Intonation in Romance sets out toward two main goals: to provide data and descriptions of the intonation of a diverse set of nine languages within the Romance family, and to provide a prosodic transcription system and phonological model that can be flexible enough to cover both high-level typological characteristics as well as dialect-specific details. To these two ends, the volume is undoubtedly successful, a product of serious collaboration among its 44 contributors. This in itself is an impressive feat, given how especially susceptible to language-, framework-, and transcription-specific idiosyncrasies intonational research can be. Indeed, the overall consistency in how the nine languages were analyzed makes variation in intonation much easier to grasp, part of why Intonation in Romance is bound to be a highly valuable resource, especially for those who have been developing a typological perspective on prosody.

Between the introductory Chapter 1 and typological Chapter 11, each of the nine internal chapters serves as a descriptive model of a different language in the Romance family, arranged alphabetically. These include languages that have had a long history of intonational study within the Autosegmental-Metrical (AM) Theory of intonational phonology and Tones and Break Indices (ToBI) annotation (e.g. Catalan, Portuguese, Spanish) or other frameworks of intonational analysis (e.g. French, Italian) to languages whose intonation has long been understudied in any modern framework (e.g. Friulian, Occitan, Romanian, Sardinian). Each internal chapter can stand on its own, in that a reader only interested in, say, Friulian intonation can skip directly to Chapter 4 without sacrificing comprehension.

One strong feature of this volume is its high degree of internal consistency. Every chapter adopts the AM Theory of intonational phonology, while still making frequent connections to relevant literature from a variety of other frameworks. Every chapter uses a common ToBI-style system of prosodic annotation, marking pitch accents, boundary tones, and perceived prosodic boundary size in text grids below waveforms and spectrograms over which clear f0 contours are displayed; in some cases this common ToBI system meant that a well-established label from the literature of one language (e.g. L*+>H in Spanish) had to be changed to line up with its use in other Romance languages (e.g. L*+<H). Every chapter is accompanied by a large set of example recordings available on the companion website. And while some of these languages have been studied for decades on the basis of scripted laboratory-style data, every model in this volume is supported by more natural, unscripted recordings using the Discourse Completion Task (DCT) method, where participants are given a passage about a specific context, and then are prompted to respond, thus minimizing the occurrence of unenthusiastic or unnatural reading intonation while also allowing for more variation based on nuanced speech contexts. Above all, the most impressive of these shared qualities across the nine internal chapters is that each chapter delves into regional variation, based on recordings of multiple speakers from multiple locations in multiple dialect areas.

Chapter 2 covers Catalan in remarkable detail, with data from speakers native to 70 cities and towns across Spain, France, and northwest Sardinia. For me, a special highlight of the Catalan chapter was its maps: while every chapter begins with one or two useful maps showing the broad geographical range of the language and its major dialect areas, nothing compares to the 12 detailed maps illustrating the regional variation of each feature of Catalan intonation. The maps very clearly reveal how Catalan dialects vary by intonation in groups that do not align with dialect boundaries determined by segmental processes, e.g. vowel reduction patterns. For example, the highest-level dialect boundary in traditional descriptions of Catalan is the East–West division, yet this line was largely irrelevant in terms of intonation: Central and Northwestern varieties form the tightest dialect grouping by intonation despite being from opposite sides of the East–West split. National boundaries, which arbitrarily cut across traditional dialect groupings but can define where different languages hold prestige over others, seemed to be quite important: dialects of Catalan spoken in France (and thus in contact with Occitan and French) and Italy (and thus in contact with Sardinian and Italian) showed signs of prosodic borrowing from locally dominant languages. Given that this is the first language-specific chapter, these informative maps set the reader up to hope that every subsequent chapter will be covered by this level of cartographic detail; while this hope is currently not met, I urge the contributors to consider incorporating maps like these for all nine languages in future publications. There are only a few errors and omissions that stand out in an otherwise excellent chapter: intermediate phrase (ip) tones and ‘initial tones’ (LHi) are mentioned in the text and appear in a handful of pitch tracks, but are not included in the tonal inventory tables in the summary and conclusion; Figure 2.2 shows L+H* H- L+<H* but the description in the text identifies L+<H* L+<H*; Figure 2.4 shows a L- tone, described in the text as a L%; Figure 2.8 shows a H+L* pitch accent, described in the text as ¡H+L*; and the images for Figure 2.13 (top) and (bottom) are flipped with respect to the caption, the description in the text, and the sound files on the companion webpage.

