‘Adverbials may be regarded as a rag-bag category in the linguistics system.’ It is hard to think of anyone who might want to take issue with the opening statement of the monograph under review. In fact, it is hard to think of any domain of grammar (not only English grammar) which is messier than the one of adverbials. For this reason alone anyone who volunteers to tackle the Herculean task of writing a(nother) comprehensive book on by far the largest subgroup of adverbials, namely adjuncts, must be admired for their courage and stamina. It must be someone who has explored the territory for a long time and acquired intimate knowledge, which is indeed the case for Hilde Hasselgård, who in 1996 published a first monograph on the two largest semantic classes of adverbials, more exactly on Where and when: Positional and functional conventions for sequences of time and space adverbials in Present-day English (see references). Hasselgård's new book stands in the tradition of Greenbaum's seminal Studies in English adverbial usage (1969) and its successor publications (in Quirk et al. Reference Quirk, Sidney, Geoffrey and Jan1972 and Reference Quirk, Sidney, Geoffrey and Jan1985, with the relevant chapters in both pillars of English reference grammars having been authored by Sidney Greenbaum, and in Biber et al. Reference Biber, Stig, Geoffrey, Susan and Edward1999) and truly complements Greenbaum's Reference Greenbaum1969 classic by focusing on adjuncts. While Greenbaum was solely concerned with linking adverbs, i.e. conjuncts and disjuncts, Hasselgård restricts her study to adjuncts, including under this heading, following Biber et al. Reference Biber, Stig, Geoffrey, Susan and Edward1999, ‘all time and degree adverbials along with focus and viewpoint adverbials’ (p. 23), i.e. adverbials classified as subjuncts in Quirk et al. (Reference Quirk, Sidney, Geoffrey and Jan1985). The overall approach Hasselgård adopts is one that is descriptive, broadly functional and more (but not too) narrowly Hallidayan, which is reflected among other things by her usage-in-text/discourse perspective on adjuncts (notably their use on the textual and interpersonal levels of communication). This perspective is also prominently stated in the cover blurb, where, besides pointing to usage differences of adverbials across text types, the reader is informed as follows: ‘In using real texts, Hasselgård identifies a challenge for the classification of adjuncts, and also highlights the fact that some adjuncts have uses that extend into the textual and interpersonal domains, obscuring the traditional divisions between adjuncts, disjuncts and conjuncts.’

The volume comprises almost exactly 300 text pages and is organised into four parts with altogether thirteen chapters. Part I (pp. 3–63) outlines the overall framework for Hasselgård's take on the field of adverbials, in general, and adjuncts, in particular. It offers the expected background information concerning the major research questions, materials and methods, and organisation of the book (chapter 1) and an overview of the classifications of adverbials as discussed in the literature (chapter 2). In chapter 3 the focus is on the syntactic positions of adverbials in clauses and sentences (essentially, initial, medial and end position) and the semantics, especially the semantic scope, of adverbials (in general and depending on their position). Part II (pp. 67–183) is concerned with the positions of adverbials. It consists of five chapters, the first three of which address adverbials in initial, medial and end position respectively (chapters 4–6). Chapter 7 is exclusively concerned with (it-)cleft focus position of adjuncts and chapter 8 with the combination of adjuncts and, especially, combination of their positions. The overarching topic of part III (187–256) is semantics, more exactly subtypes, frequencies and usage of different semantic types of adjuncts (adjuncts of time and space in chapter 9, of manner and contingency in chapter 10, and of respect, focus, degree, etc. in chapter 11). The synoptic part IV (pp. 259–305) draws together the major findings from the previous chapters, focusing on adjunct usage in different text types (chapter 12) and offering a rather extensive chapter 13 which sums up the major findings and offers some concluding remarks and ideas for future research. The body of the text is followed by a well-selected bibliography (pp. 308–14) and a very helpful general index (pp. 315–20), including a special index for each of the adjunct types distinguished in the volume. The publication dates in the bibliography, by the way, reveal that research on adverbials, and especially adjuncts, in English is largely a topic of the past: the majority of the immediately relevant titles listed were published before the year 2000. At this point it may also be mentioned that the volume has been carefully edited and produced. The only (minor) point of criticism concerns the spelling of the German linguist Holger Diessel, which is consistently misspelt as Dissel (in the text as well as in the bibliography). In the following the most noteworthy points of the monograph will be discussed in the order in which they emerge in parts I to IV.

