Another Look at Coeur d'Alene Harmony

Abstract

1. Much of the work done on Coeur d'Alene (Salish) since the publication of Reichard's (1938) comprehensive grammar has focused on the analysis of the complex retracting harmonies in this language (Cole 1987:77-103, Johnson 1975, and Sloat 1966; 1971; 1972; 1980). These efforts have resulted in a much clearer picture of the harmony processes than that presented by Reichard but still inadequately explain the data. In this paper, I make use of recent developments in phonological theory to provide a complete account of Coeur d'Alene harmony, demonstrating that what were thought to be irregularities are in fact the result of regular phonological rules applying to underspecified vowels. The analysis depends on a nonterminal Pharyngeal node (McCarthy 1988) in the feature hierarchy (Sagey 1986) which may be used to specify primary or secondary articulation. In 2, I present a brief description of the Coeur d'Alene vowels and the effects of the harmony processes; in 3, I review the various analyses that have been published and discuss the production of retracted segments; in 4, I describe the harmony processes; and in 5, I examine a number of forms that have proved difficult to account for in other analyses.