Chapter 3 focuses on French, which with its lack of lexical stress, its more restricted set of tonal patterns, and its functionally demarcative use of accentual phrases (APs) is arguably the most divergent Romance language in its prosody, more similar to many (non-tonal) languages of South and East Asia (e.g. Bengali, Japanese, Korean, Tamil). In many ways, this makes the French chapter one of the most fascinating in the volume, with the language serving as an outlier with respect to some of the core features of Romance. The high degree of constructive compromise across collaborators to propose a mutually-agreed-upon model and transcription system is somehow more palpable in this chapter, with contributors coming from very different theoretical backgrounds; this is clearly one of the great strengths of the chapter. Compared to the chapters on the two other intercontinental Romance languages (i.e. Portuguese, Spanish), however, it is noticeable that the chapter of French has a more limited geographic scope, restricting its view to Europe, and missing an opportunity to incorporate the many francophone regions of Africa and the Americas. This could partially explain why such little regional variation is found in intonation, even though the authors investigated regions in different countries and in different areas of language contact (e.g. traditionally Occitan- and Arpitan-speaking regions). It is of course likely that other factors are also at play in restricting intonational variation in French, including the exceptionally high prestige of a single standardized Parisian form for the language, and the already restricted set of tonal combinations possible in this AP language.

Chapter 4 is the first comprehensive intonational phonological description of Friulian. Having come with no prior knowledge of the properties of this language, I found the chapter extremely well written, with a clean analysis of the findings. I found it especially interesting that the authors found only minor dialect variation in intonation, and little influence of Slavic or Germanic languages, with the possible exception of the German-like non-vocative uses of the ‘chant’ contour and prosodic features outside intonation (e.g. mora-based tone bearing units, stress-timed rhythm). And while there are some questions left unanswered (e.g. why is there such a dramatic valley between L+H* and H*+L in Figure 4.10 but not in Figure 4.22 or Figure 4.23?), the overall model and analysis are well supported by the examples provided.

Much like the chapter on French, Chapter 5 covers regional variation in Italian by looking within and across traditional linguistic boundaries in Italy, in this case examining whether there could be macroregions of intonational variation based on the local vernacular language (e.g. Neapolitan, Piedmontese). Also much like the French chapter, the authors unexpectedly do not find evidence of such macroregions (e.g. some Northern varieties look more like Extreme Southern varieties than they look like neighboring dialects), but unlike French, they do find a great deal of city- and speaker-specific variation. This could reveal a greater set of possible contours available to speakers, or it could suggest that some of the transcriptions and analyses should be more closely compared. The authors do acknowledge that ‘in some cases a particular analysis has been chosen after auditory and [f0] inspection of a number of examples – not necessarily after systematic phonetic and phonological investigations’ (p. 164); thus it is possible that with continued work, clearer patterns will emerge. Overall, I found this chapter somewhat less transparent in its presentation, with lengthy footnotes and text-heavy tables, but the data are promising and I look forward to seeing continued work on this project.

The diverse varieties of Occitan are covered in Chapter 6, which despite its more preliminary status as a comprehensive intonational description, is one of the best written chapters. Occitan is unique in this set of languages in that it is spoken across a fairly large territory spanning three countries, but has been in a state of decline and low status for centuries. This could help explain why the various dialects of Occitan appear to be so especially prone to intonational transfer from locally dominant contact languages: Aranese and Cisalpine (both spoken outside France) have signs of borrowing of delayed prenuclear pitch accents, bitonal boundary tones, and the vocative chant, presumably from Catalan, Spanish, and Italian. Conversely, Occitan varieties in France show at least some evidence for APs as the basic tonal unit, much like French. It should be noted though that unlike French, where APs are obligatorily marked on both edges (except in cases of tonal overlap), the putative APs of Occitan have obligatory marking in the pitch accent only, with the left boundaries only occasionally marked. Without overt AP boundary tones, it may be that unquestionable evidence for more prototypical APs (à la French or Bengali) remains elusive; this question is certainly worth pursuing further.

