Vowel Reduction and Reduplication in Klamath

Abstract

McCarthy and Prince (1995) propose a powerful new approach to reduplication, correspondence theory, which differs from previous approaches in its ability to enforce identity between a base and its reduplicant even after copying has taken place. Base-reduplicant (BR) correspondence makes it possible to account for ostensible cases of over- and underapplication, but in doing so predicts a wide range of cases that are apparently unattested. This squib examines one prominent case of ostensible backcopying - the nonapplication of vowel reduction in intensive prefixing reduplication in Klamath (Penutian) - and demonstrates that vowel reduction failure in the intensive cannot be derived from BR faithfulness. Rather, intensive prefixation is actually a process of stem formation not subject to the same constraints as regular prefixation. This reanalysis eliminates Klamath as support for BR correspondence and makes the more general point that morphological investigation must precede any attempt to attribute the nonapplication of a phonological alternation to backcopying underapplication