Editors Ivano Caponigro (UC San Diego), Harold Torrence (UCLA), and Roberto Zavala Maldonado (CIESAS-Sureste) have gathered 14 essays, written by leading specialists in linguistic fieldwork, syntax, and semantics, together with an editorial introduction, to provide a much-anticipated in-depth discussion of headless relative clauses and related constructions in languages of Mesoamerica. Each chapter outlines the general patterns found in a particular language of the region and identifies the cases of variation within each language family or group. The book is a brilliant compendium of reliable new data on relative clauses. It is also of great use as a methodology guide for future studies dedicated to examining a particular phenomenon in a group of geographically and genetically related understudied languages.
Any review of such a lengthy volume must be selective. For reasons of space, I will not assess each of the chapters separately; instead, in what follows I will begin by briefly discussing the goal of the project, its main findings, and the general organization of the volume, then proceed by mentioning several questions raised by the presented empirical observations that could be addressed by follow-up research inspired by this book.
The volume is the result of the collaboration between the participants of a seminar and two workshops on headless relative clauses organized by Zavala Maldonado, Caponigro, and Torrence in 2016–2018. The main goal of the project, as stated by the authors, was to fill in the gap in the study of headless relative clauses in Mesoamerican languages to achieve ‘a better understanding of what is general and what is language-specific in the morphosyntactic and semantic behavior’ of such constructions and related expressions (27). The book opens with a must-read Introduction by Ivano Caponigro, ‘Introducing headless relative clauses and the findings from Mesoamerican languages’, which sets the stage and acquaints the reader with the three types of relative constructions considered in the later chapters – namely, free relative clauses (introduced by a wh-item only), light-headed relative clauses (introduced by a determiner or a demonstrative) and super-free relative clauses (lacking a wh-expression and a D-type element) – and presents a single template that was used to describe each of the languages. Thus, the Introduction becomes a companion to the other chapters, which all use the terminology consistently without re-introducing it and discuss the data in a uniform way.
The Introduction summarizes the main findings from each chapter. Some of the authors already draw cross-linguistic parallels; see, for instance, the section on Mayan languages, where Vázquez Álvarez and Coon (Ch’ol), AnderBois and Chan Dzul (Yucatec Maya), and Polian and Aissen (Tseltalan languages) mention observations made by each other. Yet it is in the Introduction where the differences and similarities between the languages are brought together in the form of several annotated tables, including the main table about the types of headless relative clauses attested in the languages and the complementary tables showing the distribution of wh-expressions in different constructions.
I would like to mention some of the comparative findings that contribute to the discussion of relative clauses across the world’s languages. Headless relative clauses are found in all languages under consideration, but even closely related languages may differ significantly in having or lacking a particular type of construction, such as free-choice or super-free relative clauses. For instance, an interesting point of micro-variation can be observed in the Uto-Aztecan family and among Oto-Manguean languages, where there are two ‘outliers’, that is, languages without super-free relatives: Southeastern Tepehuan and Acazulco Otomi. A great variety of strategies is used to build a relative clause and several patterns common in Mesoamerica are rarely attested in better studied languages. For example, maximal free relative clauses are often introduced by complex wh-phrases without a special morpheme (see Cinque Reference Cinque2017; Caponigro Reference Caponigro2019 on parallel examples in English and Italian, among others). Existential free relatives turn out to allow finite/completive/past tense-aspect-mood marking, challenging the generalization put forward by Grosu (Reference Grosu and Tomić2004, Reference Grosu, Dobrovie-Sorin and Giurgea2013) and Šimík (Reference Šimík2011), whereby such clauses are universally non-finite/subjunctive/incompletive. The light-headed relatives can be introduced by all kinds of D-elements, which determines their DP-like semantic behavior. The super-free relatives receive by default a referential and maximal interpretation. Although the discussion of the exact structure of headless relative clauses is minimal, there is no denying that the book contributes a great deal to the understanding of their semantic and morphosyntactic properties.
