Slavonic and East European Review, 95, 4, 2017 Reviews Bond, Oliver; Corbett, Greville G.; Chumakina, Marina and Brown, Dunstan (eds). Archi: Complexities of Agreement in Cross-Theoretical Perspective. Oxford Studies of Endangered Languages. Oxford University Press, Oxford and New York, 2017. xxii + 280 pp. Figures. Tables. Notes. Bibliography. Indexes. £65.00. Archi is a Lezgic Northeast Caucasian language spoken in one village, Archib (with six adjacent settlements), in the central highlands of Daghestan. Archi separated from the main body of Proto-Lezgic a very long time ago, perhaps because a family or clan fled a blood feud to what was then territory that was uninhabited owing to its isolation and limited resources. As a result of its history, Archi has evolved into a unique combination of archaic and specific features, and it is now one of the most ecologically fragile languages in Daghestan. The current volume is an important contribution to Caucasology, typology and formal linguistics as well as to the study of endangered languages. Archi, unlike the rest of Lezgic but like some other Northeast Caucasian languages, displays what the authors call ‘extreme agreement’, i.e. gender marking on a variety of parts of speech (including adverbials and adpositions) via prefixes, infixes and suffixes. Also, like the rest of Daghestanian, Archi is semantically ergative, although agreement patterns are split and bi-absolutive constructions are possible. As gender marks and controls person agreement, this adds an additional complexity to the analysis of verb phrases. Moreover, agreement sometimes occurs on an item outside the clause where the gender-controller occurs. Archi thus represents significant challenges to any scholarly approach to language, and the editors and authors are to be congratulated for producing such a broadly useful and interesting volume. The premise of the volume is to use Archi data to test three different theories of grammar by focusing on one particularly complex phenomenon: agreement. The title of the book, especially as it appears on the cover and title page, is justified: ARCHI: Complexities of Agreement in Cross-Theoretical Perspective. The unsuspecting reader might pick up the book thinking to find in it a grammatical description of Archi, and in this such a reader will not be disappointed. The ‘Introduction’ (chapter 1: Bond, Corbett, Chumakina, pp. 1–16) explains the rationale of the book, defines basic concepts and terminologies, and outlines the structure of the chapters. Chapter two (Chumakina, Bond, Corbett, pp. 17–42) outlines ‘The Essentials of Archi Grammar’ in theory-neutral terms that can be used by any linguist or student of the Caucasus. Owing to the nature of the book, this is an outline, not a reference grammar. For greater detail, the outline has good references SEER, 95, 4, OCTOBER 2017 734 to relevant works cited in the References (pp. 261–72), which contains more than 200 entries. The book also has indices of Authors (pp. 273–75), languages (p. 276), and Subjects (pp. 277–80). Thus armed with the essentials of Archi grammar, the reader can proceed with profit to the next two chapters: 3. ‘Agreement domains and targets’ (Bond and Chumakina, 43–76), and 4. ‘Competing controllers and agreement potential’ (Chumakina and Bond, pp. 77–117). In chapter three, the authors describe the basic morphology and mechanics of agreement, and in chapter four they describe agreement in phrases and clauses. Particular attention is paid to bi-absolutive constructions, i.e. where two participants are both in the absolutive case instead of one being in the ergative (or some other) case. The accounts are lucid and theory-neutral. As with chapter two, they can be used with profit by any student of language. The next three chapters examine the facts adduced in chapters 2–5 through the lenses of three competing formalist syntactic theories: chapter five, entitled ‘HPSG and the nature of agreement in Archi’ (Robert D. Borsley, pp. 118– 47), chapter six, entitled ‘Agreement in Archi: An LFG perspective’ (Louisa Sadler, pp. 150–83), and chapter seven, ‘Agreement in Archi from a minimalist perspective’ (Maria Polinsky 184–232). In each of these chapters, the authors begin by laying out clearly the premises of the theoretical frameworks in which they are working. Having assured that the reader understands...