1 Overview

The Oxford handbook of language prosody, edited by Carlos Gussenhoven and Aoju Chen, is a collection comprised of 49 chapters divided into eight parts, with contributions from 121 researchers and nearly 700 pages of content. As these numbers might suggest, this volume is expansive. In introducing the handbook, the editors note that volumes focusing on one aspect of prosody (e.g. intonational phonology) are more common than those that cover the field as a whole. Unlike a volume which offers an introduction to a particular domain of prosody research or a particular model, this handbook reads as a catalogue of the current state of knowledge in the field, as described by leading prosody researchers. As such, theoretical or methodological cohesion is not a goal, though many chapters share a theoretical and methodological approach, described below.

As an overview of the field, the handbook is certainly a success. It offers, among other things, a well thought-out and wide-ranging compilation of prosody research, which spans topics such as the role of prosody in first language acquisition, second language instruction and automated speech processing.

The eight parts of the handbook vary in size. The largest, ‘Prosody across the world’, contains 18 chapters surveying prosodic typology; the smallest, ‘Fundamentals of language prosody’, contains just two chapters. Chapters within a section can also be only very broadly related, for example the ‘Prosody and language processing’ section contains three contributions on very different topics: ‘Cortical and subcortical processing of linguistic pitch patterns’ (Joseph C. Y. Lau, Zilong Xie, Bharath Chandrasekaran & Patrick C. M. Wong), ‘Prosody and spoken-word recognition’ (James M. McQueen & Laura Dilley) and ‘The role of phrase-level prosody in speech production planning’ (Stefanie Shattuck-Hufnagel). The focus of the volume, if any, comes across as typology (‘Prosody across the world’ is by far the largest part), though given the overview-oriented structure, the reader does not get the sense that there is an intended focus on one particular aspect of prosody research.

The production quality of the handbook is good overall. Typographical errors, though present, are few and far between. Figures and visualisations are also generally well-placed, though they vary widely in their detail and quality, even sometimes within a chapter. However, given the number of contributors, this might be expected, and does not seriously detract from the content. One clear strength in the production of the handbook is the well-constructed index, which is split to cover both languages and topics; this will certainly help readers navigate and explore the volume.

2 Framing and scope

The question of how prosody is framed in any collection is non-trivial, but even more crucial in such a wide-ranging volume. The editors offer a guiding conception of prosody in the introduction (Chapter 1). In framing the domain of inquiry, they state: ‘we consider the core prosodic elements to be tone, stress, prosodic constituents, and intonation’ (p. 4). These are taken to constitute the phonological representations of language prosody, and serve to delimit the scope of the handbook. Each of these core elements receives broad coverage.

The treatment of prosodic constituents and intonation overall most closely fits with the field of Intonational Phonology (see e.g. Ladd Reference Ladd2008). As such, much of the content reflects what has been called the ‘intonational approach’ to modelling prosodic structure (Jun Reference Jun1998): the identification of prosodic constituents and heads on the basis of cues in an utterance, with a common focus on the intonational tune. The Prosodic Phonology approach to modelling prosody, which tends to focus more on the relation of prosodic domains, metrical structure and syntactic structure (exemplified by Nespor & Vogel Reference Nespor and Vogel1986), is less represented. Nevertheless, insights from Prosodic Phonology are referenced throughout the volume, notably in questions of stress assignment and the relation of prosodic structure to syntactic structure.

An investigation of prosody which follows the ‘intonational approach’ often includes detailed consideration of cues to identify prosodic domains and prominences, and thus must engage with the question of what those cues are. This too is non-trivial. As described in the handbook's introduction, the classic definition of prosody as ‘suprasegmental’ phonetic parameters (pitch, duration and amplitude) has guided much of prosody research historically, and is still pervasive in the present day. In this definition, suprasegmentals are ‘overlaid’ on segments, with segments themselves parameterised phonetically in their ‘spectral pattern’. There are of course well-known issues with this definition of prosody as suprasegmental parameters, as noted by the editors. For example, in order to be perceptible, any segment must have some duration (Lehiste Reference Lehiste1970). Further, various non-suprasegmental phonetic features encode prosodic categories and organisation, for example formant structure and spectral variations signalling voice quality (e.g. Redi & Shattuck-Hufnagel Reference Redi and Shattuck-Hufnagel2001, Georgeton et al. Reference Georgeton, Kocjančič Antolík and Fougeron2016). The complexity of the phonetic substance recruited in signalling prosody has motivated other definitions, which have been called ‘structural’ (e.g. Beckman et al. Reference Beckman, Edwards, Fletcher, Docherty and Robert Ladd1992). As stated by Cho (Reference Cho2016: 122):