Following the Introduction, the rest of the volume consists of 14 chapters dedicated to the following 15 languages that represent the largest families of the region: two Southern languages from the Uto-Aztecan family (Southeastern Tepehuan and Tlaxcala Náhuatl), four Oto-Manguean languages (Acazulco Otomi, Matlatzinca, Iliatenco Me’phaa, and San Pedro Mixtepec Zapotec), seven Mayan languages (K’iche’, Q’anjob’al, Chuj, Ch’ol, Tsotsil and Tseltal, and Yucatec Maya), the Sierra Popoluca language from the Mixe-Zoquean family, and Pesh, the only Chibchan language spoken in Mesoamerica. The sample of languages is diverse both from a genetic and morphosyntactic point of view, as it includes, for example, nominative-accusative languages (Uto-Aztecan) as well as languages with morphological and syntactic ergativity (Mayan). The only improvement that can be expected in the future is that even more languages are examined to further enrich the comparison. Thus, within the Oto-Manguean, several Western languages are described and further juxtaposed to an Eastern language; however, Eastern (Mamean) Mayan languages are left out and the Mixe-Zoquean family has only one representative in the volume. This is not intended as a critique of the book, since such thorough examination of language data as presented here already requires a tremendous amount of work, but one would hope that a second volume will appear in the future.
One of the strong points of the book is that the authors analyze primary data that come first and foremost from their fieldwork. Many of the researchers who contribute to the volume are native speakers; they consider both naturally produced texts and examples collected via elicitation and can also rely on introspection. The authors’ meticulous research has resulted in comprehensive description of the constructions that includes negative data and remarks on inter-speaker variation and frequency; see, for example, Chapter 8, ‘Headless relative clauses in K’iche’, Chapter 12, ‘Headless relative clauses in Tseltalan’, and Chapter 15, ‘Headless relative clauses in Pesh’ commenting on the frequency of occurrence of different types of headless relative clauses in K’iche’, Tseltalan, and Pesh, respectively.
Another strong point already mentioned above that makes the volume a great reference work is that all chapters share the same structure. Each part opens with a brief introduction of the language under consideration, which is followed by an overview of the main syntactic properties of headed relative dependents and interrogative constructions. After that the headless relative clauses are discussed in great detail: maximal, existential and free-choice free relatives, light-headed relatives, and super-free relatives. Adopting a single template for all chapters makes navigating the material much easier for the reader, but most importantly, the data can be naturally compared across the languages and there are no gaps left in the description. From a methodological point of view, one can only applaud the decision to present the material in this way and wish that other volumes that examine a single linguistic phenomenon within a group of languages follow the pattern.
The authors extensively discuss the distribution of particular items (wh-elements, D-items, relativizers, etc.) in relative clauses, the distribution of the clauses themselves and their interpretation, but a formal analysis of the constructions is not attempted, except for several chapters on Mayan languages, including Chuj (by Royer), Ch’ol (by Vázquez Álvarez and Coon), Tseltalan (by Polian and Aissen), and Yucatec Maya (by AnderBois and Chan Dzul). The thought-provoking material presented in the volume leaves ample room for further in-depth investigation within any theoretical framework. In the remaining part of this review I would like to point out two research problems that can be explored building on the observations made in the book.
The first puzzle concerns the mutual distribution of complementizers and wh-expressions in headed and/or headless relative clauses and the variation attested among the Mesoamerican languages, compared to the patterns found in other language families and groups, especially the doubly-filled-COMP Filter (see Bayer Reference Bayer, van Oostendorp and van Riemsdijk2015 and references therein). For example, Southeastern Tepehuan (Uto-Aztecan) allows a complementizer to precede a wh-word in all headless relative clauses; in contrast, Yucatec Maya (Mayan) bans complementizers in relative constructions. The picture is more complex in Oto-Manguean languages. In Iliatenco Me’phaa (Oto-Manguean) the complementizer can occur in a relative clause but never together with a wh-expression, while San Pedro Mixtepec Zapotec exhibits peculiar internal variation that appears to correlate with the presence of additional morphology on the wh-element. In this language, complementizers are allowed in maximal and free-choice free relative clauses, where the wh-element bears a special prefix or suffix, but they are prohibited in existential free relatives, where the wh-word is affix-less. A related question is that of the distribution of the wh-items themselves. All wh-expressions used in interrogative clauses can be found in relative clauses in at least some of the Mesoamerican languages. However, similarly to their counterparts in Indo-European languages, ‘why’ and ‘how’ appear to be exceptional and are often restricted, for example, in Sierra Popoluca and Q’anjob’al. All this presents an interesting problem that, when solved, may help us find out more about the status of different wh-items and contribute to our understanding of the structure of the clausal periphery.
