In recent work (McCarthy 1979, 1981; Halle and Vergnaud 1980; Harris 1980; Marantz, to appear; Yip, to appear) a new model of morphology has been emerging, one in which nonlinear phonological representations play a central role. This model, which I will refer to as “prosodic,” was originally developed in an analysis of the complex system of nonconcatenative morphology found in Semitic languages, Classical Arabic in particular. It has since been extended to other, typologically and genetically quite different sorts of phenomena. In this paper, we will see still further empirical consequences of the adoption of this theory.