This is an important contribution to the discussion in the field of educational linguistics on the nature of academic language proficiency. The title itself evokes some of the questions in dispute: What aspects of “language” might the author have in mind? And is there really a different kind that could be said to be “of schooling?” Chap. 1 starts by delimiting the domain of school-related language ability that will be the object of scrutiny: advanced literacy. The skeptic may object that neither “school-related” (in relation to “language”) and “advanced” (in relation to “literacy”) are coherent, well-defined categories. In fact, some educators might go so far as to describe Schleppegrell's analysis as a type of “deficit model” of literacy, a characterization frequently directed toward attempts to differentiate between conversational ability and the kind of language ability required for advanced literacy (see Edelsky 1996). Cummins 2000, for example, has been strongly criticized along these lines.

The first chapter sets the stage for the discussion by proposing an alternative approach to the way that previous investigations of academic language proficiency have applied the concepts of decontextualization, explicitness, complexity, and cognitive demand. The evidence presented in subsequent chapters will demonstrate how the general line of research outlined above is on solid ground, despite continued widespread skepticism, in particular from language arts teachers influenced by strong versions of whole-language and constructivist philosophies.

Chap. 2, “Language and context,” begins with a strong claim, in two parts: (1) All normally developing children enter elementary school with language resources that have served them well during the preschool years at home with their families and in the local community; and (2) the ability to use language for academic purposes, “schooled ways of using language,” however, is not distributed uniformly among all first-graders. A lack of experience with written texts and academic-type and classroom-type discourse can make it difficult for many students to learn and to demonstrate learning. After summarizing the research on “sharing time” and “the definition form,” the reader is reminded that for children who struggle with the higher-order literacy-related abilities, the relevant “insufficiency” is simply a question of sufficient experience with school texts, adequate instruction, and familiarity with the expectations of classroom discourse. That is, in all cases (Specific Language Impairment aside) the contributing factors are “extrinsic”; no nonstandard dialect or nonacademic primary discourse ability stands as a cognitive impediment to attaining proficiency in advanced literacy.

Chaps. 3, 4, and 5 address the core issues of the study, in which the claims from chap. 2 are substantiated. The analyses of school texts and student writing merit a careful reading because they specify what are the actual linguistic resources that academic language proficiency calls upon – aspects of grammar that are associated with the discourse of expository texts. The argument, however, isn't limited to this idea, which from a number of points of view could be taken as a safe consensus. Rather, the sentence grammar of the advanced genres of science is integratively bound up with the development of the scientific disciplines themselves; and history is a “textual construction” (125) for which primary, universally accessible, narrative ability is insufficient.

Chap. 6 concludes with a discussion of the pedagogical implications of an approach to literacy development that helps language educators understand the importance of focus on form at the sentence level and metalinguistic awareness at the discourse level. For researchers, the summary of the study's findings (limited as they are to the analysis of texts) offers a clearly formulated set of testable proposals.

Schleppegrell straightforwardly frames her analysis of academic discourse within the perspective of systemic functional linguistics (Halliday 1994). What makes this interesting is the way it is applied to the domain of language ability under consideration. Researchers often exploit the findings of their work to argue broadly against competing models, even when it's evident that they have reached far beyond their findings. This error is scrupulously avoided throughout the book, facilitating congruent readings by students and investigators who may not share the same theoretical perspective. For example, a non-functionalist, Universal Grammar (UG)-oriented reader, interested in different aspects language use and ability (as opposed to linguistic competence in the narrow sense) should find nothing objectionable, in principle, in the way Halliday's theory is applied to the analysis of texts. Every theoretical formulation in the book, without exception, leaves the door open for an objective assessment of the data – a provocation no less, compelling one to ask: How could this be? From a UG point of view, for example, what might explain a potential coincidence on the central claims? The possibility of exploring areas of common ground might even include the discussion in chaps. 3, 4 and 5 on the uneven (i.e., non-universal) access among school-age children to the grammatical resources associated with academic literacy. In fact, this section of the book should help generativists, among others, to reconsider some long-standing assumptions about the development of grammar, from the point of view of both performance and competence. The allusion here is to actual recent critical reconsiderations in the field to which functionalists might be interested in paying attention as well.

Among the many illustrative examples of complex grammar analyzed throughout the book, nominalization is one of the most interesting (pp. 67–74, 118–31, 141–44). We could perhaps all agree that the basic operation involved represents an early-acquired universal attainment of child grammar:

But the resulting noun phrase construction is then placed at the service of expository text. When noun phrases express complex propositions, comprehension becomes difficult for second-language learners and native speakers alike. An analysis must be effected rapidly to determine the properties of the head noun predicate, assign the thematic roles of each argument, and account for those that are left unspecified. We can point to a similar relation between early acquisition and subsequent complexity in the embedding of relative clauses. Consider, also, the possibilities and challenges in the conversion of sentences into noun-headed complements, which can then become the subject of a new sentence:

Barely scratching the surface here of the problem of complex grammar at the service of secondary discourse ability, one unresolved pending question is: Which aspects of comprehension difficulty correspond to knowledge of language per se (competence), and which aspects to information processing ability? (See Jacobs 1995 for discussion.)

Schleppegrell takes a different approach to the question, but the description of the phenomena provides us with a common ground starting point for further discussion. One reason is that all of her examples are pertinent illustrations of the problem at hand. Since a hard dichotomy between competence and performance needs to be abandoned anyway, the starting point might focus on the latter: a serious study of the components of different kinds of language ability. What the components actually are, how they are integrated or how they interface, and so forth, can be the empirical questions.

I will venture to make one recommendation for a second edition. In chap. 1, the concept of decontextualized language use, as it has been applied by other researchers, appears to be strongly questioned. The observation is made that each category of discourse ability and language use calls upon a different kind of context support and background knowledge. Here, it may occur to the reader that, indeed, this qualification is entirely correct. In academic texts (advanced narrative included), however, the manner in which context can be relied upon is different, and the form that context support takes requires a kind of discourse/text processing that is more cognitively demanding than situationally embedded conversational turn-taking. Chap. 3 develops this very theme, with explicit reference to “decontextualized reflection” on meaning (56) – for example, “Language is structured differently when it supports nonlinguistic aspects of a shared context than when the context is created through language itself” (48). So we might want to say that the critical treatment in chap. 1 of the concept of context-reduced language use in Olson 1994 and Snow 1990 should be taken more along the lines of a fine-tuning or elaboration, one that Olson and Snow would probably find entirely compatible with their respective models. The book fine-tunes “decontextualization” with great precision, with a useful and necessary emphasis on the grammatical features that are an integral component of secondary discourse ability.