In recent years, growing interest in patterned realizations of pragmatic functions has led to renewed attention to the notion of local grammar. Originally proposed by Gross (Reference Gross and Hoey1993) and subsequently developed in the work of Sinclair and his colleagues (e.g. Barnbrook & Sinclair Reference Barnbrook, Sinclair, Sinclair, Hoelter and Peters1995; Hunston & Sinclair Reference Hunston, Sinclair, Hunston and Thompson2000), local grammar has been employed in the past decade as an analytical approach to the study of individual speech acts and discourse acts, including evaluation, exemplification, definition, request and apology, among others. Against this backdrop, Hang Su’s Local Grammar Approaches to Speech Act Studies: Apology in Contemporary Spoken British English represents a timely and valuable contribution. Situated at the intersection of corpus linguistics, pragmatics and discourse analysis, this volume develops a local grammar approach to speech act studies, illustrated through analyses of apology in contemporary spoken British English.

The book contains eight chapters. Chapter 1 (pp. 1–16) outlines the disciplinary background by reviewing related research fields. Su begins with a concise discussion of corpus linguistics and pragmatics, insightfully arguing that corpus linguistics should not be viewed as purely quantitative, but rather as ‘a kind of qualitative study based on quantitative data’ (p. 4). As pragmatic research has traditionally been regarded as qualitative and context-dependent, recent scholarship has increasingly emphasized the compatibility between corpus methods and pragmatic research. The chapter then introduces classic theories of speech acts and identifies two limitations in previous studies: first, the lack of a widely accepted taxonomy of speech acts poses difficulties for developing annotation schemes and annotated corpora; second, early speech act studies have not sufficiently engaged with corpus data. The subsequent discussion narrows to apology, which Su considers particularly suitable for local grammar analysis because apologies are ubiquitous in everyday interaction and are realized through relatively fixed and recurrent lexicogrammatical patterns.

Chapter 2 (pp. 17–45) develops the central theoretical framework. Su draws on Hunston’s (Reference Hunston2002: 178) definition of local grammar as an approach that ‘seeks to account for, not the whole of a language, but one meaning only’. Accordingly, a key feature of local grammar is its use of functional terminologies to analyse formal elements. For instance, in the apology I am sorry for the late reply, I is analysed as the Apologizer, am as the Hinge, sorry as the Apologizing element and for the late reply as the Specification. Su argues that the advantage of such an analysis is that the resulting grammatical description becomes more transparent and context-specific. He then proceeds to discuss methodological issues in local grammar analysis. The first concerns data retrieval, namely how to identify linguistic forms that realize a certain communicative function. Su summarizes three data retrieval methods, which start respectively with lexical items (e.g. sorry in apologies), sentence structures (e.g. X means Y in definitions) and recurrent lexicogrammatical patterns (e.g. V it ADJ that in evaluations). The second methodological issue concerns component labelling, that is, how to identify functional elements such as Apologizer, Hinge and Specification. Su proposes the principles of transparency and economy, arguing that functional labels should clearly reflect communicative meanings and that unnecessary proliferation of categories should be avoided. A particularly interesting theme in chapter 2 is the comparison with several related analytical frameworks, including FrameNet, corpus pattern analysis, transitivity analysis in systemic functional linguistics and construction grammar. The final part of the chapter discusses the implications of local grammar for grammatical description and its applications in language teaching and natural language processing.

Chapter 3 (pp. 46–71) presents a more detailed discussion of issues in speech act annotation. Su reviews established taxonomies of speech acts and annotation schemes such as the Speech Act Annotated Corpus, the Dialogue Annotation and Research Tool, and the SPICE-Ireland Annotation Scheme. He notes that these existing schemes, which annotate only at the utterance or clause level, are genre- or task-specific. He suggests that the taxonomy proposed by Edmondson et al. (Reference Edmondson, House and Kadar2023), which comprises a relatively minimal set of speech acts, may help avoid the proliferation of ad hoc categories and that speech act annotation should include smaller fragments or units that carry pragmatic meaning. The chapter subsequently introduces the local grammar approach to speech act annotation, using apology and request as illustrative examples. Su further proposes two annotation principles: the first is form–function correspondence, meaning that every functional label should correspond to an identifiable form; the second is semantic completeness, which requires annotators to infer implied meanings in context. This is followed by a brief discussion of the possibility of automating speech act annotation with Large Language Models (LLMs). Finally, Su argues that the advantage of the local grammar approach is that it allows for more delicate and transparent annotation and that it facilitates quantitative and comparative corpus-pragmatic research. However, there are also challenges, such as the difficulty of identifying functional labels consistently and the practical difficulty of large-scale annotation.