Chapter 7 expands work mostly concentrated on the Lisbon variety to cover an array of Portuguese dialects in both Portugal and Brazil, finding notable intonational differences both within and across the two countries. Despite the long history of studying the intonation of Standard European (i.e. Lisbon) Portuguese, this is the first comprehensive look at Portuguese as spoken on both sides of the Atlantic. Overall, the work is impressive and far-reaching, finding considerable variation in pitch accent distribution and tune–meaning mappings across the two countries. There are a few somewhat minor gaps: the audio quality of the Brazilian Portuguese examples does not match that of the European Portuguese examples or of the other languages in the volume, and the text of example sentences is provided only in orthography (and not in IPA) in the text grids. I was intrigued by what the authors call an ‘additional tonal event . . . in pretonic syllables in longer words’ (p. 244), found in statements and questions with narrow focus in Brazilian Portuguese, e.g. in Figure 7.7 (marked with the uncertainty diacritic ‘?’), Figure 7.18, and Figure 7.19. Given that these rises occur on lexically unstressed syllables, should they be considered leading tones for the following pitch accent (e.g. L+H+L*+H), or word/phrase-initial tones optionally used in narrow focus independent of pitch accent placement and choice (e.g. L+Hi . . . L*+H)? The authors do not expand on this in the chapter, and the short example words under narrow focus do not allow us to disambiguate between these two possibilities. I found myself wondering whether this feature is a characteristic that separates Brazilian and European Portuguese, and how the feature should be analyzed in a model that proposes only pitch accents and boundary tones. It would be especially useful to compare these pretonic rising contours to similar phenomena found in Catalan and explicitly incorporated into the models for French, Occitan, and Sardinian. The authors note that their analysis is still in progress, based on a subset of the data collected, so I look forward to seeing more findings in the near future.

The model of Romanian in Chapter 8 is more preliminary relative to the other chapters, being based on only four speakers from two locations. Particularly notable is the claim that the nuclear accent in yes/no questions is borne on the verb even when not ip/IP-final, argued to be a borrowing from Slavic languages. I do find, however, that the data provided to support the claim seem to show a salient pitch accent on the post-verbal object (e.g. Figures 8.12–8.15), in line with other Romance languages; furthermore, the fact that the main intonational difference in yes/no questions between Transylvanian and Moldovan dialects lies entirely in the contour spanning the object suggests to me that it is the object that bears the nuclear contour. Direct comparisons with corresponding declaratives (where the object is expected to bear the nuclear accent) would help in this regard. More examples would also help demonstrate the distinction between H* . . . L% and H*+L . . . L% (e.g. Figure 8.8), and support the positing of ip boundary tones in Figures 8.9– 8.11. I noticed that some of the examples appear to have unlabelled tonal targets, e.g. Figure 8.17, where the high targets between the L* accents are unaccounted for in the transcription, and only referred to in the text as ‘rhythmic’; could these be leading (H+L*) or trailing (L*+H) tonal components of the pitch accents? Of course, given that this is my first exposure, I admit that many of these questions could stem from my unfamiliarity with Romanian prosody.

Chapter 9 compares the Logudorese and Campidanese varieties of Sardinian, which show some unique pitch accent types identified by the authors as extra-high ¡H+L* and ¡H*+L, and very strong cues for secondary stress, which lead the authors to explicitly incorporate pitch accents for secondary stress in certain cases. The positing of ¡H+L* appears to be especially useful in labeling what otherwise sounds like a typologically atypical strategy for marking yes/no questions, i.e. higher pitch on the prenuclear syllable. I am curious if it would further strengthen the model to show more declarative examples with clear H+L* accents, and to provide more information for how to properly distinguish an extra-high ¡H target from its plain H counterpart. One can imagine the nuclei of some declarative and imperative examples reanalyzed to have a monotonal L* (e.g. Figure 9.2, Figure 9.3) or downstepped !H+L* (e.g. Figure 9.15), thus allowing for the ¡H+L* to be reanalyzed as H+L* in contrast with L* and !H+L*; further examples might disambiguate these possibilities. Overall, I was impressed by the level of detail given to those features that separate Sardinian from its sisters, and to the thought that went into devising a practical orthography that could be applied across dialects.