Many of the opening remarks made above are fleshed out in the scene-setting chapter 1 (pp. 3–13). For example, the author is keen to emphasise ‘that the study is data-oriented more than it is theory-oriented’, its primary aim being ‘to arrive at descriptions of, and explanations for, adverbial usage that can have predictive power and thus be useful in, for example, text production and language teaching’ (p. 12). As to be expected, a substantial part of this chapter is concerned with data and methodology. The main source of her texts was the ICE-GB corpus, with a subset of these (forming the so-called ‘core corpus’ consisting of about 60,000 words; for details see the Appendix on pp. 306–7) having been analysed quantitatively (and in many cases also qualitatively) for adjunct adverbials (pp. 6–7). On the whole, Hasselgård analysed 4,470 adjuncts and classified them according to their position, semantic category, realisation type and discourse pragmatic properties across text types (for very informative visual representations of the adjunct data base see figures 2.1–2.3 on pp. 34 and 38 as well as figures 4.1–4.3 on pp. 68–71). The different text types she investigated from an adjunct perspective are exactly those six distinguished in ICE-GB, namely conversation, sports commentary, social letters, fiction, news and academic writing.

Very helpful in chapter 2 (pp. 14–39) is the discussion of the various (syntactic, semantic and functional) criteria which have been used in previous research for defining and distinguishing between adjuncts, disjuncts and conjuncts, culminating in table 2.1. This table offers a comparison of the classification schemes in the three major reference grammars of English (Quirk et al. Reference Quirk, Sidney, Geoffrey and Jan1985 and Biber et al. Reference Biber, Stig, Geoffrey, Susan and Edward1999 as the two primary sources of inspiration of this monograph, as well as Huddleston & Pullum et al. Reference Huddleston and Pullum2002) and in Halliday (Reference Halliday and Mathiessen2004), followed by a detailed account of semantic categories of adjuncts in these four grammars. To some extent the author is quite sober in judging the outcome of these comparative discussions concerning definitions and classifications of adverbials when stating the following: ‘While the present chapter may have done little in terms of clarifying the status of adjunct once and for all, it has hopefully contributed to an understanding of the view of “adjunct” taken in the present study’ (p. 39). Well, this modest (but at the same time realistic) aim has certainly been reached. Hasselgård adopts a ‘relatively wide definition of “adjunct”. . ., much like the category of “circumstance adjuncts” in Biber et al. (Reference Biber, Stig, Geoffrey, Susan and Edward1999: 776)’ (p. 286). In general, the definitions used in the book follow Biber et al. (Reference Biber, Stig, Geoffrey, Susan and Edward1999) in being ‘mainly meaning-based’ (p. 39; for illustration see on the same page table 2.2, which offers a list of the semantic types of adjuncts used in the book).

Chapter 3 ‘Some syntactic features of adverbial placement’ (pp. 40–63) has a scene-setting function for part II in that it puts centre stage the three major positions of adverbials (initial, medial, end). A major finding of Hasselgård's data analysis is that the end position is the vastly preferred position of adverbial adjuncts (about 77%). This is illustrated in various illuminating tables and figures, showing among other things how different semantic types of adverbials behave concerning their placement in the sentence (see e.g. figures 3.2 to 3.4; on adverbial clauses see also figure 4.5 on p. 87). The chapter ends with a survey of factors (clause-internal or textual) determining adverbial placement for those adjuncts whose position in the clause is variable (see table 3.3).

