This volume presents a typological/theoretical introduction plus eight papers about ergative alignment in 16 Amazonian languages. All are written by linguists with years of fieldwork and comparative experience in the region, all describe details of the synchronic systems, and several also provide diachronic insight into the evolution of these systems. The five papers in Part I focus on languages from four larger families with ergative patterns primarily in morphology. The typological contribution is in detailed consideration of unusual splits, changes in ergative patterns, and parallels between ergative main clauses and nominalizations. The three papers in Part II discuss genetically isolated languages. Two present dominant ergative patterns in both morphology and syntax, the other a syntactic inverse system that is predominantly ergative in discourse. In each, the authors demonstrate that identification of traditional grammatical relations is problematic. These data will figure in all future typological and theoretical debates about grammatical relations.

This volume presents a typological/theoretical introduction plus eight papers about ergative alignment in 16 Amazonian languages. All are written by linguists with years of fieldwork and comparative experience in the region, all describe details of the synchronic systems, and several also provide diachronic insight into the evolution of these systems. The five papers in Part I focus on languages from four larger families with ergative patterns primarily in morphology. The typological contribution is in detailed consideration of unusual splits, changes in ergative patterns, and parallels between ergative main clauses and nominalizations. The three papers in Part II discuss genetically isolated languages. Two present dominant ergative patterns in both morphology and syntax, the other a syntactic inverse system that is predominantly ergative in discourse. In each, the authors demonstrate that identification of traditional grammatical relations is problematic. These data will figure in all future typological and theoretical debates about grammatical relations.