The title of Neil Smith's book (Language, bananas and bonobos) suggested to me a book focusing on the issue of nonhuman primate language, and the title attracted me to the book for this reason. While Smith deals with that issue in one chapter, the range of issues explored in this book is much wider than I expected, which made it a more interesting read for me. Smith suggests an audience of non-linguists, and I agree that the book would make an interesting excursion for an academically minded reader with only a passing knowledge of linguistics. I also think there are many linguists who would find the book a catalyst for debate and reexamination of linguistic theory and practice.

The book is divided into three sections: (I) “Problems,” (II) “Puzzles,” and (III) “Polemics.” The volume collects revised magazine columns, and as such each chapter is quite short. Though a collection of essays risks lacking the coherence of a singular work, the essays for each chapter are well selected. The only problem in this respect is the overlap of material in chaps. 3 and 4.

I found the most attractive chapters to be those in the “Problems” section. The issues discussed here are those that the general public, students in linguistics, and many linguists will find most intriguing. These include discussions of a system of language for people without the faculties of sight, voice, or hearing, and the extraordinary abilities for the acquisition of linguistic systems among people with very low-level nonlinguistic skills (“savants”). In that discussion, the author challenges traditional notions of “intelligence” while supporting, according to Smith, the idea of a universal grammar. Other chapters in this section focus on interesting disassociations between hearing and perception, allowing the reader to consider why it is that one person can hear and perceive language but not other environmental sounds, while another can hear sounds but can no longer perceive language in the form of sound. Synesthesia and autism are other phenomena introduced in this section. With such interesting topics, the reader may leave each short (4–6 pages) chapter wishing, as I did several times, for more.

Section II, “Puzzles,” presents information on a diverse selection of topics, including but not limited to the connectionist model of language, “politically correct” language, linguistic mechanisms for accomplishing humor, and relevance theory as a resource for the interpretation of imagery. These issues are discussed from Smith's theoretical perspective, which is made clear up front. While only section III is titled “Polemics,” a polemical tone is present in other parts of the book, particularly chap. 12 in section II, in which Smith criticizes extreme claims made by some applied linguists that “the whole realm of theoretical linguistics should be replaced by the study of texts” (72). Yes, that is extreme, and such views can be criticized. But there are many linguists who would consider equally extreme Smith's claim that “we need to distinguish knowledge of language (the initial domain of linguistic inquiry) from the use of that knowledge in particular areas” (9). The reader looking for bridge-building between “formalist” and “functionalist” approaches to the study of language will not find it here.

Section III, “Polemics,” is made up of chapters that seek to persuade the reader that linguistic inquiry is best accomplished from the perspective that language is a modular system, independent of general cognitive capability. This section presents interesting evidence for the idea of modularity as well as criticisms of applied, cognitive, and connectionist perspectives on language.

The opening chapter of section III, “Bonobos,” criticizes strong claims made by researchers of ape language (Savage-Rumbaugh, Shanker & Taylor 1998) that we may need to rethink linguistic theory based on evidence of language competence displayed by one bonobo named Kanzi. Smith is right to criticize the methodological opaqueness of the reporting of Kanzi's grammatical competence, but he attempts to discount interesting findings by criticizing the researchers' perspective on language, calling their adoption of a Wittgensteinian perspective “deeply troubling” (85). Unfortunately, Smith leaves out of the discussion what I find to be the most fascinating aspect of the story of Kanzi – the fact that Kanzi was not originally the subject of the research. Kanzi started picking up language after observing teaching sessions between human researchers and his mother. Although not explicitly taught, Kanzi ended up more proficient than his mother, the subject of the explicit instruction. Kanzi's acquisition occurred, the researchers speculate, because of the modified environment in which Kanzi grew up. The living environment for the researchers and bonobos ensured that much of the everyday, mundane interaction that humans negotiate among themselves was, in this case, negotiated with the bonobos at the research site. The unique relationship between humans and bonobos (as an infant, Kanzi was carried many places by the human researchers) also allowed Kanzi free use of his hands, which in turn allowed him more opportunities for gestural communication. Smith is right that there is evidence for a modular system for human language (presented in earlier chapters), but he seems critical of the ape language research in large part because the data collected on ape language do not fit existing human-specific theories of language. I do not see the need to suggest that, if we look at the communication of a bonobo and reconsider dominant linguistic theory, there might be “dire implications” (84) for the field of linguistics.

Other chapters in section III include discussions of a critical period for language development and the potential conflict between Chomskyan linguistics and the theory of natural selection. The section also includes a review of Pinker's Words and rules which describes that popular book as offering a balanced view of linguistic inquiry (the lexicon does play a role in grammar), and a final chapter which is a light-hearted suggestion that linguists' names seem to share the phonological feature of velarity (Smith calls on all the forces of phonological theory to make this humorous stretch).

Some linguists reading the book may find themselves slightly miffed by the lack of acknowledgment that linguistics existed before Chomsky, but that shouldn't deter any potential reader. This slender compilation of articles addresses some of the more interesting linguistic territory – from a particular theoretical perspective. No doubt the book will persuade the non-linguist of the fascinating realm where the discipline makes its home.