This is a valuable addition to Cambridge University Press’s well-known Textbooks in Linguistics series, and like all books in the series, it raises questions about how linguistics should be taught, and hence about what sort of textbooks are most likely to be useful. My own view is that it is important to get an in-depth understanding of some aspects of some specific language to see just how complex any language is (even the student’s own language). It is also important, in my view, for students to get some idea of what a precise and explicit analysis of some aspect of language looks like and how it differs from the informal descriptions of traditional grammars. With the former, there are clear implications about what is and is not possible (which can be tested), whereas with the latter, it is often not clear what follows because it all depends on how undefined terms are understood. But it is also important to get some idea of just what sorts of things are possible in language (most of which are not realized in any individual language). Languages may not vary without limit, as Joos (Reference Joos1957) famously claimed, but they vary a lot. Moreover, this is a complex matter since they may look different when they are actually quite similar and may look similar when they are actually rather different. A book like Croft’s is invaluable here, and not just to students.
Arguably, this is more of a handbook than a textbook. It has no exercises, although it has a list of terms defined at the end of each chapter. Maybe it is harder to offer relevant exercises in this area than some, for example, syntactic or phonological theory, but I note that Clackson’s (Reference Clackson2007) textbook on Indo-European linguistics includes exercises. It is also longer than many textbooks, for example, Croft (Reference Croft1990), but this is probably inevitable given the combination of breadth and depth for which it aims. Essentially, it seeks to describe how the most important grammatical constructions are realized in the world’s languages, and that is a big task.
The task obviously requires a cross-linguistically applicable descriptive framework, and this probably has to be meaning-based. Croft outlines a framework based on what he calls information content and information packaging. On the one hand, there are object, property, and action concepts; on the other, there are reference, modification, and predication. Typically, object concepts are involved in reference, property concepts in modification, and action concepts in predication, but Croft stresses that ‘any semantic content can be packaged in any information packaging function’ (13). Thus, thorn in the thorn’s colour denotes an object concept but is involved in modification, and in it is a thorn, it is involved in predication. Similarly, the word sharpness denotes a property concept but is involved in reference, and sharp in those thorns are sharp denotes a property concept but is involved in predication. Finally, scratching in the scratching of the thorns denotes an action concept but is involved in reference, and what looks like the same form in the thorns scratching me denotes an action concept but is involved in modification.
The information packaging functions are central to the organization of the book. Following an introductory Part I consisting of two chapters, Part II consists of three chapters, focusing on constructions used in first reference and then modification, dealing with such matters as the semantics of reference, pronouns, articles and definiteness, and various kinds of modification and their origins. Part III has nine chapters, looking at constructions used in predication, that is, clauses of various kinds. This deals, among other things, with the nature of events, various kinds of argument encoding, voice, topic, comment and related matters, and speech act constructions and polarity. Finally, Part IV has five chapters exploring constructions used for multiple predications, that is, complex sentence constructions. This considers coordination and subordination, comparatives, equatives and conditional constructions, and complement clauses and relative clauses.
It is not possible in a short review to discuss the content of 19 chapters. Instead, I will make some general points about the book. A central feature of the book is its focus on constructions. It is perhaps worth noting that much work in syntactic theory assumes that there are ‘no grammatical constructions of the traditional sort within or across languages’ (Chomsky Reference Chomsky1995: 6). The Chomskyan view has long been that the complexity of syntax is a product of the lexicon and some very general combinatorial mechanisms, mainly external and internal merge. Something similar is assumed in categorial grammar. But there is also work in syntactic theory in which constructions play a central role, such as Head-driven Phrase Structure Grammar, which seems rather closer to Croft’s position, and looks more promising to me (Borsley Reference Borsley, Bárány, Biberauer, Douglas and Vikner2020). Croft is naturally aware that not all linguists will agree with the analyses proposed in the book but makes the important point that ‘the basic crosslinguistic facts and patterns presented here are likely to represent lasting empirical generalizations, and must be explained in some way, no matter what syntactic theory one follows’ (xxvi).
It is probably worth adding that Croft assumes a fairly abstract notion of construction. Constructions are associated with certain meanings, and a specific construction may take quite different forms in different languages. The different forms are called strategies. I assume this means that a formal pattern that does not seem to be associated with a specific meaning is a strategy and not a construction. An example might be auxiliary–subject order in English, given the arguments in Borsley & Newmeyer (Reference Borsley and Newmeyer2009) that it is not associated with any specific meaning.
Another notable feature of the book is its emphasis on diachronic matters. To mention just a few examples, it is noted that case affixes are historically derived from adpositions (117), that adpositions are derived from either nouns or verbs (119), and that classifiers originate primarily in nouns (127). There is also considerable emphasis on ‘recruitment’, where a strategy used in one construction is extended to another. For example, a locative construction may come to be used to express possession. It is often pointed out in connection with such remarks that speakers generally have no knowledge of such matters, and hence, they are not relevant to the attempt to characterize a speaker’s grammatical system. This seems right, but grammatical changes stem from the interaction of language use and language acquisition, so they are very relevant to the study of those domains.
The book is also notable for some non-traditional terminology. For example, ‘indexation’ is used instead of ‘agreement’ and ‘non-person indexation’ instead of ‘concord’. Similarly, it has ‘balanced’ versus ‘deranked’ instead of ‘finite’ versus ‘non-finite’. But reasons are given for these terminological choices, and it is not hard to get used to them. It should be noted, however, that ‘deranked’ is a broader notion than ‘non-finite’, referring to any way in which subordinate clauses differ from main clauses. Non-finite clauses, subjunctive clauses, and nominalized clauses are all deranked.
Probably quite a few readers will have quibbles about the book, thinking either that it is not quite right in some area or that it should have said more (or less) about some topic. For example, the account of the meaning of questions given in Chapter 12 seems less satisfactory to me than the account developed within Relevance Theory in Wilson & Sperber (Reference Wilson, Sperber, Jonathan Dancy and Taylor1988) (which extends to rhetorical questions and speculative questions, among others). I would also have liked to see more about gerunds, verbal nouns, and masdars, and about free relatives. The latter just receive a very brief mention at the end of Chapter 19. I was also disappointed to read that ‘[t]here is no typological survey of the form of comparative conditionals’ (545), having done some work in this area, for example, in Abeillé & Borsley (Reference Abeillé and Borsley2008).
But whatever quibbles one might have, this is a very useful book, and not just for students. It is also valuable for linguists who are not typologists, including linguists like myself, who have worked on a number of languages but are a bit hazy about all the others. With this book, we can become quite a lot less hazy.
