Last year, Alexander Clark and Shalom Lappin published a book under the title Linguistic Nativism and the Poverty of the Stimulus. The background of the book was a course on "The Poverty of the Stimulus, Machine Learning, and Language Acquisition", which the authors gave at the LSA Summer Institute at Stanford in 2007. In the preface, the authors thank an impressive collection of linguists, computer scientists, psychologists, and philosophers "for helpful discussion of many of the issues that we address in this monograph".
An hour-long discussion between the authors and Chris Cummins was just released as a podcast in the New Books in Language section of the New Books Network. Cummins' online intro to the podcast opens with this bit of snark:
In linguistics, if a book is ever described as a “must read for X”, it generally means that (i) it is trenchantly opposed to whatever X does and (ii) X will completely ignore it. Alexander Clark and Shalom Lappin, Linguistic Nativism and the Poverty of the Stimulus (Wiley-Blackwell, 2011) is described, on its dust-jacket, as a “must read for generative linguists”. Apparently generative linguists have so far taken the hint. This is a great pity, as this book is not only very pertinent, but also succeeds in eschewing most of the polemical excess that tends to engulf us all in this field.
Richard Sproat, who contributed the cited jacket blurb, quipped in an email that "Apparently I did you a disservice by saying it was a must read".
In the first place, I don't believe that this book has been ignored — for example, Nick Moore contributed a thoughtful 3/4/2012 review on linguistlist. But to the extent that the book has had less impact than it should, the fault is not Richard Sproat's, but rather Wiley-Blackwell's. If you want to read this book, you'll have to shell out $104.95. You can get it in ebook form from publisher for $84.99, or from amazon for $79.99. That's a lot for a 216-page monograph.
In comparison, the other dozen books featured so far in the New Books in Language series have an average price in paper form of $30.43, or $19.65 as ebooks. Those prices are still on the high side, but they're well within the range of what a non-specialist might pay for a book of likely interest, and they're also reasonable prices for a book to be assigned as part of the reading list for a university seminar.
The business model of many European publishers who specialize in scholarly or scientific books seems to be to enlist eminent authors to produce intellectually serious and even important books, and to depend on the fact that a world-wide collection of about 500 research libraries will buy one each of everything they publish, sight unseen. Then a price of $100-$200 brings in $50-100K per title. The authors get nothing, of course, except another item for their bibliography; and the world gets a "must read" book that almost no one reads.
This has always been a morally dubious enterprise, in my opinion. In the days when typesetting and printing were expensive, and other ways to circulate such content didn't exist or were of markedly inferior quality, it made some sense; though even then, the various lower-priced alternatives were often used as they became available — ditto copies in the middle of the 20th century, and xerox copies later on.
But today, this model doesn't facilitate scholarly and scientific communication in any way — it purely retards it. Isn't it past time for a prestigious scholarly press or scientific society to experiment with an Open Access digital monograph series?
And yes, I know that there would be real costs associated with such an enterprise, and that a way would have to be found for them to be paid. But in the world as it exists today, the present way of handling the publication of scholarly and scientific books and monographs strikes me as pretty nearly the worst method that could possibly be invented. Authors are paid (essentially) nothing and get fewer and fewer readers; interested readers are asked to pay so much that they are forced to ignore most offerings; academic disciplines are encouraged to become more and more ingrown; and libraries' hard-pressed budgets are required to subsidize the whole charade.
One alternative solution would be a monograph series sponsored by a scholarly and scientific society — the Association for Computational Linguistics already sponsors a prestigious Open-Access journal. A slightly more conventional alternative would be a series that made an honest attempt to calculate non-volunteer publication costs, and set prices accordingly.
I've contributed chapters to a number of high-priced edited volumes in the past — sometimes without intending to, as documented in "Echoes from the Dance of the Elephants", 8/19/2007 — but I try not to do so any more.