Reviewed by Edward Vajda, Western Washington University
Tlingit forms a coordinate primary branch alongside Athabaskan-Eyak within the Na-Dene family. With about 845 speakers, all middle-age or older, this important language is highly endangered. No full-length grammar has yet appeared, and new descriptions of any aspect of Tlingit structure are noteworthy. This is particularly true of analyses targeting the complex morphophonology of verb prefixes, a phenomenon that has hitherto defied cogent explanation. Cable provides a succinct account of these alternations in the form of a single hierarchy of constraints operating across different subportions of the prefix string. His approach reveals that most apparent idiosycrasies in Tlingit verb-prefix shape accrue from general principles observable in other languages as well.
The study examines the Northern Coastal dialect taught in Alaska and owes much to three seminal works dealing with Tlingit verb morphology: The schetic categories of the Tlingit verb (Jeff Leer, Ph.D. dissertation, University of Chicago, 1991), A morphological study of Tlingit (Gillian Story, Dallas: SIL, 1966), and Tlingit verb dictionary (Gillian Story and Constance Naish, Fairbanks, AK: SIL, 1973).
This treatment adopts an optimality theory constraint-based framework to examine alternations between syllabic (CV) and nonsyllabic (C) allomorphs in Tlingit verb prefixes. Its major thesis is that prefix allomorphy is a natural product of more general phonological output constraints that act to maintain the overall metrical well-formedness of the verb word. Understanding them renders Tlingit verb structure much less erratic than commonly assumed.
One of the strengths of this analysis is that the author avoids direct appeal to the shape of particular morphemes whenever possible. Instead, the general phonological principles discovered elegantly explain many morphological quirks. For example, the 1st person subject prefix χa lacks the final /t/ found in the 1st person object prefix χat due to a rule disallowing coronal codas across a certain subportion of the prefix string. The failure of the first syllable of the distributive prefix daGa to syncopate to *dGa is explained in the same way.
The prefix position class ‘template’ is cited merely as a descriptive convenience and receives no theoretical significance. Nevertheless, the existence of six partly overlapping phonological subzones in the verb string—including the classifier + root complex (called the ‘stem’), the ‘inner conjunct’ prefixes (positions 10 to 2), the ‘outer conjunct’ prefixes (positions 11 to 14), and the nonconjunct prefix 15—must crucially be stipulated when defining and ordering the constraints that govern Tlingit prefix-vowel syncope. Why these domains should exist at all cannot be understood from any general phonological or metrical principle and remains an interesting question for future research. The partially overlapping subdomains of the Tlingit verb contradict accepted conceptions of discrete prosodic levels. Why Tlingit morphophonology should employ such a typologically unusual structure begs further investigation.
This study represents an important contribution to the understanding of Tlingit phonology and verb morphology that offers particular insights into the language’s prosodic structure. It also has implications for understanding Athabaskan verb-prefix morphophonology, with which the author draws many parallels throughout the work.