hard pressed to find a satisfying explanation.
If on the other hand, you find this claim fairly self obvious (maybe you don't know anybody who doesn't play cricket for instance) and are in fact scandalized anybody could think otherwise, you'll probably enjoy the book. show less

Part nuts-and-bolts computational linguistics articles and part cri de coeur from a researcher who is exasperated at the hand-wavy route much theoretical linguistics has taken. I am sympathetic to the latter, but bought the book for the former, specifically the chapter "Good-Turing frequency estimation without tears", a detailed tutorial on this smoothing technique that I couldn't find anywhere else.