The eight chapters in this book analyse the various ways of measuring the rhythm of speech, and it presents the results of applying a range of metrics to a comparison of the rhythm of Indian and British English. Chapter 1 is an introduction, and chapter 2 provides background material on Indian English. Chapter 3 describes the concept of rhythm and the different ways of measuring it, while chapter 4 provides an overview of rhythm in varieties of English. Chapter 5 introduces the data and methods of analysis for the current study, and then chapter 6 presents the results for all of the different metrics when comparing the rhythm of Indian and British English. It is assumed that these two varieties occupy different places on the stress-/syllable-timed continuum, so it is important to evaluate which metrics most effectively separate them out. Chapter 7 describes two perceptual experiments, one involving accent identification after re-synthesis of various features of the speech and the other considering the ability of listeners to understand instructions which are masked by other speech (the cocktail party effect). Finally, chapter 8 provides an overview of the findings and a summary of the implications of the research. The two key chapters in this book, each about fifty pages long, are chapter 3, which carefully analyses the various metrics for measuring rhythm, and chapter 6, in which the results of applying the metrics under a range of conditions are presented in considerable detail.

One unusual feature of this book is that the list of references for each chapter is presented at the end of the chapter rather than being collated at the end of the book. While this is normal practice in edited volumes, as it allows individual chapters to be distributed independently (and it also allows the publisher to sell them separately), it is somewhat unusual in a monograph such as this, and it does not facilitate finding the references, as one has to look for the end of the chapter. Furthermore, it is hard to see that many of the chapters will be distributed independently of the rest of the book. While chapter 3 might be useful on its own, and also maybe chapter 7, it is hard to see how chapter 6 makes any sense without a description of the data (presented in chapter 5), or indeed how the comparison of the results for different metrics (chapter 6) is meaningful in the absence of the detailed presentation of those metrics (chapter 3).

But this is a minor quibble. The book is packed full of exceptionally valuable material on different metrics for the measurement of rhythm based on the duration of vowels, consonants and syllables, as well as other acoustic variables such as intensity, fundamental frequency and sonority, and the results analyse in admirable detail the effects of the minimum number of syllables, the inclusion or exclusion of the final syllable, and the use of read speech or conversational speech.

Inevitably, there are places where I would have liked to know more. For example, on p. 141, it is stated that the frequency of creaky voice causes problems for voicing-based metrics. How frequent is creak in the data? And in many places (e.g. pp. 93, 98 and 157), it is noted that variability in consonant durations is greater in Indian English than in British English, and this seems counterintuitive, as one would expect greater variation in a stress-timed language such as British English. So why does it occur? On p. 56, it is suggested that it may be linked with proficiency, but why would this happen if all the speakers are highly proficient, as is claimed?

In addition, in many cases further details about implementation of the metrics would have been helpful. For instance, in measuring vowel durations, it is stated that ‘[b]oundaries between vowels and approximants were placed at the midpoint of change in formant structure between two phonemes’ (p. 106). But what happens if there is L-vocalisation, so a final /l/ in the coda of a word such as cold is pronounced as a vowel, or sometimes completely deleted? How should that be dealt with? Should we conclude that there is no /l/ in the word, so we measure everything from the /k/ to the /d/ as a vowel? And what should we do when there are two consecutive vowels with similar quality, as in the phrase more of with no linking [r]? How is the boundary between the two vowels determined in this case? In fact, a huge number of issues such as these arise in the measurement of rhythm, and while this book provides lots of splendid material and guidance, there still remain many things that need to be resolved. For example, how is syllabification determined in uncertain cases, such as towards and fire (both of which may be one syllable or two)? On what basis do we make a decision and avoid a huge element of subjectivity? Although this book provides lots of exceptionally important material about the measurement of speech rhythm, there remain many issues that are unresolved.

Despite the impressive detail and meticulous presentation in this book, there are a few flaws with the phonetics. On p. 23, the vowels of Indian English are presented, but the start vowel is missing, and this must surely be an oversight. Furthermore, I wonder about showing all transcription in phonemic slashes. For example, on p. 24, the unaspirated allophones of /p, t, k/ are shown as /p˭, t˭, k˭/, but if they are allophones, surely they must be shown in phonetic square brackets [] not phonemic slashes //? And on p. 47, it is stated that, as a result of the elision of the vowel in the second syllable, /bɒtəl/ may actually be pronounced as /bɒt/, but if this represents actual pronunciation rather than the underlying representation, shouldn't it be [bɒt]?

In addition, there are occasional flaws in the text. For example, on p. 138, in discussing measurements of sonority for representing rhythm, it is stated that ‘alveolar and fricative /r/ occupy opposing extremes’, and presumably alveolar should be approximant. Then, on p. 146, when comparing the use of loudness and duration in IndE (Indian English) and BrE (British English), it is stated that ‘BrE uses these cues for prominence more often together than BrE’ – one assumes that one of these tokens of BrE should be IndE.

Despite a few errors such as these, the book provides a massive amount of invaluable data about the measurement of rhythm, and it will constitute a vitally important text for all scholars and students involved in this area of research. I just wish it had been twice as long, so that a few more of the contentious issues could have been elaborated in more detail. However, that does not detract from the impressive quantity of excellent material that is presented in this book, which is certain to become essential reading for everyone involved in research on the rhythm of speech.