Dialect geography of Syria-Palestine, 1000-586 B.C.E. By W. Randall Garr (review)

726 LANGUAGE, VOLUME 62, NUMBER 3 (1986) H succeeds admirably in his main goal, which is to present all the relevant information he has on the dialect in a readily accessible form. His notation (which includes some morphophonemic information) is easy to decipher, and the tables and over-all organization are clear. A footnote (410-11) indicates that a cassette tape is available, containing recordings of some of his texts; thus interested scholars can check the pronunciations. H also provides numerous examples for all the inflectional and accentual categories, and he gives variant inflectional forms with general predictions (where possible) about the likelihood of occurrence of each variant . Some readers will regret the omission of any discussion of verbal aspect; H treats aspectual pairs as separate verbs, on which aspect is not indicated. One thing H does not do is give reasons for his analytic decisions, even when they deviate from standard practice among Slavists. Thus he posits no affricate phonemes, on the ground that syllable-initial stop + fricative sequences 'can be interpreted biphonemically in all positions' (14); but a biphonemic analysis results in wordinitial clusters /ts/ and /c$/—the only such clusters in which a stop is followed by a voiceless fricative. H sets up three gender-based declension classes for nouns and merges the remaining old ¿-stem feminine nouns with the general feminine class, which primarily continues the old astems . It may be that too few of the ¿-stem endings are preserved to justify a separate declension class; however, some are used consistently . Verbs are grouped into nine conjugation classes, primarily on the basis of the thematic vowels ofthe infinitive and present tense forms; this organization certainly makes the description easier to read than a Jakobsonian one-stem analysis would be; but a one-stem analysis, with a concomitant reduction in the number of verb classes, would probably come closer to the actual structure of the dialect's verb system. Quite reasonably, given the scope ofhis book, H does not explore the comparative or historical implications of Orlec dialect features. Aside from the important prosodie phonemes and their morphophonemic alternations, Orlec has a number of other noteworthy points. Thus it shows an unusual innovation in the animacy category: in addition to the usual SC (and Proto-Slavic) pattern in which the mase. an. ace. sg. = genitive , but the mase. inan. ace. sg. = nominative, Orlec has extended the animate/inanimate distinction from the singular to the plural of masculine nouns, and from there to the plural of feminine nouns. However, in the plural of both masculines and feminines, the accusative = genitive pattern is restricted to nouns denoting people; non-humans are grammatically animate in the mase, sg., but inanimate in the mase, and fern, plural. This innovation may be offairly recent date in Orlec: a 1909 description ofanother Cres dialect showed the animacy distinction only in the mase, sg., and in 1910 a somewhat more distant Northwest Cakavian dialect had an animacy distinction in the mase. sg. and pi., but not in the fern. pi. In sum, this book is a valuable contribution to SC dialectology. Slavists who study accentology , as well as general linguists interested in the nature and history of complicated prosodie systems, will want to read it. [Sarah G. Thomason , University ofPittsburgh.] Dialect geography of Syria-Palestine, 1000-586 B.C.E. By W. Randall Garr. Philadelphia: University of Pennsylvania Press, 1985. Pp. xiv, 289. $40.00. G has written an outstanding book comparing the Northwest Semitic (NWS) 'speech forms' of the first millennium B.C.: Phoenician, Hebrew , Aramaic, Moabite, Ammonite, Edomite etc. G chose the dates 1000-586 B.C. because those four centuries reflect the period when ancient Israel was a viable entity—from King David until the fall of Jerusalem. To deal with a greater time span would probably have complicated the task, although G might have provided a chapter on Ugaritic since so much knowledge of NWS comes from that language. Ugarit was destroyed around 1200 B.C., only 200 years before G's beginning date. A major value of this book is in the classification and subclassification of NWS. Z. Harris surveyed the linguistic history ofthis area in his Grammar of...