Introduction to Indo-European linguistics By Oswald J. L. Szemerényi (review)

456LANGUAGE, VOLUME 76, NUMBER 2 (2000) Department of Linguistics 619 Williams Hall University of Pennsylvania Philadelphia, PA 19104-6305 [cardona@unagi.cis.upenn.edu] Introduction to Indo-European linguistics. By Oswald J. L. Szemerényi. Oxford: Clarendon Press. 1997. Pp. xxxiii, 352. $90.00. Reviewed by Joseph F. Eska, Virginia Polytechnic Institute and State University This long-awaited English translation ofOswald Szemerényi's Einführung in die vergleichende Sprachwissenschaft (first published in 1970) is based on the fourth edition of that volume (1991) and incorporates new notes and references. Happily, the title of the volume finally reflects the contents of the volume accurately, unlike the various German editions and the previous translations into Spanish, Russian, and Italian, as reviewers of those volumes have repeatedly noted. Sz was one of the best known practitioners of Indo-European (IE) linguistics, and so it is good to have his views in their final form in English, for until recently, with the publication of the translated volumes of Beekes (1995) and Gamkrelidze and Ivanov (1995), students in North America had little recourse but to begin their study of this intricate field through a foreign language, notably German. As an introductory textbook to such a wide-ranging and complex field as IE linguistics, the book warrants a cautionary note at the outset. Individual scholars, naturally, may not agree with the communis opinio on certain points or may have particular theories from which they are reluctant to part, and these are unquestionably in evidence in this volume; but Sz's sometimes involve such basic matters as the composition of the phonemic inventory or the number of 'laryngeals' to be reconstructed, matters on which broad consensus has been reached. Unfortunately, perhaps owing to considerations of space, Sz sometimes does not present an argument for his own views (or for disagreement with those of others) but simply asserts his opinion. Furthermore, while new notes and references keep up to date, more or less, with advances in the field, this is not always the case with the main text, which, as a result, more than a few times is out of sync with the opinions expressed in the notes. AU scholars interested in IE linguistics should read this volume and will profit from it; but it should not be read in isolation. This is not only because of the reasons given above but also because in such a large language family as IE, no one scholar can control all of the available data and have a critical opinion on the published literature. As one whose primary area of specialization in IE linguistics is Celtic—not one of Sz's primary areas—I noticed an uncomfortable number of assertions made or opinions reported that can hardly be considered mainstream today, e.g. that the IE syllabic nasals were realized as both /aN/1 or /eN/ in the Celtic languages (48)—in fact, */aN/ is now considered to be the proto-Celtic treatment and /eN/ a (sporadic) secondary development in various parts of the Celtic speech area, and that the mysterious set of inflectional desinences in the Old Irish imperfect can be directly traced back to the IE imperfect (298). The volume opens with three short introductory chapters: Ch. 1, 'Introduction' (1-13), presents views on language held in antiquity, attributes the foundation of historical linguistics to the publication of Jakob Grimm's Deutsche Grammatik (1819-37), and reviews the composition of the individual IE subgroups very briefly. Ch. 2, 'Language in change' (14-30), focuses on the regularity of sound change and employs the familiar example of the Germanic sound shifts discovered by Rasmus Rask and Jakob Grimm, noting—importantly—that exceptions to sound 1 N = any nasal consonant. REVIEWS457 laws can exhibit some regularity, too; see the essays in Durie and Ross (1996) for further work in this area as well as on important alternative methods of explaining (so-called) irregularity in sound change. Ch. 3, 'Tasks of LE linguistics' (31-36) states that the first objective of the LEist is to work back to the fullest possible reconstruction. Only once late IE has been reconstructed can one go on to the reconstruction of deeper levels of LE...