Nomi Erteschik-Shir's new work appears in a series of ‘surveys’: critical overviews of topics in linguistics aimed at researchers and graduate students, and thus neither traditional textbooks nor monographs. There is no doubt that a clear survey of the literature on information structure (IS) is highly desirable: this area spans diverse linguistic phenomena – syntactic, morphological, phonological and semantico-pragmatic – and recent years have seen an explosion of work on it carried out in a wide variety of theoretical frameworks, with notoriously varied terminology. Such a work would be a valuable resource in teaching as well as research, IS being all too often tacked onto discussions of semantics and pragmatics at textbook level, leaving students and researchers in other fields to discover the complexity and diversity of studies on IS only upon tackling a considerable range of primary literature for themselves.

Erteschik-Shir's book goes some way to filling this gap in the market, but potential readers should be aware that it provides somewhat idiosyncratic coverage, even when allowing for the difficulties of surveying such a diverse literature. It will fulfil at least one of the functions of a survey inasmuch as it will leave almost any reader with an awareness of some work that he or she was not previously familiar with, and it will help to clarify how some apparently separate ideas may be interrelated. It will not necessarily perform the job of giving a balanced view of the relevant literature. It seems to me that at least some of the blame for this lies in the particularities of the ‘critical survey’ format: one senses that Erteschik-Shir was torn between the expository side to the work and the desire to write a monograph on developments in her own ‘f-structure’ approach to IS. This leaves one with a feeling of missed opportunity, as Erteschik-Shir undoubtedly has the depth of scholarship and the insight to have produced a more well-rounded work of either kind.

The resulting imbalance is visible in the basic organisation of the book. Chapters 1 and 2 (‘Introduction’ and ‘Architectures and information structure inventory’) together form an introductory overview of typical IS phenomena, some basic notions such as topic and focus, and major approaches to IS (with a clear emphasis on Erteschik-Shir's own analytical framework). Chapter 3, ‘Configurations’, concentrates on the expression of IS by word order, and includes a robust rejection of the ‘cartographic’ approach employed by Rizzi (Reference Rizzi and Liliane1997) and similar lines of analysis. This chapter also considers Scandinavian object shift and contains some fairly detailed discussion of scrambling in certain languages (in particular Russian, Japanese, Dutch and Persian/Farsi). The inclusion of the latter may surprise some linguists, but Erteschik-Shir makes a good case, based on existing literature, for treating the possible involvement of (perhaps quite subtle or unconventional) IS factors as a key question in research on scrambling. So far, this all seems quite in keeping with the purpose of the volume. The next two chapters, however, concentrate on topics whose inclusion under the banner of IS is much more dependent on Erteschik-Shir's f-structure approach.

Chapter 4, ‘Information structure constraints’, argues (largely following Erteschik-Shir Reference Erteschik-Shir1997) that such traditional concerns of syntactic theory as island constraints on movement and superiority effects are attributable to IS-based constraints, rather than abstract structural ones (elsewhere in the book, Erteschik-Shir also revisits her 1997 argument that quantifier scope can be read off f-structure, thus rendering Logical Form representations unnecessary). It is certainly quite legitimate to air these arguments, which to my mind offer an intriguing alternative to conventional syntactic ways of thinking and merit further investigation (even though in this presentation they do appear to rely on a number of questionable acceptability judgements). Nevertheless, the amount of space devoted to them seems rather disproportionate. Given that chapter 6, ‘The division of labor between syntax and IS’, is a short concluding chapter, these arguments form the subject matter of one of just four longer chapters in the book, implying that they are far more central to the study of IS than they are recognised to be by the vast majority of researchers.

Chapter 5, ‘Aspectual focus’, is an even starker case: this is dedicated to what is effectively a theory of argument structure, developed in a series of works by Erteschik-Shir & Rapoport (see their 2005 article and references therein). Again, the aim seems to be to demonstrate that IS is responsible for much more than linguists commonly recognise. This may be an interesting argument, but it seems plainly disproportionate to give it a chapter-length exposition in a survey of IS literature. It is not even clear how relevant an argument it actually is, since Erteschik-Shir notes that her key notion of ‘aspectual focus’ is not identical to ‘informational focus’ (196). The precise relationship between the two notions is never made clear in this exposition, and the later sections of the chapter contain few or no references to focus of either kind. The chapter is both too long, in the context of the survey format, and too short, as an exegesis of a new approach to a complex issue. As such, it neither enlightens the reader greatly nor (one imagines) does justice to Erteschik-Shir & Rapoport's ideas. This ultimately reflects the broader problem with the book as a whole: attempting to be at once survey and monograph in the space of 217 pages compromises its effectiveness on both sides.

Concentration on Erteschik-Shir's own ideas upsets the balance of the volume at other levels too. While Erteschik-Shir's ‘file-card’ framework has its virtues, it certainly has a number of features that are decidedly unconventional. Most notable among these is the idea that all cases of contrast, including what is traditionally thought of as ‘contrastive focus’, contain a topical element. This is because the contextually-supplied contrast set is thought of as a kind of topic in itself (being cognitively ‘available’, background, etc.). This is crucial to a number of Erteschik-Shir's analyses and speculations, as it allows her to suggest that many kinds of expressions are underlyingly topics, contrary to appearances; for example, the subject in A BOY found the book (where capitalisation represents the falling stress usually associated with focussing). Based on this, Erteschik-Shir goes on to speculate that all left-peripheral IS-related positions might be varieties of topic position, given the common contrastiveness of preverbal foci cross-linguistically (124). Whatever one thinks of the idea of contrast-as-topic, it is undeniably unorthodox, yet it is treated more or less as an established principle in many parts of the book, again compromising the balance of the book as a survey of the IS literature.