Chapter 4 (pp. 72–102) presents the core corpus analysis of the volume. The analysis begins with searching for eight typical illocutionary force indicating devices (IFIDs), namely sorry, pardon, afraid, excuse, apologise/-ize, apology/-ies, forgive and regret, in the spoken components of the British National Corpus 1994 (Spoken BNC1994) and the British National Corpus 2014 (Spoken BNC2014). The functional elements in each concordance line are then manually annotated, yielding 27 local grammar patterns, organized around two main strategies, namely making apologies and seeking forgiveness. The patterns identified range from short and highly formulaic items such as Apologizing (e.g. sorry) and Forgiveness-seeking (e.g. forgive me), to more expanded structures such as Apologizer–Hinge–Intensifier–Apologizing–Specification (e.g. I was awfully sorry about that spider affair). Normalized frequencies and percentages are reported to reveal the distribution of the identified patterns. Su suggests that these results ‘can reveal the correlation between the complexity of linguistic realisations and the severity of offense or social distance’ (p. 87). However, I believe this correlation may not be as straightforward as suggested, since local grammar analysis alone cannot fully determine the nature or severity of the offence without further consideration of the larger discourse and situational context. To illustrate, Forgive me may imply a more serious offence than the structurally more elaborate I’m sorry I’m late. The chapter then shifts focus to speech acts in languages other than English by discussing exemplification in Chinese and apology in Italian. While the shift here appears somewhat abrupt, these sections effectively demonstrate the cross-linguistic applicability of the local grammar approach to speech acts. In the concluding discussion, Su reiterates three advantages of the proposed approach: it can provide systematic descriptions of recurrent realization patterns, clarify the relationship between linguistic form and communicative function, and contribute to methodological developments in corpus pragmatics.

Chapter 5 (pp. 103–27) broadens the scope of discussion by exploring the relationship between local grammar and construction grammar in speech act research. Su argues that their shared emphasis on recurrent form–function pairings justifies treating recurrent realizations of speech acts not only as local grammar patterns but also as constructions. Notably, Su’s ‘speech act constructions’ are identified at a mid-level of specificity. As such, they can be distinguished from Pérez-Hernández’s (Reference Pérez-Hernández2021) request constructions, which range from highly abstract constructions such as interrogatives to highly specific ones such as I would like you to do X. In Su’s analysis, the Apologizer–Apologizing construction comprises the patterns Apologizer–Apologizing, Apologizer–Hinge–Apologizing and Apologizer–Intensifier–Apologizing. The construction is formally realized through the structures Person BE (adv) sorry and Person (adv) performative verb (e.g. apologize, regret), with examples including I’m terribly sorry and I sincerely apologize. Su further demonstrates that local grammar constructions have important implications for construction grammar, speech act studies and language pedagogy in EFL/ESL contexts. The chapter also outlines several future research directions, especially empirical research on the psychological reality of speech act constructions.

Chapter 6 (pp. 128–52) further extends the corpus analysis from synchronic to diachronic description by examining changing patterns of apology in contemporary spoken British English. Su first reviews diachronic research on speech acts and highlights several methodological challenges, namely the difficulties in accessing diachronic spoken data, comparing speech functions across time and quantifying speech act realizations diachronically. Against this background, Su argues that local grammar can provide a more reliable basis by offering two pieces of evidence: (i) utterances containing different IFIDs may realize the same local grammar pattern and (ii) utterances containing the same IFID may realize distinct patterns. His own corpus findings show that apologies in contemporary spoken British English are decreasing overall, while at the same time becoming more conventionalized and semantically weakened. Specifically, Su finds that patterns typically realized through short and routinized expressions show an increasing tendency of use. By contrast, patterns involving elements such as Specification and Forgiveness-seeking, which are more closely associated with sincere apologies, decline significantly in frequency. That is to say, expressions originally functioning as genuine pleas for forgiveness have gradually evolved into more casual acknowledgements of relatively minor transgressions. The chapter concludes by noting that the observed changes may either reflect decreasing social distance or indicate growing social stratification. Su also cautions against directly interpreting these changes as evidence for a shift in British society from positive politeness to negative politeness, as proposed in previous research (e.g. Kopytko Reference Kopytko1995; Culpeper & Gillings Reference Culpeper, Gillings, Love, Brezina and Aijmer2018).

Chapter 7 (pp. 153–79) further examines the impact of speaker sex and culture (or region) on speech act realization in contemporary English. Su first reviews previous research on sex differences in apologies and then investigates the distribution of apology patterns in the British National Corpus (BNC) divided by sex. The frequency findings reveal a notable diachronic shift: whereas women apologized more frequently than men in the Spoken BNC1994, men were found to apologize more frequently than women in the Spoken BNC2014. Su argues that the finding may suggest that increasing emphasis on gender equality has begun to influence discourse practices in British society. The second half of the chapter exemplifies the local grammar approach to intercultural speech act studies, with data drawn from the International Corpus of English (ICE). Focusing on three Asian Englishes (Singaporean English, Indian English and Hong Kong English) in comparison with British English, Su presents the findings of an earlier study, Su & Lu (Reference Su and Lu2023), which reveal a mixed picture of how apologies are performed across the four regional varieties. The results are taken to suggest that the three Asian Englishes may have undergone a process of language indigenization and pragmatic nativization.