Though such a structural view of prosody is certainly mainstream in the field and espoused throughout the handbook, as noted by Fletcher (Reference Fletcher, Hardcastle, Laver and Gibbon2010), suprasegmental parameters are still prioritised in prosodic modelling and description in general. In line with this, the handbook's contributions tend to focus on these parameters, though to varying extents. The reader will also find clear critiques of the suprasegmental understanding of prosody, for example in McQueen & Dilley's chapter, which highlights how traditionally defined ‘segmental’ cues like voice onset time in stops simultaneously signal segmental and prosodic information via domain-initial strengthening (Keating Reference Keating, Harrington and Tabain2006, Cho et al. Reference Cho, McQueen and Cox2007). This particular chapter more generally emphasises the necessary interrelatedness of segmental and prosodic structure as they are manifested in the speech signal, problematising the suprasegmentals view. Thus, much like the field as a whole, a structural view of prosody is prevalent in the handbook (and an alternative suprasegmental definition is rebutted in several chapters), though pitch, duration and amplitude remain central as phonetic parameters used in describing and modelling prosodic structure.

3 Fundamental topics: Parts I, II and III

To examine the diversity of contributions in the handbook, I outline Parts I, II and III, which exemplify dimensions along which prosody research varies.

Consider first the extent to which prosody research is tied to continuous phonetic data (e.g. acoustic or articulatory data), as compared to phonological or symbolic data. The handbook offers a full range of contributions in this regard. Various chapters present descriptions and analyses which are primarily (or only) symbolic, employing autosegmental tonal representations (e.g. ‘Tone systems’; Larry M. Hyman & William R. Leben), symbolic metrical and stress representations (e.g. ‘Stress, meter, and text-setting’; Paul Kiparsky) and optimality-theoretic analysis (‘Prosodic morphology’; John J. McCarthy). Other chapters offer a tighter connection to phonetic data, some of which deal with issues that are typically less of a concern in phonological analysis or modelling, for example ‘Articulatory measures of prosody’ (Taehong Cho & Doris Mücke) and ‘Fundamental aspects in the perception of f0’ (Oliver Niebuhr, Henning Reetz, Jonathan Barnes & Alan C. L. Yu). The latter two chapters constitute Part I of the book, ‘Fundamentals of language prosody’, and both present thorough overviews, especially impressive given the complexity of the topics. Both stand alone as well-constructed introductions, though connections to information presented in this section are mostly absent in the sections that follow (for example, chapters that deal with spoken language prosody are essentially concerned with acoustic data, such that information from Cho & Mücke's chapter lacks much connection to subsequent chapters). Other chapters in the handbook are also more focused on phonetics, most obviously chapters in Part III, ‘Prosody in speech production’, which contains ‘Phonetic variation in tone and intonation systems’ (Jonathan Barnes, Hansjörg Mixdorff & Oliver Niebuhr), ‘Phonetic correlates of word and sentence stress’ (Vincent J. van Heuven & Alice Turk) and ‘Speech rhythm and timing’ (Laurence White & Zofia Malisz). Moreover, many chapters in the handbook offer both a description of phonetic facts and their link to phonology, in line with long-standing traditions in intonational phonology research, which are often characterised by a dual focus on phonological representation and phonetic implementation. In summary, as one might expect given the breadth of this volume, readers with research interests in symbolic/phonological approaches to prosody research and in the phonetics of prosody, or both, will find chapters of interest in the handbook.

