- Agreement beyond phi by Shigeru Miyagawa
In this monograph, Shigeru Miyagawa expands, sharpens, and further substantiates the theory of clausal operations argued for in his 2010 monograph Why agree? Why move? Unifying agreement-based and discourse-configurational languages, exploring its consequences in many new domains of syntax and the syntax-discourse interface. Based on Chomsky’s (2001) uniformity principle, Miyagawa 2010 argued that movement operations across languages are governed by strong uniformity, which dictates that the same set of grammatical features occurs and is overtly manifested in all languages. The current monograph defends a version of strong uniformity, according to which phi-features (φ-features) and discourse-configurational features (δ-features) all start out on the phase head C and may or may not be inherited by T as a parametric option (cf. Chomsky 2005, 2007). The different combinatorial possibilities lead to four basic types of languages, depending on where the different sets of grammatical features are hosted: (I) φ-feature on C, δ-feature on T (Japanese); (II) δ-feature on C, φ-feature on T (English); (ΙΙΙ) both φ-feature and δ-feature on T (Spanish); (IV) both φ-feature and δ-feature on C (Dinka). In this simple and elegant typology, the variation we find across languages depends on how φ-features and δ-features interact with each other and with items in their local domain, the C-domain and the T-domain.
The book investigates a wealth of empirical phenomena in many typologically unrelated languages and arrives at a set of surprising new generalizations, which are accounted for in terms of the strong uniformity hypothesis, thus shedding light on the interaction between the agreement systems and the discourse-configurational properties of languages. It is argued that politeness markers in Japanese, allocutive agreement in Basque, sloppy interpretations of null subjects in Japanese, Chinese, Malayalam, and Spanish, ‘why’ and ‘how come’ in Japanese, Chinese, English, and Spanish, and focus in ga/no conversion in Japanese can be understood only if the locus of φ-features and δ-features in the clause is taken into account. The book thus initiates an exciting new research program in the study of formal syntax.
Ch. 1 introduces the basic theory argued for in the book: strong uniformity, the predicted typology of four language categories, and some preliminary empirical evidence for the unconventional claims that δ-features reside in T in Japanese and Spanish, while φ-features reside in C in Dinka (Nilo-Saharan; van Urk 2015). Chs. 2–5 are case studies illustrating how the typology developed in Ch. 1 can be fruitfully exploited to account for a range of phenomena that do not seem to be connected to one another. Ch. 6 concludes by summarizing the main findings of the monograph.
Ch. 2 argues that, contrary to what is assumed in the literature, Japanese is not a language lacking agreement; rather, it is a language where φ-feature agreement resides in C, being manifested as the politeness marker -mas-. Unlike, for example, French politeness agreement, which is sensitive to the clausal subject and can occur in all kinds of embedded clauses, the distribution of politeness marking in Japanese is sensitive to properties of the addressee and is limited to environments matching the notion of the Root (Emonds 1969). M argues that the differences between the two language types can be accounted for if φ-features are hosted in C in Japanese and in T in French, as predicted by the strong uniformity typology. He furthermore analyzes Japanese politeness marking as an analogue to allocutive agreement in Basque dialects. Basque allocutive agreement displays sensitivity to properties of the addressee (e.g. gender, status) and is limited to Root environments, just like Japanese politeness markers. M analyzes them as agreement in C, adopting Speas and Tenny’s (2003) speech act projection (SAP). SAP can only be present when C is unselected, thus providing a straightforward explanation for the fact that this type of agreement is limited to Root environments.
Ch. 3 investigates pro-drop and agreement. The empirical question addressed concerns the nature and availability of sloppy readings of zero subjects in three different types of...