This book is a very welcome addition to the field of sociolinguistic research on English in Scotland. Over the last couple of years, studies have focused on several sociophonetic aspects of Scottish English in different parts of the country (see, e.g., some of the contributions in Lawson Reference Lawson2014). Most studies are interested in the interplay of Scots and Scottish English described in the well-known model by Aitken (Reference Aitken and Trudgill1984). Schützler takes a different approach in his study of the sociophonetics of Scottish Standard English (SSE) as spoken in Edinburgh. He is not interested in the Scots–SSE continuum, but rather focuses on another continuum, that between SSE and Southern Standard British English (SSBE). Whether or not SSBE carries prestige in Scotland and/or influences SSE has been debated in the literature going back at least as far as Aitken (Reference Aitken, Aitken and McArthur1979: 110–11), but to date none of the claims has been tested empirically.

The introductory chapter provides a brief overview of previous research on Scottish English and then locates the study in the context of Edinburgh. Schützler quite rightly argues that because of the demographic and economic make-up of the city, speakers are overall more prone to contact with speakers from south of the border than in Glasgow and other Scottish cities because they are more likely to be part of one or more mixed Scottish–English social networks (p. 5).

This is followed by a more linguistic introduction and motivation for the study. Schützler looks at three phonetic–phonological variables – (e), (o) and (r) – which may be considered to be the heart of the SSE–SSBE continuum. Scottishness may be indexed by using monophthongal variants of the diphthongs, face and goat in Wells’ (Reference Wells1982) terminology, being consistently rhotic and using more tapped variants in syllable onsets. Pronouncing the vowels with a diphthong, being non-rhotic and using approximants in onset /r/ may be seen as English features. The chapter is rounded off by placing the study within the broader area of sociophonetics and outlining its overall contribution to theoretical (dialect contact) and methodological (advocating the inclusion of more internal factors) frameworks.

Chapter 2 provides a concise, yet well-informed and clear introduction to the concept of SSE in the Scottish linguistic landscape and its status as a ‘double contact variety’ (p. 23), which means that it is part of two linguistic continua – with Scots on the one hand and SSBE on the other. This is followed by a brief summary of the sound inventory of SSE.

Chapter 3 is concerned with the theoretical factors influencing accent variation and change. Schützler positions his study in the dialect contact (Trudgill Reference Trudgill1986) and ‘change-by-accommodation’ (Auer & Hinskens Reference Auer, Hinskens, Auer, Hinskens and Kerswill2005: 336) frameworks and assumes ‘that there is a considerable amount of contact between middle-class speakers of SSE and speakers of SSBE’ (pp. 34–5). This is a central notion of his work, but whether this is really the case in the situation that Schützler describes is arguable. I will return to this point below. The remainder of the chapter summarises the importance of external and internal factors in language variation and change in the context of the present study. The author presents his proposal for a model of SSE–SSBE contact which integrates both internal and external factors. This model serves as an interpretative framework for the current study. Schützler argues that the range of variants available in situations in which speakers of SSE and SSBE come into contact and therefore try to accommodate linguistically is determined to a large degree by the variants each of the speakers has at their disposal. The range of variants is dependent on the social characteristics of the speaker, without going into detail what these characteristics entail. Knowing about the range of variants and being able to use them in specific contexts may open up ‘points of inception for lexical diffusion or the catch-on of socially motivated variation and possibly change’ (pp. 45–6). The model provides a good basis for analysis, however, the description is rather dense and not always straightforward. An example illustrating the main points would have clarified the model's aims.

The fourth chapter, data and methodology, provides the reader with more detailed background information with regard to data collection procedure, the linguistic analysis of the variables and the statistical models used. The study was conducted within a variationist research paradigm and is based on speech recordings of 27 speakers stratified by three social factors and five speech styles.

