Child language acquisition: Contrasting theoretical approaches (review)

The field of language acquisition has long suffered from a lack of serious, real dialogue between competing approaches. It is to be welcomed, then, that researchers of Ambridge and Lieven's caliber are spearheading a movement toward more collaboration in understanding the complexities, advantages, and disadvantages of different viewpoints. Child language acquisition (CLA) is a monumental effort to kickstart more unbiased, comprehensive comparisons between competing generativist and constructivist theories of language development. The book's intended audience is researchers and instructors for undergraduate/graduate courses, although as we see below, some caveats apply.

CLA contrasts the two major approaches to language acquisition—generativism/nativism and constructivism/empiricism—across selected subdomains of development, and examines critically the empirical support for each view in each domain. It does not advocate a 'third way' that would reconcile both approaches, because in many important respects nativism and empiricism are incompatible: a particular kind of knowledge is either present at birth or it is not, hence learned. CLA's main contribution is in its painstaking attention to the assumptions and predictions of each theory, whether and how these have been supported by empirical data, and how we can continue to try to adjudicate between opposing views in each area.

After Ch. 1, which introduces the two main approaches and summarizes different research methods for reference, the book is organized around subareas of language acquisition that figure prominently in generativist/constructivist debates (pragmatics and formal semantics are missing). It is not intended to provide a chronological overview of language development as is the case with many introductory textbooks. CLA proceeds by examining evidence for each approach and laying out ways to advance our knowledge in each area, a goal it accomplishes admirably well. CLA is not intended to advocate a particular viewpoint: in this respect it departs from the genre's traditions, where authors in general advocate for their own theories (with some exceptions). In many areas, there is no clear 'winner' anyway, and A&L say so. CLA succeeds at another important goal: showcasing in detail serious and well-supported constructivist alternatives to what many still take to be the 'mainstream' generative paradigm.

The presentation of each approach is simply outstanding in its accuracy and detail. Ch. 4, for instance, presents a somewhat simplified summary of modern generative syntax in the Chomskyan tradition (X' theory, etc.) that is better than most anything else I have laid eyes on. This is a very difficult task because Chomskyan theory is sort of a moving target, where analyses of acquisition are heavily dependent on features of the theory that regularly become obsolete. A&L's discussion of syntactic bootstrapping in Ch. 3 is, I thought, particularly good, mixing methodological and conceptual critiques (see for instance n. 10 on p. 378). The book does present some minor linguistic inaccuracies, but these do not detract from its meticulousness. Ch. 9, for example, dismisses the uniquely human character of language as evidence for innateness by comparing it to activities like driving. This careless analogy is unfortunate because it offers an easy target for critics of constructivism (i.e. that constructivists do not really understand how complex language is). A&L would do well to eschew this kind of reasoning and unwarranted use of loose analogies (which are also sometimes found in generativist writings). However, these very few places where a linguist might take exception with A&L's characterization of language (another place is their discussion of the presumed advantage of nouns over verbs because the former are 'virtually infinite') are not enough to be of concern.

Although A&L are known constructivists, they provide fair appraisals of positions they are sympathetic to: to wit, their criticism of L's own studies on verb-specific schemas (Ch. 6, pp. 218-19), their discussion of the acquisition of sentential complements (Ch. 7, p. 310) where they conclude that empirical data do not fully support either position, or their critique of the social-pragmatic approach...