First published in 1997 under the title Educating Eve but now revised and expanded and with a useful foreword from Paul Postal, this book presents and critiques all the main arguments that have been offered in support of the thesis that a body of language-specific knowledge is innate in the human child. Sampson finds the thesis unproven on detailed grounds of logic or evidence, and proceeds to sketch an alternative basis for first-language development, rooted in Popperian science and nurture. As the new title indicates, the book purports specifically to respond to Steven Pinker's The language instinct, but somewhat surprisingly Pinker's book isn't extensively debated here: It seems that the publishing demigod is mostly objected to for having renewed the intellectual standing of the nativist stance in the 1990s, and matters are certainly further confused by Sampson's frank admiration for Pinker's more recent The blank slate.

Although the book is ostensibly written for the educated general reader, I found it heavy going in places. Main ideas and preparatory remarks are repeated several times over, and there is also a structural reiteration in the text's explaining, in turn, how each of a number of arguments used to underpin the nativist hypothesis just does not stand up to careful scrutiny (in Sampson's view). Notwithstanding the title, in one sense this is not a debate, and I increasingly felt the absence of dialogue was regrettable. Sampson rehearses at some length and reasonably fairly what various UG adherents have written at one point or another in support of innatism; but this is still reported speech, as it were, with Sampson as narrator. Then, in rejoinder, rebuttal, refutation, comes his own panoply of evidence and counter-arguments – monologue, not dialogue. How much more rewarding might it have been if, after reading the author's laying into the arguments of Derek Bickerton (or Ray Jackendoff) for a few pages, we had been given Bickerton's own response to Sampson's critique.

Sampson sets out the main lines of the nativist case well. A distinguished linguist who has published a number of informative, independent-minded, and readable books over the years (I particularly recommend Schools of linguistics and Making sense, both from 1980), he has read the centrally relevant publications by Chomsky and other UG adherents with care. There is a sense of a linguist (and political and social theorist) striving to articulate a coherent theory with an integrated explanation of language development, the potential of the individual and the (limited) shaping power of society, and the roles of the mind and brain. It is a complex fusion of liberal-individualist politics, Popperian “guess-and-test” science, Herbert Simon's natural hierarchizing, and Bergsonian emergence, with no place for an imposed level playing field of innate linguistic knowledge.

The analysis often takes the form of reporting some aspect of the innateness argument (speed of acquisition, critical age hypothesis, poverty of the data, etc.), working through aspects of that claim by pointing out the inexplicit or dubious logic that – in Sampson's view – is at work in the adherents' articulation of it, pointing out the limited amount of robust supporting evidence, and finally discussing at length some of the evidence that is incompatible with the claim but, Sampson argues, is entirely compatible with a largely environmentally enabled Popperian picture of language ontogeny. But in general there is not here that systematic, forensic analysis, involving a careful probing of each part of each claim from a range of perspectives, which can be found in other (granted, non-populist!) critiques of UG and innatism.

When the nativists Hornstein and Lightfoot say “People attain knowledge of the structure of their language for which no evidence is available in the data to which they are exposed as children” (quoted on p. 44), Sampson's concern is to hunt for “evidence” in child-encountered “data,” so that he can conclude that children could attain their knowledge of structure postnatally. But here, as you can see, I cannot resist use of scare quotes, as the emphasis on evidence and data in my view imposes much too objectivist and segregationist a framework on what is involved in growing up and developing language. We can alternatively regard children as undergoing socialization and acculturation, learning how to mean and learning how to be, communicating their needs and feelings, orienting to other in addition to self. In doing so we may come to regard the language activity that surrounds the child (and the play activity, the meals routines, etc., and all these not distinct but profoundly interwoven: no meal that doesn't involve kinds of speaking and not-speaking and listening, and kinds of playing and not-playing) as kinds of exemplification rather than evidence. Children are learning how to do things with words alongside learning how to do things with sand, balls, cups, peers, and caregivers. Nor do their lives, as experienced, divide straightforwardly into a linguistic part (aux-fronting in interrogative-construction) and a nonlinguistic part (the sand, the meals, the cups); rather, at every point in time it is a seamless integration. In becoming linguistically proficient, of course there is more than imitation involved; of course cognitive resources facilitating generalization, categorization, hierarchizing, grouping, summarizing, and metonymizing are involved – but nothing so restrictively interpreted as a UG language module.