Finally, chapter 8 (pp. 180–96) concludes the volume by recapitulating the major contributions of the local grammar approach to speech act studies. Su revisits key issues discussed throughout the book, including speech act annotation, the development of local grammars of speech acts, the identification of speech act constructions, and the application of local grammar to contrastive and intercultural speech act research. The concluding discussion also highlights broader implications and applications. Su argues that research on local grammars of speech acts may contribute to the development of a function-based communicative grammar and to the teaching and learning of speech acts. Particularly helpful is Su’s proposed five-session teaching procedure for apology, which includes: (1) introducing key concepts, (2) exploring the functional terminologies of apology, (3) production practice, (4) consolidation and (5) exercises and assessment. The final remarks point toward several promising avenues for future research, most notably large-scale investigations of speech acts across different languages and cultures.

Overall, a major strength of this volume lies in the clarity with which complex theoretical and methodological issues are presented. The writing is fluent and reader-friendly, with empirical findings consistently connected to broader questions. Sue convincingly demonstrates that local grammar opens up new possibilities for investigating synchronic variation and diachronic change in speech act realization. In my view, this approach is not only useful for the study of speech acts, but can also be extended to a wide range of discourse practices, such as multimodal, technical and professional discourse, and computer-mediated communication. At the conceptual level, local grammar can be productively integrated with usage-based linguistic theories. Methodologically, it is strongly data-driven and context-sensitive, which gives it considerable potential for further integration with statistical methods and large language model techniques. Together, these strengths make local grammar an important conceptual and methodological advancement in corpus linguistics within the Sinclairian tradition.

Nonetheless, perhaps precisely because of its ambitious scope, the analysis presented in this volume raises several challenges, some of which Su also openly acknowledges throughout the book. First, the well-cited characterization of local grammar as an approach that ‘seeks to account for, not the whole of a language, but one meaning only’ (Hunston Reference Hunston2002: 178) invites the question of whether a single meaning can ever be isolated from the multifaceted and overlapping pragmatic meanings found in actual language use. Relatedly, the identification of speech acts itself remains methodologically challenging. In the present study, apologies were retrieved via searching for conventionalized IFIDs, an approach that risks overlooking implicit realizations of apology, even if such cases may be infrequent (see also Wen et al. Reference Wen, Wang, Yao and Averyanov2026). Conversely, the presence of an IFID does not guarantee that an utterance functions as a genuine apology: excuse me, for instance, may also convey irritation or anger rather than apology. In addition, the present work does not fully address functional variation within the same pattern: expressions such as forgive me, excuse me and pardon (me) were regarded as instantiations of the same pattern, despite their distinct uses. Future research would therefore benefit from engaging more closely with extended discourse context and situational context. Quantitative analyses could also move beyond descriptive statistics to incorporate more sophisticated modelling of pattern distributions. This said, I regard local grammar patterns not as the endpoint of linguistic analysis, but rather as a productive starting point for understanding broader and more deep-seated regularities of language use.

Finally, one limitation of the volume concerns a noticeable degree of repetition across chapters and sections. This seems to be understandable, as the volume synthesizes and further develops a series of the author’s earlier publications produced at different stages of the project. For example, section 2.5 (‘Local grammar and speech act studies’) appears somewhat unnecessary because it is already the central theme of the entire volume. Overlapping methodological concerns are discussed in section 2.2.2 (‘Methodological issues concerning local grammar research’), section 3.2 (‘Issues concerning speech act annotation’) and section 3.5 (‘A local grammar approach to speech act annotation: Potentials and challenges’). Similarly, section 4.2 (‘Why a local grammar approach to account for speech act realisations?’) and section 4.6 (‘Developing local grammars of speech acts: Benefits and challenges’) largely repeat the motivation for the study, much of which has already been introduced in the preceding chapters. Repetition also appears in chapter 5, whose focus is the connection between local grammar and construction grammar. In this context, section 5.2.1 (‘Revisiting the concept of local grammar’) and section 5.3.1 (‘An overview of the local grammar approach to speech act studies’) seem redundant, given the extensive earlier discussion of these topics. Additionally, some of the frequency tables in chapters 6 and 7 contain overlapping information, with the same raw and normalized frequencies appearing in separate tables, one with and one without significance testing results (e.g. table 6.8 overlaps with table 6.9, and table 7.9 with table 7.10). In these cases, the presentation could have been streamlined by merging the tables to improve conciseness.

Despite these minor limitations, Local Grammar Approaches to Speech Act Studies remains a highly original and thought-provoking contribution. The book will be of considerable interest and value to practitioners, researchers and students in pragmatics, corpus linguistics, applied linguistics, discourse studies and related fields.