The second puzzle is that of the nature of light-headed relative clauses. A distinction must be made between genuine light-headed relative clauses and those with a silent nominal head. This becomes especially relevant for Mesoamerican languages, since many of them allow pro-drop and nominal ellipsis and have determiners that are used as stand-alone pronouns and/or wh-words that are also used as pronouns or nominals, as in Matlatzinca and Chuj. To my knowledge, the problem was first addressed by Citko (Reference Citko2004), who introduced the notion of light-headed relative clauses and discussed them in comparison to the headed relative clauses in Polish; see also Epps (Reference Epps, Comrie and Estrada-Fernández2012) and Gutiérrez-Bravo (Reference Gutiérrez-Bravo, Comrie and Estrada-Fernández2012), among others, arguing that at least some seemingly headless relative clauses should be reanalyzed as headed in various languages of the Americas.
The authors of the volume are keenly aware of this. As acknowledged in the Introduction, as a rule of thumb they assume that ‘silent N heads need to be licensed by syntactic and semantic/pragmatic conditions, as is the case for all silent elements’ (21) and continue treating headless relative clauses as such when no extra restrictions are observed. Yet extensive research into this problem would be welcome in the future. Some of the authors do address the issue further and come up with various language-specific tests aimed at detecting the silent N. Let me list those, to emphasize how diverse Mesoamerican languages are and how much their close examination can contribute to the general discussion of relative clauses.
Flores-Nájera notices that demonstratives in light-headed relative clauses generally lose their deictic readings and suggests that in Tlaxcala Náhuatl the relative clauses introduced by the items inon-keh ‘those’ and non ‘that’ are ambiguous between the light-headed (only a D-element) and null-N-headed (D + NØ) structures; a similar claim is made by Hernández-Green regarding Acazulco Otomi.
Examining seemingly headless relative clauses in Tseltal and Tsotsil, Polian and Aissen propose that only [+D, +Wh] clauses are truly headless, while in [+D, −Wh] clauses the nominal head is likely present but silent. To distinguish between the two cases they note that [–Human] wh-words (‘what’) are not allowed in headed relative clauses with an overt N, and thus their presence indicates that the clause is truly headless. A similar strategy is used by Can Pixabaj: in K’iche’ the elements jachin ‘who’ and jawi ‘where’ can occur as relative pronouns in headed relative clauses but never in light-headed relatives. In addition to this, she observes that light-headed relatives cannot be introduced by indefinite determiners, while those are perfectly compatible with an overt nominal head.
Another test is offered by Vázquez Álvarez and Coon for Ch’ol, where the relativizer bä appears in all headed relatives ([−Wh]), with an overt or covert N head; when bä is absent the clause ([+D, −Wh] or [−D, −Wh]) is analyzed as genuinely headless. Finally, López Márquez mentions that in Sierra Popoluca a headed relative clause can either precede or follow the head but light-headed relative clauses must always go after the D-item. While this might be attributed to the phonological weight of the head, the diagnostic is worth mentioning.
To summarize, the book offers a unique comparative perspective that sheds light on the nature of headless relative clauses and related constructions (headed relative clauses and content questions) in the world’s languages. It is a perfect example of a collective work on understudied languages, and the authors’ wealth of knowledge provides an essential tool to scholars working on syntax and semantics, especially to those interested in typology and descriptive fieldwork.