The kind of evidence Sampson spends most time on includes the following (the example is taken from the only entirely new chapter, using corpus data, in this second edition). Consider “Subordinate clause within main clause Subject” sentences, such as the following:

The assumption (of nativists and Sampson alike) seems to be that someone – child or adult – hearing this utterance will associate it with the declarative “counterpart” sentence, namely

And this hearer will notice that the encountered interrogative version involves fronting not of the tense-carrying part of the leftmost verb (has), which is in a subordinate clause, but of a tense-carrying trace of the later main verb. This “structure dependence” in the framing of interrogatives is not, nativists claim, something that English-acquiring children hear around them and can model their own question structures on; therefore, the instruction must be innate knowledge in some form. The UG poverty-of-stimulus position entails an empirical claim that Sampson tests: Is it true that children never hear questions with (what I would call) complex NP Subjects before producing correct ones of their own? Sampson attempts to answer this, looking in the nearly 5-million-word British National Corpus sample of demographic spoken English (why he didn't look in CHILDES is unclear). The wrinkle emerges that in the BNC demographic corpus Sampson finds no examples of “Sub clause within main clause Subject” questions; in written English samples, on the other hand, there seem to be plenty of examples. So he concedes that it may be the case that children hear few or any of the desired analysis-informing constructions. But if the requisite type of interrogative is absent from the BNC spoken corpus (always assuming the corpus is adequately representative) then it is, contra nativist assumptions, absent from adults' spoken English too. Neither position seems fully supported by the patchy findings.

Besides, while children grasp the Subj-Aux inversion rule before they begin to read, it is harder to be sure they do so without benefit of written English. What about all those books that are read to them, all those audiotapes? Such “spoken written English” will not appear in a standard corpus of spoken English, but it may have a considerable impact on the model of English that children learn and apply. But to my mind this whole exercise is too limited to hunting for intervening clauses and their non-fronted tense-carriers. On what grounds do we exclude, as instructive exemplifications of how to frame a question, innumerable exchanges involving VP-ellipsis like

where intonation and stress and pausing may help the listener orient to the late-occurring main verb, despite the extent of material following the Subject head? The Sampson (and nativist) claim here seems to be that children have to encounter interrogative constructions in which a leftward but lower-cycle tense-carrier is not fronted in order to learn structure dependence. But why should they not learn structure-dependence and hierarchical structure (in particular the nuclear role of the tensed main verb) from a host of contributory factors, including caregivers' intonation, use of ellipsis, question tags, and so on?

Sampson sometimes does a good line in gratuitous offense, which may keep the volume of course adoptions down. On p. 105, for example, à propos of Bickerton, he has an entirely pointless pop at “Green students in Hawaii.” The tone fluctuates somewhat: mainly sober academic, with occasional lapses into colloquialism and journalese. But I liked the idea of the reader skeptical of Platonic Forms responding, “Come off it, Plato, pull the other one” (6). Sampson also manages to put the hype in hypotaxis. In his own eyes bravely defying the p.c. brigade, he treats as “well-founded” the idea that a language with a larger vocabulary or various resources of clausal subordination is indicative of a society with greater “cultural advancement” (101); but this is baffling to me, a credulous associating of incommensurables. Other arguments are also insufficiently worked through to be convincing – for example, one against an innate basis for the complex-NP constraint (a sketch of Gazdar et al.'s Foot Feature Principle and suggestions that movement is possible – if still clumsy – where it puts focus on “a genuine characteristic” or essence of an entity rather than something more temporary). Similarly, in chap. 6, which is on “the creative mind” in general but with a specific objection to the restricted notion of rule-governed creativity associated with the UG-programed brain, the topic seems too large for the pages allotted to it; if one declares that language learning is done by minds and not by brains, and that “minds are not a topic open to scientific theorizing,” more supporting explanation is needed.

There is a brief concluding chapter that attempts to explain why Chomskyan nativism took hold in linguistics departments in the mid-1960s and has remained strong ever since; the jaundiced picture drawn of university culture in the United States and the United Kingdom is too partial to offend, and does not convince. Still, this book was well worth writing and is well worth reading. Perhaps the author tilts at too many windmills along the way, but in formulating the innatist hypothesis as falsifiable and then striving to falsify it by all reasonable means Sampson has treated UG nativism with a scientific respect tellingly absent from the avowals of some of its adherents.