The first three chapters in part II on ‘Adverbial positions: theme, cohesion and information dynamics’ offer much descriptive detail and statistical underpinning concerning each of the three main adjunct positions (initial position in chapter 4 (pp. 67–95), medial position in chapter 5 (pp. 96–114), and end position in chapter 6 (pp. 115–51)), giving most room to the discussion of functional factors and motivations potentially responsible for placing adjuncts in the relevant position (for initial position see e.g. sections 4.2 to 4.4; pp. 72–86). Essentially, there is nothing strikingly new which the interpretations of the detailed statistical and richly illustrated analyses tell the reader (see e.g. for initial position, the summary in section 4.4.8 on p. 86). What is comforting about these chapters is that the author dispenses with the more and more widely found multivariate analyses and all the statistical machinery that goes with them, which, in most cases, ultimately explain hardly more than a fraction of the observable variation – and, if at all, little more than a rather traditional descriptive approach as adopted in this monograph. A special position of adjuncts is addressed in chapter 7 (pp. 152–68), namely ‘The cleft focus position’, as in Maybe just that it's this week that there aren't enough people around (p. 152) or It wasn't till I was perhaps twenty-five or thirty that I read them and enjoyed them (p. 160). On the basis of an analysis of the information dynamics and discourse functions of all relevant 51 examples that the author found in the entire (!) ICE-GB corpus, it emerges that ‘it-clefts with adjuncts seem to differ from other it-clefts in having an unexpected information structure’ (p. 167). This shows, above all, in an overrepresentation of the given + new pattern in so-called informative presupposition clefts (according to Prince's classic Reference Prince1978 article, these are it-clefts where the cleft clause provides new information while the clefted constituent may contain given or new information). Part II ends with chapter 8 (pp. 169–83) on ‘Combinations of positions’ once two or more adjuncts are used in a clause. As expected (see above), ‘the great majority of combinations (91%) involve end position’ (p. 170). The brief summary of the chapter in section 8.10 may also be read as a conclusion for the whole of part II in that it summarises the major syntactic and, especially, discourse/communicative functions which are associated with the three major slots for adverbials in the clause and thus offer a high prognostic potential. For the end position, for example, this reads as follows: ‘obligatory/predicational adjuncts; foreground; focal (N-rheme); long/complex adjuncts; neutral/default position’ (p. 183).

Part III on ‘Semantic types of adverbials: subtypes, frequencies and usage’ offers usage explorations involving, above all, the distribution of the various semantic adjunct types across text types and Halliday's six process types (material, relational, mental, verbal, behavioural and existential processes). For space and time adjuncts (chapter 9, pp. 187–217), by far the two largest semantic adjunct types (68% of all 4,470 adjuncts analysed by the author), compare pp. 188–95 for space adjuncts and pp. 204–9 for time adjuncts. Discourse functions are also explored for the individual adjunct categories, revealing for time adjuncts, for example, that they serve discourse-organising functions to a greater extent than space adjuncts (especially in narrative), whereas space adjuncts (again, not surprisingly) have more metaphorical meanings and uses (p. 216). The next largest semantic category of adjuncts (22%) is constituted by ‘Adjuncts of manner and contingency’ (chapter 10, pp. 218–39), the former of which (pp. 218–28) are typically non-clausal and denote quality (‘in a ____ way’). By contrast, so-called contingency adjuncts (pp. 228–39) are predominantly realised by clauses and semantically cover five major types of adverbial (or: semantic interclausal) relations: the three big C's (Cause, Condition, Concession) as well as Purpose and Result. What is slightly surprising concerning Hasselgård's account of contingency adjuncts is that she largely relies on just two authorities concerning what is known about the semantics, realisation types, function and usage of causal, conditional, etc. clauses, namely on Quirk et al. (Reference Quirk, Sidney, Geoffrey and Jan1985) and Biber et al. (Reference Biber, Stig, Geoffrey, Susan and Edward1999). Far more detailed empirical studies on adverbial clauses in English from the last twenty-odd years (even corpus-based studies exploring different aspects of adverbial clauses from a functional, textual or interactional perspective) play either a minor role in this chapter or, more frequently, are not mentioned at all. The semantic part of this book ends with chapter 11 ‘Other adjunct types: participant, respect, focus, degree, situation, comparison and viewpoint’ (pp. 240–56). In other words, this chapter deals with rather little-used semantic adjunct types, which together add up to no more than 10 per cent of all adjuncts in Hasselgård's data.

