Language use in performances has traditionally been considered to lie outside the scope of sociolinguistics as it does not square well with the sociolinguistic concept of the ‘vernacular’, the most unmonitored and unconscious language use of a speaker. However, in more recent years, we can observe a shift from investigating the everyday vernacular to analysing ‘the non-everyday and the non-vernacular’ (Bell & Gibson, Reference Bell and Gibson2011: 558). Wilson's 2014 study on language use in classical choral singing in Trinidad and Westphal's 2017 study on linguistic variation on Jamaican radio blend in well with this recent trend in sociolinguistics, while also shifting the focus of attention from Western monolingual settings to the Caribbean, where English coexists with a local creole.

Wilson analyses the attitudes of singers, conductors, and members of the audience towards the use of different accents in classical choral singing in Trinidad and relates these to actual accent variation that can be found in this genre (Chapters 4 & 5). To obtain attitudinal and linguistic data, Wilson adopts a mixed-methods approach by making use of questionnaires, interviews, and participant observation as well as by recording language use in choir rehearsals (Chapter 3).

Her study shows that British English still enjoys considerable prestige in Trinidadian classical choral singing, especially among singers and conductors (p. 101). These practitioners are only tolerant of endonormative forms in local songs, while audience members are also open to ‘Trinidadian English/Creole’ (TE/C) in Western classical songs (pp. 123–4). Apart from accent preferences, Wilson also elicits what singers, conductors, and members of the audience perceive to be difficulties in achieving the target accent in classical choral singing. In this way, she leaves room for speakers' perceptions, which also becomes visible in the many illustrative statements of singers, conductors, and audience members she integrates in her book. The perceived difficulties fall into two categories, i.e. accent-related versus style-related (p. 172). The first category includes features that are influenced by TE/C and are therefore perceived to be an impediment to conforming to the British English norm, e.g. pronouncing thin like tin. In the second category, we can find perceived deviations from the specific pronunciation conventions of classical choral singing, e.g. an unrounded rather than the preferred rounded realisation of the kit vowel.

After discussing practitioners' and audiences' perceptions of language use in singing, Wilson describes actual accent features used by classical choral singers. One of the most interesting findings of her quantitative analysis is that the singers' ideological preference for British English does not always directly translate to actual language use because singers also incorporate local pronunciations in their songs (p. 229). A similar trend towards endonormativity can be found in the corrections of conductors towards TE/C variants (Chapter 6). Based on her analysis of accent features in Trinidadian choral singing, Wilson establishes features of acrolectal and mesolectal TE/C and shows that there is an emergent local standard in Trinidad ‘that can be distinguished both from the metropolitan norm and from the creole norm’ (p. 314). She concludes that ‘Trinidadian choral musicians claim to target SBE [Standard British English] […]. For the most part, however, what they are actually targeting is a perception of British English, and what is in fact the emerging local standard’ (p. 304). Apart from accent features used in singing, Wilson also finds an inventory of phonological features that constitutes what she labels ‘classical choral singing style’ (p. 320), i.e. a specific style of singing which consists of features that are neither part of British English nor part of TE/C. Wilson's data from Tobago, which she did not include in this study, could provide an interesting testing ground for analysing whether the same non-national singing style exists side by side with an acrolectal variety of Tobagonian English/Creole in classical choral singing on the island northeast of Trinidad. With regard to the Trinidadian data of this study, an appendix with linguistic data could have been a very useful addition to the monograph.

Like Wilson, Westphal analyses norm competition in the English-speaking Caribbean in performance data, more specifically in Jamaican radio. Furthermore, both studies share an interest in the intricate relationship between language use and attitudes, and pay attention to their genre- and context-specificity. Westphal investigates two genres of the Jamaican radio landscape in more detail, i.e. radio newscasts (Chapters 6 & 9) and talk shows (Chapters 7 & 9). To get an idea of language variation in these genres, readers of his book are invited to listen to sound files that have been made available online.