Another distinction in prosody research is the degree of focus on what is commonly called lexical prosody (e.g. lexical tone, word stress, lexical pitch accent) vs. postlexical prosody (e.g. phrasing, phrasal prominence, intonation). Part II of the handbook, ‘Prosody and linguistic structure’, offers an illustration of its breadth in this regard. The reader will find chapters on ‘Tone systems’ (Larry M. Hyman & William R. Leben) and ‘Word-stress systems’ (Matthew K. Gordon & Harry van der Hulst), which each offer well-constructed overviews of these facets of prosody, with a focus on word-level phenomena. The subsequent chapter, ‘The Autosegmental-Metrical theory of intonational phonology’ (Amalia Arvaniti & Janet Fletcher) gives a thorough examination of the Autosegmental-Metrical (henceforth AM) theory, and deals primarily with postlexical intonational questions. These three chapters thus serve well to orient the reader to a fundamental understanding of domains which are relevant in many subsequent sections of the handbook, particularly Part IV, ‘Prosody across the world’, where the AM-theoretic accounts are plentiful, and where concepts related to tone and word-stress systems are also relevant. The remaining chapters of Part II are ‘Prosodic morphology’ (McCarthy) and ‘Sign language prosody’ (Wendy Sandler, Diane Lillo-Martin, Svetlana Dachkovsky & Ronice Müller de Quadros). Unlike the other chapters in Part II, these chapters do not connect to much of the material which follows in the handbook; however, they stand alone as excellent introductory overviews.

As one might expect given the breadth of the volume, at times chapters in the handbook can feel at odds with one another. One case of this is the handbook's engagement with rhythm classes and ‘rhythm metrics’ (e.g. measures developed to distinguish stress-timed vs. syllable-timed languages). White & Malisz's chapter provides a history of the rhythm class hypothesis, and foregrounds research which argues against it, discussing for example, studies finding comparable variation within a rhythm class rather than across classes based on speech rate and style (Loukina et al. Reference Loukina, Kochanski, Rosner, Keane and Shih2011, Arvaniti Reference Arvaniti2012). The chapter as a whole, as the authors state, strongly questions ‘the validity and usefulness of categorical rhythmic distinctions’ (p. 171) and metrics designed to quantify them. In contrast, other contributions in the volume hew more closely to the notion of rhythm classes, or assume their validity implicitly, in for example discussing how language acquisition and infant- and child-directed speech relates to a language's rhythm class (Chapters 39–41). Given the critical view outlined in the chapter introducing speech rhythm, the reader may feel a certain tension between the approaches in these contributions. Another domain that exemplifies differences in approach across chapters relates to the theoretical treatment of pitch movements. Going back to foundational work such as Pierrehumbert (Reference Pierrehumbert1980), models of intonation often adopt a ‘target and interpolation’ approach to phonetic implementation. As summarised in Féry (Reference Féry2016: 1):

As is common practice in intonational analysis, tonal representations are generally posited as the targets in an F0 contour, and indeed a target-and-interpolation model of intonation is implicit in many contributions throughout the book. However, elsewhere in the handbook some critiques of a strictly target-and-interpolation view are given, mostly in Barnes et al.'s chapter, which provides an overview of how various phonetic aspects in F0 implementation are important in perceived target alignment and scaling. These include contour convexity/concavity, the more general concept of Tonal Centre of Gravity (e.g. Barnes et al. Reference Barnes, Veilleux, Brugos and Shattuck-Hufnagel2012, Barnes et al. Reference Barnes, Brugos, Veilleux and Shattuck-Hufnagel2021), and the intensity of the speech signal during an F0 movement, all of which question the idea that transitions between targets are only interpolative, and are unimportant for phonological analysis. As the authors of that chapter state, this ‘makes clear the need to view individual aspects of the f0 contour as part of a larger constellation of cues working together to realize the contrasting categories of phonological systems’ (p. 144). This chapter details convincing evidence to this effect, and thus offers a valuable perspective that orients the reader to a more holistic understanding of F0 contours and their relation to tonal structure than the target-and-interpolation model that is implicit in other contributions in the handbook.

Of course, differences of the sort described above are only natural and healthy in the development of theory and the field, and fit with a conception of the handbook as a survey of the current state of prosody research.

