It has generally been accepted that the Principle of Locality is a syntactic concept if locality is viewed from the perspective of the Closeness Condition or the Minimal Link Condition, while the Principle of Prominence belongs to the domain of semantics, although it also involves factors concerning pragmatics and cognition and includes the syntactic concept of constituent-command (c-command). Prominence and Locality in Grammar, by Jianhua Hu, challenges this consensus and advances the viewpoint that factors pertaining to locality and prominence are often interrelated in the actual production and interpretation of linguistic expressions. Using wh-questions and reflexives, two classical topics in generative grammar, Hu examines how the interaction between locality and prominence can account for some basic syntactic–semantic constraints that underlie the formation and interpretation of linguistic structures and expressions. While the focus is on Mandarin, he also looks at the roles that locality and prominence play in English. The notions of prominence and locality are employed in both wh-interpretation and reflexive binding but the formation and interpretation of wh-questions in Mandarin are constrained more by the principle of locality, whereas the interpretation of reflexives is constrained more by the principle of prominence.

It is difficult to find universal definitions of locality and prominence. Prominence is a more flexible concept, so it is more difficult to give it a unified definition even for different linguistic structures in the same language. To some extent, locality can be regarded as a representation of prominence. In other words, what is closer is more prominent. However, locality in terms of Closeness is a syntactic constraint, whereas prominence incorporates not only syntactic conditions like c-command but also semantic conditions like the animacy hierarchy.

Apart from the introduction and the conclusion, the remainder of the book is divided into two parts, i.e. ‘Wh-questions’ (Part I) and ‘Reflexives’ (Part II). The introductory section provides a brief overview of how Chomsky and other generativists understand locality and prominence. Inspired by these ideas, Hu argues that although the two principles seem to be simple, their actual performance and interaction in natural language is quite complicated, which results in variation within the same language. Things will become even more complicated in language comparison studies since prominence may be defined differently in different languages.

The wh-questions part examines how wh-questions are formed and how wh-elements are interpreted in Mandarin and English. Hu investigates movement and non-movement approaches to wh-questions. The former proposes that all wh-expressions in natural language undergo movement to some clause-initial position to form an operator–variable relation, though languages may differ in terms of where this movement applies, in overt syntax or at logical form (Huang Reference Huang1998). The non-movement approach was first proposed by Baker (Reference Baker1970) who claimed that a question-morpheme, as an operator, can unselectively bind all the free variables in its c-command domain. While agreeing that both approaches have their respective theoretical advantages and empirical implications in explaining wh-questions, Hu argues that the formation and interpretation of wh-expressions are also constrained by two conditions of economy, i.e. the Pure Clausal Typing Condition (PCTC) and the Principle of Economy (PE), both of which involve the principles of locality and prominence. PCTC is a locality condition that stipulates that a clause is typed by the wh-phrase that is closest to the left periphery of the clause. PE involves the interaction between locality and prominence and stipulates that in multiple wh-questions, the wh-phrase closer to the left periphery of the clause should be first moved to the [Spec, CP] position, because this movement better conforms to the locality condition and the moved wh-phrase is the most prominent on the Prominence Hierarchy. Hu assumes that if a wh-question involves two wh-phrases, one must function to generate an accessible set so that the other one can be interpreted. On the prominence hierarchy, a subject is more prominent than a nonsubject and an argument is more prominent than a nonargument. The more prominent a wh-phrase is, the more possibly it functions as a set generator. For instance, *What did who buy? is grammatically unacceptable since what is less prominent than who. Finally, he analyzes A-not-A questions and the asymmetry between weishenme ‘why’ and other wh-expressions in Mandarin. He argues that when weishenme and the A-not-A element occur in a strong island, they cannot take wide scope since they cannot be interpreted with the relevant C via either the Agree operation or the choice function. This part concludes that the formation and interpretation of wh-questions are constrained by two economy considerations, specifically, PCTC and PE, which incorporate two factors: prominence and locality.

The ‘Reflexives’ part examines the roles that locality and prominence play in the constraints on reflexives. Hu provides a unified theoretical analysis of the bare reflexive ziji ‘self’ and noncontrastive compound reflexives such as taziji ‘he/she-self’ in Mandarin. He defines the binding domains of ziji and taziji via the means of ‘feature search engines’. Taziji only lacks referential features and depends on the referential feature search engine to define the most prominent NP for its binding domain. Ziji, however, lacks both phi-features and referential features and has to depend on the phi-feature search engine and the referential feature search engine to do so. He concludes that the binding domains for both bare and compound reflexives are determined by the candidate set, a group of potential members related to the most prominent NP chosen by the relevant search engines. On the basis of an assumption put forward by Huang & Tang (Reference Huang, Tang, Koster and Reuland1991), Hu argues the bare reflexive ziji can undergo long-distance binding because it lacks both phi-features and referential features. Hence, the binding domain of ziji is determined by the union of the two candidate sets related to the two most prominent NPs chosen by the two search engines. Using the two search engines, Hu provides a new account of the blocking effect exhibited by reflexive binding and argues that the blocking effect results from the most prominent NP. That is to say, the blocking effect will be derived when reflexives are constrained by the binding domain that is defined by the most prominent NP.

Previous studies on wh-questions and reflexives have attached much importance to the role of locality in grammar. However, less attention has been paid to that of prominence in the grammatical system of natural language. The author provides a unified analysis with the principles of locality and prominence in natural language and shows that prominence also plays a crucial role in grammar, and the role it plays is even more important than that of locality in the grammar of Mandarin. In this sense, Hu’s research constitutes both challenges and innovation for the approach to wh-questions and reflexives.

The book challenges the current consensus on the analysis of wh-questions and reflexives from the perspective of the syntax–semantics interface and proposes an integrated approach combining analyses of the interaction between different levels of linguistic knowledge. It gives an in-depth description of the mechanism and conditions underlying our linguistic knowledge and cognition by using a set of theoretical and descriptive tools developed at the syntax–semantics interface.

For all the insights this book brings to linguistic researchers and interested readers, its theoretical depth and academic terminology may make it a little difficult for some readers who are not very well versed in this topic. Such readers are advised to do some preliminary reading if they are interested in this book. In addition, some descriptions in the book seem to be repeatedly emphasized, which helps consolidate the main ideas of the book on the one hand but adds workload for readers to process the information on the other hand. Finally, since the discussion focuses on two classical topics in grammar, i.e. wh-questions and reflexives, the applicability of Hu’s theoretical proposal may need to be further tested on other linguistic structures in both Mandarin and other languages, if possible.

All that said, the book is an important theoretical contribution to the syntax–semantics interface studies and can serve as a valuable text for graduate students and scholars in the field of Chinese, linguistics, and cognitive science.