All speakers are of a middle-class background. They were recruited at the University of Edinburgh and one of Edinburgh's private schools. The concept of social class is very fuzzy and the question of how to group speakers often problematic. In this study, younger speakers are middle class ‘by birth’, many of the older speakers on the other hand had only one middle-class parent (p. 5). This, however, is not considered problematic because – as I mentioned above – these speakers would be part of mainly middle-class social networks. Holding the class variable constant is reasonable, as Schützler is not interested in the Scots–SSE continuum but that between two standard accents. While one might argue that SSE is an option for working-class speakers, SSBE certainly is not. Groups were separated according to two age brackets (17--22 and 40–62) and gender as well as a third variable ‘contact’, referring to ‘exposure to and interaction with speakers of SSBE beyond the sporadic everyday encounters’ (p. 38). He identifies three factors leading to higher exposure: 1. English parents, 2. having studied in England for an extended period of time and 3. having worked in England for a longer period. All three are valid, yet somewhat fuzzy, and not clearly defined. Schützler, for example, does not explain what ‘an extended period’ is. Apart from that, it will probably make a difference if the speaker spent time in England as a young person or an adult. Furthermore, having lived in England does not necessarily mean that the speakers had regular face-to-face communication with SSBE speakers, as is tacitly assumed throughout the book. Most middle-class speakers outside the South-East will not usually speak SSBE, and in the English ‘North’ monophthongal variants in face and goat are by far the most common and widespread amongst speakers of all social backgrounds (Beal Reference Beal, Schneider, Burridge, Kortmann, Mesthrie and Upton2004: 123). What is more, there are only five speakers in total who represent the contact group. Although the problem of little or no analytical power is acknowledged (p. 48), given the theoretical angle of the study with its strong focus on the impact of dialect contact, this is somewhat unsatisfactory.

Five different speech styles (wordlist, reading passage and three types of careful free speech) were elicited, but there was no ‘classic’ sociolinguistic interview in the Labovian sense. As Schützler is mainly interested in more standard speech, this approach is justified. The number of tokens is sufficiently high for all variables to allow for statistical analyses.

The vowel data were analysed acoustically in Praat, while the analysis of the variants of (r) was carried out auditorily. The methodology used for measuring the vowels is described in a very comprehensible way, illustrating the approach very clearly. In the following section, Schützler proposes a new method of vowel transformation and – by extension – vowel normalisation. Taking Critical Band Rate and Bark (Bk) as a starting point for describing psychoacoustic differences between vowels, he suggests that in ‘a second step, a conversion of Bark into octaves relative to 1 Bk is applied’ (p. 56) for which he proposes a unit called ωBk. The advantage is that this scale can be considered to be linear, so that a difference of, say, 0.5 ωBk refers to the same distance between two tokens no matter what their formant frequencies are, which is clearly an advantage over previous methods. Following the transformation, the vowel data are normalised using a formula that combines the advantages of the psychoacoustic bark scale with the sociolinguistic robustness of Nearey's (Reference Nearey1977) method. Unfortunately, there is no guideline as to the scale or magnitude of the values yielded after normalisation, which made the interpretation of the variation patterns unnecessarily more complicated.

The variable (r) was analysed auditorily, as it is of a rather categorical nature. Like most other researchers, Schützler treats it as a different subvariable according to position in the word (onset, coda and linking). Compared to other studies of postvocalic /r/ in Scotland, his categorisation is comparatively basic, as he only differentiates between three broad categories: [ɾ], comprising tapped, flapped and trilled variants; [ɹ], containing approximants and rhoticised vowels; and zero variants for the coda and linking contexts. He does not go into greater detail about the motivation of his classification system. One might argue that the processes underlying /r/-loss in non-SSE varieties of Scottish English are different from those in the present study and that therefore a more fine-grained classification would be beyond the scope of the study, but this is only speculation on my behalf.

The final section of the chapter contains a detailed and well-informed outline of the statistical models employed, a variation of the now widely used mixed-effects modelling which takes into account both categorical factors such as age group or position in the word, but also speaker-based variation.

Chapter 5 gives the reader a short, yet precise outline of the acoustics of diphthongs and the historical development of /e/ and /o/ in English as well as previous studies of the two vowels in the Scottish context. This is followed by the research questions and expected results, which are given in a very readable, comprehensible way. In summary, the external factors of age, gender and style are likely to replicate previous findings and contact is expected to provide a main source of variation. Four internal factors are taken into account: speech rate, lexical frequency, prepausal position and whether the word had diphthongised variants previously in the historical development.

Chapter 6 provides the descriptive and inferential statistical analyses of (e) and (o). The section on the descriptive results is rather short at only four pages, but the variables are discussed in greater detail in the following subsection. The descriptive results are well presented, but as there is no explanation about the scale for the normalised values, the statement that for the F1 of (e) ‘[t]he difference between T1 [onset] and T2 [offset] in this dimension is .232 ωBk, while F2 only changes by .061 ωBk’ (p. 79) left me slightly puzzled. Only by looking at a vowel plot including peripheral vowels (e.g. figure 6.1, p. 80) could I make some sense about these values and their magnitude. The following plots of external factors already give a good impression of the findings, which are overall more uniform for (e), but for both variables they suggest that contact may play a role in using more diphthongal variants.