Disproportionate coverage of Erteschik-Shir's own ideas would be no real problem in an otherwise balanced survey. There are also some rather surprising omissions, however. In particular, the entire formal semantic literature on IS phenomena is missing (bar one brief and very approximate glance towards Mats Rooth's ideas (28) and a few citations in footnotes). It might be argued that this merely reflects the official nature of the book series, as ‘surveys in syntax and morphology’. Then again, the ‘aspectual focus’ material of chapter 5 is in essence lexical semantico-pragmatics. Indeed, morphology is another area that receives scant attention. In chapter 2 (though, strangely, hidden in a section entitled ‘Summing up focus properties’), Erteschik-Shir introduces the idea that some languages have specialised topic and focus morphemes and presents a few examples of apparent focus-marking from the Papuan language Wambon and the Bantoid language Aghem (40), but this important idea is never revisited. This is a pity, since considerable theoretical weight is placed upon the existence of such morphemes by those who assume that topic and focus are grammatical primitives. On the other side, work like that of Lecarme (Reference Lecarme, Rebuschi and Tuller1999) on Somali suggests that the idea of simple focus-marking morphemes should not be accepted uncritically. More discussion of this point would therefore have been not only appropriate in terms of cross-linguistic balance but also of some theoretical significance.

Erteschik-Shir is generally too willing to accept the analysis of a given linguistic phenomenon as a simple grammatical instantiation of a focus or topic feature. Though there are many analysts of IS who do this with a good deal less critical acuity than Erteschik-Shir, it becomes a particular problem for her when it leads to inconsistency in her presentation. For example, Hungarian is referred to several times as an example of how focus can be expressed through syntactic movement, which in turn is viewed as evidence that focus as such is a linguistic primitive (e.g. 40, 81, 85). However, Erteschik-Shir also correctly notes (86) that the so-called Hungarian ‘focus position’ is now widely accepted to express something other than simple focus (qua information update). She attempts in other ways to reconcile the more specialised nature of this position with her view of IS, but this does not change the fact that Hungarian demonstrably does not display ‘focus movement’ as such, and therefore should not be casually cited as evidence for a universal grammatical focus feature. The point goes well beyond Hungarian: as Kidwai (Reference Kidwai, Rebuschi and Tuller1999) notes, it is characteristic of both ‘focus movement’ and ‘focus morphemes’, but not of prosodic stress, that they relate to narrow foci. These are not only foci of restricted scope but are also associated with particular interpretations, like contrast/exhaustivity. A cross-linguistic view of IS should therefore consider possible connections between such factors and should not simply assume that some basic notion of focus unites the prosody of languages like English with syntactic positions and morphemes in other languages.

Some of Erteschik-Shir's judgements regarding cross-linguistic generalisations are almost certainly coloured by her more general assumptions regarding universality in language. The book contains a number of assumptions, explicitly or implicitly, that are far from uncontroversial, yet are presented as though they were. The most startling of these is the apparent equation of ‘competence’ with ‘universal principles of grammar’ (193). This is stated without discussion, yet it seems to me that it is not the position of any major framework – competence, as ‘knowledge of language’, surely includes language-specific as well as universal facts. Another such assumption is that only the establishment of a minimal set of universal primitives, along the lines of topic and focus, will constitute a desirable outcome of research into IS. With respect to any less universalist position, Erteschik-Shir refers to ‘the unhappy conclusion that there are several types of topics and that languages may mark some but not all of them’ and states flatly that this ‘would not provide an explanation’ (27). The trouble is that Erteschik-Shir does not offer any criteria for a good explanation (she does not appear to be attempting any kind of learnability argument, for example). In the absence of such criteria, arguments from ‘description versus explanation’ can be wielded more or less arbitrarily and as such carry no real force.

One feels that a number of the problems outlined above might have been reduced or entirely prevented by more thorough editing. The book is essentially well written and a number of complex ideas are nicely distilled along the way, but there is regularly a sense that certain changes could have improved it considerably. There are several jolts when the discussion leaps unannounced from the highly specific to the very general, or suddenly moves into a summary of a large section, without being flagged as such. Similarly, it is not always easy to tell whether some exposition of another analyst's work is given in the course of developing an argument or simply as an addition to the ‘survey’ aspect of the book. This is especially noticeable when an entire section turns out to be quite inconclusive – this would be legitimate in a consistent overview format, but sits awkwardly within this semi-monograph. There are also numerous points at which Erteschik-Shir's argumentation is either too dense or too superficial to do justice to her ideas. Often one feels that she attempts an overly condensed outline of some idea where a simple pointer to other work would have been more appropriate (especially when the idea in question is a detail of her own work). Attention to such issues could have significantly improved the final product and thus given a better return on Erteschik-Shir's unquestionable scholarship and insight.