Generative linguistics and Arabic metrics

Abstract

The classical theory of Arabic metrics has been reinterpreted numerous times in a variety of theoretical frameworks, notable among them Metrical Phonology (Prince 1989 a Schuh 1996) and Optimality Theory (Golston a Riad 1991), which we briefly examine to show that they do not permit an adequate account of the structure of Arabic verse-patterns. The analyses that have been proposed, by simply reinterpreting classical analyses, suffer from the same failings as the classical theory, whose descriptive adequacy (or inadequacy) is never discussed. To these failings we can add the problems inherent in the basic presuppositions of these theories (hierarchical organisation of meter, constituency and, above all, binarity), whose relevance (their explanatory power) and universality deserve to be seriously questioned. Finally, it is shown that a detailed descriptive analysis of actual instances of classical verse-patterns in terms of free and fixed metrical positions reveals a system which relies on principles and constraints that consistently depart from the classical framework and should thus serve as the basis for future analyses of Arabic meter.