For readers who are interested in statistical detail the output of the regression models is presented in two large tables. The remainder of the section contains a well-structured analysis of the predicted values for the social, stylistic and external factors.

To summarise the findings, in both vowels there is social stratification in the data along the lines of age (younger speakers use less peripheral variants) and for (o) there is also gender-based variation (females produce higher onsets and offsets); overall the two vowels are rather monophthongal. Contact seems to have a relatively strong effect on a speaker's pronunciation of (e) and (o). Both vowels show longer trajectories with more open onsets. In addition (o) is also more advanced, thus resembling SSBE variants. Stylistic and language-internal factors contribute to ‘low-level micro variation’ (p. 99) leading to an SSE-internal continuum between conservative and innovative speakers. However, these effects do not indicate ongoing Anglicisation, as – at least for (o) – ‘[o]nly the exceptional case of actual, measurable SSBE-contact incites speakers to produce variants of the vowels that more closely resemble SSBE forms’ (p. 99).

In chapter 7 the research context for (r) is presented in great detail and much more comprehensively described than for the vowels. This may be explained by the larger body of research on (r), particularly in the Scottish context, and by the fact that it is linguistically more complex than the vowels. All factors mentioned above for the vowels are also considered here. In addition, prosodic prominence, i.e. whether the /r/ occurred in a stressed syllable or not, preceding and following vowels and position in the word were considered.

Chapter 8 provides a detailed account of the statistical analyses of (r). The results may be summarised as follows. In line with previous research on (r) in Scotland, approximants are the majority variant in all three subvariables. Derhoticised variants are coming in strongly and account for over 30 per cent of all tokens in the coda data. Linking (r) is fairly common. Looking at the results in detail, a more complex picture emerges. For coda and linking (r), the social effects are quite strong. Female speakers more readily adopt vocalised variants overall, with young women strongly leading in the adoption of zero variants for linking (r). The effect of contact is remarkably strong on coda (r). About half of the tokens are vocalised compared to only 25 per cent in the non-contact group. If (r) is articulated, however, no difference is found. Stylistic effects show an overall greater usage of vocalised /r/ in less formal speech styles. Of the language-internal factors, particularly higher speech rate and unstressed position favour derhoticisation, while high lexical frequency overall favours SSE variants. The argumentation is clear and logical.

The final chapter summarises the main findings of the current study and asks the question of whether the SSE–SSBE continuum is fact or fiction. Contact may play a role in the use of features potentially being taken over from SSBE, but overall there is little indication of an ongoing Anglicisation of SSE:

With regard to a potential SSE–SSBE continuum Schützler offers two interpretations: (i) that what he observes are the initial stages of a profound change towards a more SSBE-like standard accent and (ii) that SSE is not being abandoned, but rather enriched by SSBE features, particularly derhoticisation. I agree that the second interpretation is more likely. He adopts Trudgill's (Reference Trudgill1986: 86) view that in a contact situation variants may either be levelled or become more diffuse and result in a continuum with variants typical of the varieties as their poles. This, he argues, should not be seen as an extension of Aitken's (Reference Aitken and Trudgill1984: 519–23) model of Scottish speech, but rather as a continuum of its own as it is only available to a small proportion of the population.

Overall, this study is a remarkable piece of work. The question of how much SSBE or any other form of a non-Scottish accent may influence SSE has been asked by various scholars (e.g. Aitken Reference Aitken, Aitken and McArthur1979; Abercrombie Reference Abercrombie and Abercrombie1991 for opposing views), but no one has so far looked into this empirically. Schützler provides a detailed and comprehensive analysis of three sociophonetic variables which are most prone to being affected by SSBE if we accept that a non-Scottish model may have some impact on linguistic choices of SSE speakers. What I particularly like about this study is that it manages to include a wide range of social, stylistic and internal factors but never loses track of the bigger picture. The study is well integrated into the relevant sociolinguistic and phonetic frameworks and the statistical analysis is very profound. The two problematic points I mentioned above – too few participants in the contact group and the assumption that the English contact variety would almost exclusively be SSBE – cannot diminish the good impression of the book and I definitely recommend it to anyone interested in accent variation in Scotland and dialect contact in general.