4 Typology: Part IV

With the aforementioned sections treating what might be seen as foundational topics in language prosody, the remainder of the handbook delves into applications of prosody research in typology, language acquisition and technology. Part IV, ‘Prosody across the world’, provides a typological overview, with 18 geographically organised chapters, each describing what is known about the prosodic systems of languages that lie within a particular region. Some of the variety in these contributions boils down to their engagement with particular aspects of prosody (tone, stress, intonation), in line with the geographical distribution of these features and research traditions in the field. For example, ‘Sub-Saharan Africa’ (Larry M. Hyman, Hannah Sande, Florian Lionnet, Nicholas Rolle & Emily Clem) has a focus on lexical and grammatical tone, ‘North Africa and the Middle East’ (Sam Hellmuth & Mary Pearce) mostly examines stress and tone and ‘Southern Europe’ (Mariapaola D'Imperio, Barbara Gili Fivela, Mary Baltazani, Brechtje Post & Alexandra Vella) is primarily dedicated to intonational phonology. Beyond this, contributions in this part of the handbook vary in depth and breadth of coverage. For example ‘Asian Pacific rim’ (Sun-Ah Jun & Haruo Kubozono) overviews the prosodic systems of two languages, Korean and Japanese. This particular chapter offers a targeted and in-depth look at both languages’ prosodic systems, while also describing dialectal variation. Chapters like this will be of great use to researchers who want a fairly deep introduction to a given language. In comparison, chapters like Hyman et al.'s ‘Sub-Saharan Africa’ and ‘North America’ (Gabriela Caballero & Matthew K. Gordon) are much broader, outlining the prosodic features of a great many languages. Chapters of this sort in Part IV tend to be organised by linguistic feature (e.g. stress, tone), with representative examples from various languages given to exemplify the patterns which are discussed. ‘North America’, for example, provides an excellent overview of the typology of stress, weight-sensitivity and tone in North American languages, and will be a great resource for readers who want to understand an overall picture of what is known about prosodic features in the region. Both the more targeted and the broader descriptions are valid and well thought-out approaches to overviewing prosody research (and some chapters fall in between the extremes described above). However, readers who are looking specifically for depth or breadth in this section will find that the extent to which chapters meet these criteria varies.

The plate section of the handbook contains a map illustrating which geographical regions are described in which chapters; this will certainly help with navigating Part IV, and point readers with specific regional interests to the correct chapter. Nevertheless, one consequence of the geographical organisation of this section is that researchers who are interested in surveying a particular linguistic feature (e.g. pitch accents, stress systems) will need to look through and collate information from relevant chapters.

Jun's (Reference Jun2005, Reference Jun2014) volumes on prosodic typology, which also examine prosodic systems with a typological orientation, provide a relevant comparison. Unlike these volumes, in which contributions are more uniform and share a clear theoretical orientation, Chapter 4, like the handbook as a whole, offers an answer to the question: ‘what is the current state of scholarship in the study of prosody?’. As such, it reads as a broad synopsis of cross-linguistic research, and, unlike Jun's collections, does not seek to describe a particular set of languages in depth.

5 Other sections and chapters

The additional sections of the handbook present applications of prosody research in a variety of domains: communication (Part V), speech processing (Part VI), acquisition (Part VII) and technology and the arts (Part VIII). Part VII, ‘Prosody and language acquisition’, contains eight chapters, the largest number of chapters outside of the typological section. As such, acquisition is treated in relatively more depth than other topics, with contributions describing the acquisition of word-level and phrase-level prosody, prosody in infant-/child-directed speech, and prosody in second language acquisition and instruction. For any prosody researcher who wants to learn about acquisition, or any acquisition researcher who wants to learn about prosody, this particular section of the handbook offers a multifaceted and fairly thorough overview.

The other parts of the handbook, though certainly more varied in their composition, contain valuable introductions to a given topic, providing a high-level overview of, for example, ‘Tone–melody matching in tone-language singing’ (D. Robert Ladd & James Kirby), ‘Automatic prosody labelling and assessment’ (Andrew Rosenberg & Mark Hasegawa-Johnson) and ‘Visual prosody across cultures’ (Marc Swerts & Emiel Krahmer). I envision sections such as these as being valuable reading materials for a student or researcher who is mostly unfamiliar with a topic and wants to understand fundamental concepts therein.