Chapter 10 analyzes the variation across the most widely spoken Romance language, Spanish, and draws from a massive corpus of DCT-style recordings from across the Spanish-speaking world: the Interactive Atlas of Spanish Intonation (AIEE) and chapters of Transcription of Intonation of the Spanish Language (TISL). I noticed that unlike those of the other chapters, the authors of this chapter took extra care to not jump to conclusions, reminding the reader that just because recordings from two different locations have different pitch contours, that does not automatically argue for a dialect difference; there could be stylistic, syntactic, or pragmatic factors at play. Partially for this reason, the bulk of the examples in the chapter are produced by the first author (Hualde) in his native Peninsular dialect rather than being taken directly from the TISL/AIEE corpora; the authors acknowledge that their chapter effectively overemphasizes European Spanish (pp. 388, 390), and the choice does separate Chapter 10 from the other chapters, which relied almost entirely on naturalistic DCT-elicited data. The authors do make significant mention of Latin American varieties, largely drawing from the findings of TISL, and at least a handful of example pitch tracks from AIEE are included in the chapter's recordings. Still, even when the authors do direct the reader to TISL examples by page and figure number, this only points to images and text; to listen to the example, one must then navigate the AIEE database, which could be one step too many for some readers. Still, I would be remiss if I did not add that the chapter presents a convincing, carefully presented model, is written in a very clear style, and includes many useful links to findings established in TISL, AIEE, and other sources.

The final chapter takes a typological perspective to investigate what prosodic traits are generally shared across the Romance family, and which characteristics are more language-specific. The typical Romance language can be seen as having lexical stress (exception: French) on one of the last three syllables (exception: Occitan). Almost every content word is marked on its stressed syllable with a pitch accent (exception: Standard European Portuguese), creating strong macro-rhythmic properties. Many Romance languages also bear pitch accents on secondary stresses (although different chapters covered these to different extents). The strongest (nuclear) pitch accent is ip/IP-final (exception: Romanian yes/no questions, Sardinian short wh-questions). There are both ip boundary tones (exception: European Portuguese) and IP boundary tones marking the right edges of prosodic units. The authors argue that when considered along with syntactic and information structure properties, these facts distinguish Romance languages from West Germanic languages, which tend to have sparser pitch accent assignment and longer postnuclear contours. The authors go on to discuss the frequencies and uses of different tonal combinations, morphological and syntactic properties that interact with prosody, and the geographic extents of these features. Again, most interesting is the level of prosodic borrowing apparent in the maps in areas of language contact.

This volume is not written for those who are looking for a primer on the phonology or transcription of intonation. The reader should come with a knowledge of AM Theory's assumptions and constraints and literacy in ToBI annotation style; do not expect to have a pitch track and spectrogram explained, do not expect to see pitch accent, boundary tone, or downstep defined, and do not expect to be told how symbols such as ‘%’, ‘*’, or ‘<’ are used. In that sense, this volume is not recommended as a first reading in intonation, but I do not see this as a problem, given the substantial literature available for those looking at prosody for the first time. I also noted that the authors do not make any assumptions about the reader's expertise in the specific intonational properties or general linguistic structure of any individual language (although familiarity with common Romance roots can help one navigate the occasional loosely glossed example, e.g. Sunt vènnitos tres òmines ‘Three men came’), and they are careful to formally introduce even basic terms and concepts from pragmatics and information structure; it is refreshing to see the authors explicitly define echo question, broad focus, vocative, epistemic bias, etc., in practically every chapter. Overall, I found this volume to be highly impressive in its breadth, depth, and consistency. It will surely serve as a go-to source for the intonation of Romance languages, whether the (prosodically-informed) reader is looking for a single model of a language overall, a detailed description of regional variation, or a typological analysis across languages.