The International Corpus of English (ICE) is radically changing our concept of English. We now know more about English in Asia, however, than we do about English in the Caribbean where the language has had a history of almost four hundred years. Dagmar Deuber helps to redress this imbalance. The book is the product of field and corpus work that Deuber had been conducting for almost a decade before publication and her longterm collaboration with other scholars working on Caribbean Englishes (CarE) and Creoles.

The monograph comprises eight chapters with numerous tables and figures, a map of the Caribbean, two appendices, and an index. Of the eight chapters, six are appearing for the first time. Chapter 5 is an extended version of Deuber (Reference Deuber, Jucker, Hundt and Schreier2009a), while sections of chapter 7 appeared in Deuber (Reference Deuber2010). The first chapter introduces the reader to the book and its aims, and raises some of the chief issues in the sociolinguistics of CarE, such as what exactly constitutes English in communities where English exists in a dynamic relationship with an English-related Creole. This leads to a discussion of the (post)creole continuum, standards of English, and style in CarE.

Chapter 2 covers the use and status of English in the Caribbean, a comparative sociohistorical overview of Jamaica and Trinidad, and a brief discussion of some morphological and syntactic features of Jamaican Creole (JamC) and Trinidadian English Creole (TrinEC). The chapter also contains a useful summary of work on language attitudes in the English-official Caribbean, focusing on research on Jamaica and Trinidad. Deuber points out that much of the existing work on language attitudes in the region focuses on Creoles; however, she counter-balances the dearth of research on English by extrapolating from studies that focus on Creole. Unfortunately, the linguistic history of both territories is too brief—surprising seeing that Deuber notes that ‘Jamaica and Trinidad not only have a somewhat different linguistic make-up and different histories but they also represent different settings in the anglophone Caribbean economically and socially’ (p. 29). This difference, relevant to the study of language as social practice in these territories, demands lengthier treatment.

Chapter 3 comprises an overview of the literature on the sociolinguistics of codeswitching and style, and an overview of corpus-based studies of register. It also contains an overview of work on style in so-called ‘Creole continua’. That the majority of existing studies in these areas focus on Creoles and not English demonstrates the timeliness and importance of Deuber's work, and the urgent need for more of its kind. While the review of the various approaches demonstrates that the author has an impressive knowledge of the literature, the theoretical discussion is not always explicitly made relevant to the study. For example, Deuber discusses critical studies by Youssef (Reference Youssef1990, Reference Youssef, Simmons-McDonald and Robertson2014) and Hinrichs (Reference Hinrichs2006) but does not always lay out their relationship to her work.

Chapter 4 covers methodology: data sources and collection procedures, the ICE corpus design and the Jamaican and Trinidad and Tobago (T&T) subcomponents, the structure of the spoken component of the corpus, and the author's goal of combining quantitative and qualitative sociolinguistics with corpus linguistics methods.

Chapters 5–7 contain the analyses that form the core of the book. Chapter 5 on Jamaica and chapter 6 on Trinidad are organised in roughly the same way. The first major section of both chapters implements the study's quantitative goal by providing frequency counts for around eight morphological and syntactic variables. Each chapter has a convenient table (Table 5.10 and Table 6.12) summarising the approximate frequency of the Creole variants in several text types (e.g. news broadcasts, class lessons). The reader should note that although roughly the same variables are investigated in both chapters, the chapters are not exact copies of each other because some text types are explored in the T&T corpus that are not explored in the Jamaican one, and the author relies on research on Jamaica for which Trinidad has no equivalent.

Deuber found that for both countries more Creole variants got incorporated as one moved from formal to informal text types. The main difference between the national varieties is that fewer overt Creole forms show up in the Trinidadian conversations than in the Jamaican ones.Footnote ¹ Deuber attributes this to TrinEC having fewer overt Creole forms than JamC. I could find no significant difference, however, in the number of overt Creole variants between JamC and TrinEC. Deuber probably arrived at this conclusion by equating ‘overt Creole features’ with those traditionally labelled ‘basilectal’. This would explain why she claims that locative de is ‘the only overt Creole copula form in Trinidad’ (p. 140), and has only a zero copula in Table 6.12 with NP predicates.Footnote ² The critical finding of chapter 5 is the cline of Creolisms in JamE conversations showing that there is a formality-sensitive hierarchy of how Creole features are incorporated into JamE conversations.

