The phonology of Armenian By Bert Vaux (review)

REVIEWS The phonology of Armenian. By Bert Vaux. Oxford: Clarendon Press, 1998. Pp. 280. Reviewed by Amalia Khatchatrian, Yerevan This impressive book is actually the first comprehensive examination of the sound system of Armenian in its two literary forms—Eastern and Western Armenian—and particularly its numerous dialects. The investigation is unique in that it has been carried out from a definite theoretical position—the methods and argumentation of rules and representation theory (RRT)—a modernized version of generative phonology. Clearly structured and well substantiated it not only makes strong theoretical claims but also presents detailed analysis ofa number ofphonetic issues hitherto viewed separately or unexplained. From the traditional point of view, however, this is a phonology without time and space. The Saussurean dichotomy between synchrony and diachrony is no more relevant for the author. He freely makes use of facts of Classical Armenian and dialects as well as modem literary language referring to different epochs and geographical areas. The book consists of eight chapters. The first one, Survey of Armenian phonology, elegantly summarizes the essence of the book and points to specific sections in which more detailed analysis and arguments are given. It is clearly written and presents the contemporary state of Armenian phonetics and phonology. Based mainly on East Armenian, it deals with 36 dialects (7-8). Under the paragraph 'Historical dialectology' Vaux concentrates on the most important linguistic innovations in Armenian since the post-classical period, particularly in phonology, in which a series of consonant shifts have occurred. V succinctly summarizes the vast linguistic literature and the discussion waged in Voprosyjazykoznaniya, on this issue thus making available the results of Armenians' work to Western readers. One may wonder about the author's extensive and solid comprehension of such a wide range of linguistic and factual material on Armenian. Ch. 2 exposes the theoretical background of the book and the innovations the author has brought in RRT worked out by a group of linguists at Harvard University. The essence of this theory is well described in literature (Halle and Marantz 1993, Kenstowicz 1994, Calabrese 1995, and others). V applies a modified version of it which includes the idea of hierarchical structure of phonological features and a graphical organization of these features in a tree-like structure which reflects the vocal tract configuration. V often makes references to other authors for some more new ideas (sensitivity, markedness, cyclicity). Much effort is exerted to justify the features ATR and RTR (advanced and retracted tongue root) widely used in RRT theory. Ch. 3, the core ofthe book, presents a most comprehensive study ofsyllabification in Armenian. Its depth of analysis far surpasses other descriptions of the problem and observes it from a new viewpoint. Indeed, one is amazed at the richness of data that the author was able to analyze while uncovering the wealth of dialectal facts, even doing some field work. Syllabification in Armenian, a vast field of research, is closely connected with the rules of reading. The apparent discrepancy between the orthography and pronunciation may seem striking, but it is governed by the rules which are being taught at school and are presented in a simplistic way, sometimes even conflicting with the pronunciation. From the phonetic point of view, the consonant clusters have not been tackled since phonetically they cannot be considered as such. The first linguist who examined this problem phonologically was Bella Gulakian1 whose name, surprisingly, is not mentioned in the book, though she put forward a view very close to the one expressed in this book. In discussing the interrelation between underlying representation and phonetic representation, 1 Tackling the phonological system from the positions of transformational grammar Bella Gulakian excluded the vowel h/ from the vocalic system of Armenian, a conclusion coinciding with the one expressed in this book (p. 70 and elsewhere) 433 434LANGUAGE, VOLUME 76, NUMBER 2 (2000) V proceeds from the premise that the graphical form of words is actually the underlying representation in which all the relevant features of sounds are represented. The changes in orthography that have taken place in both Standard East Armenian (SEA) or West Armenian (SWA) are not taken into account; thus, final mute j, which is not preserved...