The two chapters constituting part IV successfully pull together and to some extent perspectivise findings from the previous chapters. The focus of chapter 12 ‘Adverbial usage across text types’ (pp. 259–85) is exactly what its title promises, namely an account of the distribution of adjuncts (i.e. their various realisation types, semantic types, positions, sequences, specific functions in discourse) across the six text types in the corpora of the ICE family (i.e. conversation and commentary as spoken registers and letters, fiction, news and academic writing as written registers). Figures 12.1 to 12.6 (pp. 260–6) and, as a summary, table 12.4 (p. 284) provide all the essential information. Perhaps the most interesting general observation this chapter offers is that ‘the spoken–written dimension does not seem to be a crucial parameter as regards the use of adverbials’ (p. 283). As with many of the preceding chapters, chapter 12 is full of informative figures, tables and authentic examples accompanied by (largely descriptive) comments by the author. Finally, chapter 13 ‘The grammar of English adjuncts’ (pp. 286–305) consists of a summary of the major findings of the book and some concluding remarks. There is not really much new the reader takes away from this chapter. In fact, on being confronted with the ‘Main findings of the preceding chapters’ in section 13.2 (pp. 286–95), some readers may ask themselves in retrospect what, in the first place, the importance was/is of exploring and meticulously quantifying certain aspects of the syntax, semantics and discourse use of adjunct adverbials. For example, it never becomes clear what exactly we can learn from the distribution of adjuncts across Halliday's six process types. This interaction of, for example, the different semantic types of adjuncts with Halliday's process types is summarised on pp. 286–9 – but what is the deeper significance of the preferences that certain semantic types of adjuncts have (or do not have) for certain process types? Is there anything surprising to be observed, especially from a systemic functional perspective? There is also not really much news value in stating that ‘because of the range and complexity of meanings expressed. . . [it] thus seemed more and more reasonable that the semantic categories should be regarded as prototypes rather than watertight categories’ (p. 287). As for the way in which adjuncts are syntactically realised, Hasselgård's study shows the expected distribution: the bulk (49.1%) are coded by prepositional phrases, 31.5% by adverbs and adverb phrases, and about 10.7% by finite clauses (p. 289) – and thus finds confirmed almost exactly (see p. 38) the corresponding statistics in Quirk et al. (Reference Quirk, Sidney, Geoffrey and Jan1985) and Biber et al. (Reference Biber, Stig, Geoffrey, Susan and Edward1999). Academic writing and news texts have the longest adjuncts (measured by number of words) – which is not really surprising either, even to the novice in the world of adverbials. The same applies to the preference of individual text types for certain ‘text strategies’, such as the preference for a spatial text strategy in sports commentaries, for a temporal strategy in letters and fiction, or for a contingency, specifically conditional, strategy in academic writing (p. 293), and to what is said about the various roles that adjunct adverbials can play in the information structure of the clause (pp. 293–4). Besides the summary, another important part of the final chapter is concerned with revisiting the definition and fuzziness of the category ‘adjunct’ vis-à-vis the categories of conjuncts and disjuncts (section 13.4, pp. 297–303). Since there are, of course, adjuncts whose semantics not only cover the domain of circumstantial meanings, but also the textual domain (home territory of conjuncts) and the interpersonal domain (home territory of disjuncts), it does not come as a surprise that in section 13.4.5 Hasselgård ultimately suggests prototype definitions for each of these three major classes of adverbials, with a special focus on semantic and pragmatic criteria on top of the well-known ones suggested in, for example, Quirk et al. (Reference Quirk, Sidney, Geoffrey and Jan1985). This reads as follows and is subsequently illustrated in figure 13.4: ‘prototypically, adjuncts express circumstantial meanings and belong primarily to the experiential metafunction of language; disjuncts express modal and evaluative meanings and belong primarily to the interpersonal metafunction of language; and conjuncts express relations between events and/or pieces of discourse and belong primarily to the textual metafunction of language’ (p. 302). It would be hard to disagree with anyone who ventured to say that these might well be statements one could have expected at the very beginning of a comprehensive study of adverbials. Not that there is anything wrong or to be argued here, but there is little of news value either. The last section of chapter 13 offers some brief suggestions for further research, the most interesting of which is the one advocating cross-linguistic (especially contrastive) studies on adverbial usage (pp. 304–5).

In sum, this monograph has one attribute written all over it: solidity. With its focus on (richly illustrated) syntactic and, above all, semantic and discourse(-pragmatic) analyses of adjuncts and their embedding and special functions in discourse and different text types – all based on a broad and representative set of data – this study is, in an entirely positive way, reminiscent of descriptive book-length studies on English grammar as they were mainstream until the 1990s. Not very much is necessarily new, but everything has been solidly compiled, exemplified, quantified and argued. It belongs to the major merits of this book that, at the end, (adjunct) adverbials can claim a less rag-bag status as a category in the linguistics system than they appeared to have at the outset. Undoubtedly, Hasselgård's 2010 monograph is a worthy addition to the classics on English adverbials written in the Greenbaumian tradition and will for the foreseeable future remain the standard reference on the topic.