The second major division of chapters 5 and 6 presents the qualitative part of the research that looks at inter- and intratextual variation/codeswitching in four text types. For Jamaica, Deuber found that code switches into JamC are determined less by topic than the attitude of the participants towards the topic and the meaning it holds for social interaction. For Trinidad, there was a cline of incorporation of Creole features from zero or none in news broadcasts to several in conversations.

Chapter 7 provides a detailed analysis of the modals can/could and will/would in the two varieties, and compares the results with British English. Deuber found that with regard to these modals JamE tends to pattern more with BritE than does TrinE. TrinE sometimes uses could in contexts where BritE would have can, and also exhibits greater interchangeability between will and would in particular contexts than JamE or BritE. The difference between TrinE and the others is attributed to influence from TrinEC.

The final chapter discusses the relationship between style and Creole-influenced grammatical features in spoken CarE, the implications of style for the (post)creole continuum model, stylistic variation in identity construction, and the link between standards and style in spoken CarE. The chapter ends with a discussion of these topics in relation to ‘angloversals’.

Now I turn to a critique of some of the issues that stood out for me. The first point has to do with imbalance in the data. Although Deuber does an excellent job in chapters 5–7 of pointing out the similarities and differences between the findings for Jamaica and Trinidad, there remains an imbalance in the number of text types used for each territory. Four are used for Trinidad (p. 71) while only one is used for some of the analysis for Jamaica (p. 79). This imbalance is due to data for some parts of the corpus not being available at the time, although Deuber tries to offset this by incorporating the findings of previous studies (e.g. Sand Reference Sand1999).

The second critique concerns the over-concentration on Creole in the book. Mair (Reference Mair and Leitner1992) and Deuber (Reference Deuber2009b) herself have argued that it is difficult to look at CarE without contending with Creole. There is too much focus on Creole features, however. For example, section 2.1.2 gives a contrastive presentation of several morphological and syntactic features of TrinEC and JamC. Most of the features presented in the section are the Creole variants of the variables dealt with in chapters 5 and 6. Futhermore, the description of the variables in those chapters almost always starts off with the Creole variant(s), and in some sections, there is more discussion of Creole than English (e.g. 5.1.1). This problem arises again in the summary tables where the features presented are those traditionally associated with the respective Creoles. A more neutral treatment would have presented them, at the start and throughout, in terms of variables with English and Creole variants.

With reference to the variables, Deuber points out that ‘These selected aspects do not include all variables which show variation between English and Creole variants in the data, but they do cover the most important Creole features that can be subjected to quantitative analysis’ (p. 80). Given the interactional dynamics of English and English-lexicon Creoles in the Caribbean, it is certainly wise to include variables that ‘show variation between English and Creole variants’; however, focusing only on these variables suggests that this is the only way variation manifests in CarE, and that the incorporation of Creolisms is the only way in which CarEs indicate style. This could have been solved by avoiding the syntactocentric approach that the book uses, and also considering lexical and phonological stylistic variation. For example, the study could have looked at the use of many vs. a lot of, or negative construction forms such as isn't and wasn't vs. their long forms.

Another consequence of the focus on Creole features is that texts that contain only grammatical forms that agree with International Standard English get overlooked. In fact, such texts are rarely quoted in the book. One gets the sense that if the ‘exotic’ features are not present in a sample of speech, it is considered monostylistic, and hence, uninteresting.

Despite these issues, we should not miss the point that this volume represents pioneering work, and like most pioneering work, the terrain is difficult to navigate. Those interested in qualitative variationist sociolinguistics will find the sections on inter- and intratextual variation in chapters 5 and 6 particularly useful. Scholars working on morphosyntactic variation within a quantitative framework will find lots of important data and findings in the quantitative sections. Those looking to marry corpus linguistics and sociolinguistics will find a successful example here. Methodological choices have seriously limited what Deuber set out to accomplish with regard to style in CarE. Nevertheless, she has increased our knowledge of variation in spoken CarE, and the ways in which interaction with English-lexicon Creoles can account for that variation. World Englishes is in her debt.