This book aims to evaluate the morpho-syntactic structures of clauses expressing causative and applicative meanings in several polysynthetic languages. This ambitious monograph presents very rich and diverse material in a comprehensive way, [End Page 133] wherein various superficially unrelated facts about the verbal morphology of a language fall into an ingeniously interwoven system of interdependent relations.
The author is a native speaker of Georgian, the language which constitutes the core of the analysis, with two other languages of the Southern Caucasus—Mengrelian and Svan—mentioned occasionally, to provide a more extensive outlook on Kartvelian languages as an historically connected group. Lomashvili verifies her data against the resources of descriptively oriented basic works on Kartvelian languages: Aronson (1990) for Georgian, Kajaia (2001) for Mengrelian, and Topuria (1967) for Svan.
The first chapter, "Preliminaries", introduces the linguistic tradition in which Lomashvili's work is situated. The basic tenets are drawn from Distributed Morphology and based on the works of Marantz (1984), Halle and Marantz (1993), Kratzer (1996), Harley (2008), Pylkkänen (2002), Cuervo (2003), and Embick (1997), among others. This approach does not distinguish word formation from clause formation in terms of the structures, the rules, or the units that the rules manipulate. Such a theory is especially convenient when we deal with polysynthetic languages, where sentence and word structures intertwine and are frequently one and the same thing. Consequently, isomorphism and correspondence of semantic content, proposed grammatical structures, and inserted morphological elements (here called Vocabulary Items), are expected. Lomashvili proposes such a system of representation for Georgian causative and applicative constructions, where the ultimate criterion to treat the data as relevant is semantics, i.e., any kind of meaning involving an additional causer automatically calls for the introduction of the causative structure and any kind of benefactive for the applicative one. Some other relevant theoretical concepts such as cyclic spell-out and phases, different types of functional heads (e.g.,Voice, Causative, and Applicative), as well as argument-introducing roots, are also briefly discussed.
In the same chapter, Lomashvili provides a brief overview of Georgian morphosyntax. Undoubtedly, here some reference to the literature on Georgian morphology would be in order, if only to direct the reader to more in-depth descriptions of the grammar of an understudied language. Conspicuously absent in this monograph is Anderson (1984), which, even disregarding its theoretical contribution, offers a very good and thorough description of Georgian morpho-syntax, especially given that it ventures into the intricacies of inverse (dative-subject) constructions, which Lomashvili analyzes.
Chapter 2, "Themorphosyntax of causative alternations", deals with the richness of causative clauses in Georgian as reflected by verbal morphological complexes. The clauses are analyzed along the lines of Pylkkänen's (2002) theory of causatives. The whole chapter pivots around the behavior of a causative Vocabulary Item (VI), which shows allomorphy with some classes of roots, while no allomorphy appears in other contexts: when the VI combines with inchoative roots (complements of the Cause head) it shows a-/ø-allomorphy. For such cases, Lomashvili adopts Harley's (2008) notion of lexical causatives, whose forms are conditioned by the phonological structure of roots—complements of lexical Cause heads. The roots and the heads appear in the same phase; they are in the appropriate structural proximity to [End Page 134] evoke the alternation. On the other hand, if the Cause head appears with unergative verbs, it is the whole vP which constitutes its complement, and no allomorphy results. Causative unergatives are syntactic causatives, and they show no causative allomorphy, since the conditioning factors—defective, consonantal roots for a-, and non-defective, vowelful roots for ø-allomorphs—are too deeply embedded in the structure to evoke the allomorphy in VIs.
For her causative structures, Lomashvili proposes that the external argument be introduced by Voice, above the Cause head, and not by the Cause head itself. This proposal aligns with her overall conception of Georgianmorpho-syntax; namely, that there are causative constructions without external arguments in the language, hence causation and the source of external...