ABSTRACT
In the Lexical Phonology model, the two generative components of the lexicon—phonology and morphology—are hypothesized to interact with each other in a way that amounts to cyclic rule application. This chapter discusses the controversy surrounding cyclic analyses and the various attempts to predict the class of cyclic rules. Diphthongization and Vocalization must be unordered with respect to each other in order to account for both surface variants of the gho-gho forms. These forms are compatible with either a cyclic or a non-cyclic analysis of the rules. The chapter argues that possessive suffixation and Nasalization must apply cyclically. The cyclic application of Schwa Lowering would bleed w Vocalization, because the output of Schwa Lowering would create a form which is identical to one which contains the second person plural subject prefix/ah/, in which Vocalization does not apply. Two Sekani rules, a Deletion and Schwa Lowering, are demonstrably noncyclic. This implies that cyclicity 'turns off' at some point in Sekani.