Westphal's sophisticated acoustic analysis of accent variation in the genre of newscasts in Chapter 6 reveals a three-way norm competition between exonormatively influenced standard English varieties, which he labels ‘American-influenced English’ (AIE) and ‘British-influenced English’ (BIE), and the endonormative standard variety of Jamaican English. His analysis shows that Jamaican Standard English prevails over BIE and AIE on Jamaican radio (p. 104). This conversely means that Jamaican Creole (‘Patwa’) has not made inroads into the more formal genre of newscasting, an observation that applies equally to the rather formal genre of classical choral singing. Apart from establishing accent-related features, Westphal also identifies features that are typical of newscast speech more generally, which could be labelled ‘newscasting style’ in analogy to Wilson's ‘classical choral singing style’. This style is for example characterised by a clearly noticeable articulation of the last consonant in words such as best (pp. 86–7). In contrast to newscasting, language use in talk shows reveals that Patwa is more strongly represented in this genre (Chapter 7). Talk show hosts use styles that are characterised by different degrees of creoleness to construct their respective host personas (pp. 141–2). Their styles range from the almost exclusive use of Jamaican Standard English to project a neutral expert persona to a consistent blend of Jamaican Standard English and Patwa to portray a multifaceted academic, teacher, and rebel persona.

Apart from actual language use, Westphal also analyses listeners' perceptions of linguistic variation in radio newscasts and talk shows in Chapter 9. To find out about listeners' overt and covert attitudes, he uses a mixed-methods approach that includes direct questioning and folk-linguistic interviews as well as a variety rating study (Chapter 8). His attitudinal study reveals that overt language attitudes towards Jamaican Standard English and Patwa are strongly context-bound. While Jamaican Creole is perceived to be inappropriate for more formal genres such as newscasts or advertisements for banks, it is more strongly accepted in more informal contexts such as comedy programmes or soft drink advertisements (p. 168). In general, he finds that listeners openly express an ideological orientation towards Jamaican Standard English in newscasts and perceive an external standard as inauthentic when asked directly (p. 171). However, as far as covert attitudes are concerned, he observes continued deference towards an external standard and that speakers who use an external model for pronunciation are not associated with the negative Jamaican term of ‘twanging’, i.e. ‘putting on an inauthentic foreign accent’ (p. 145). With regard to talk shows, Westphal finds that speakers who include more Jamaican Creole in their speech rank high on liveliness but low on prestige and that listeners generally evaluate the use of code-switching positively (p. 186). On a more critical note, it must be mentioned that the sample of informants in the questionnaire study including direct questioning and variety rating tasks is largely restricted to students (p. 152). It therefore remains an open question whether their attitudes are generalizable to the broader Jamaican society.

To conclude, Wilson and Westphal clearly break new ground at the interface of sociolinguistics and World Englishes. Synthesising the findings of their studies, at least three important contributions in the contact zone between the two fields can be identified. First, their studies reveal that endonormativity in the English-speaking Caribbean is not only visible in an increase in the use of creole but also in the emergence of a local standard variety. Their studies show a two-fold process of establishing local norms: a more wide-spread use of creole in less formal domains (‘destandardisation’) and a shift to a new local standard in more formal domains, referred to as ‘demotisation’ by Westphal following Mattheier (Reference Mattheier, Mattheier and Radtke1997: 24). Second, they show the importance of integrating speakers' perceptions in studies of World Englishes. Both observe a phenomenon that could be labelled ‘attitudinal conservatism’, a state in which the recognition of a new standard lags behind speakers' actual language use. Third and most importantly, they demonstrate how recent theoretical developments in sociolinguistics such as an agentive notion of style or the concept of indexicality (cf. Silverstein, Reference Silverstein2003) can be successfully applied to studies in the World Englishes paradigm. Their studies therefore illustrate the potential for research at the interface of modern sociolinguistics and World Englishes, to which researchers have paid surprisingly little attention so far (cf. also Sharma, Reference Sharma, Filppula, Klemola and Sharma2017: 232). Schneider (Reference Schneider and Mesthrie2011: 353) states that ‘[f]or sociolinguists of World Englishes, a world to win is waiting out there’ and metaphorically-speaking, Wilson and Westphal have planted their flags on this world with the publications of their sociolinguistic studies on the English-speaking Caribbean.