Describing the raising process of Old Spanish and the theory of the feature geometry

Abstract

This paper considers the raising process that affected the vowels of Late Latin in their development into Old Spanish; this process has been traditionally linked to the presence of the palatal glide /j/ in the environment. Two different analyses are taken into account and compared. On the one hand, following Clements 1991a model of vowel height, together with his formalization of scaler rules, we can describe the raising process in a very direct and economical way. On the other hand, using a standard feature system, where vowel height is characterized using different formally unrelated features, the raising rule of vowel height assimilation would have to involve changes in three different features. Undoubtedly, the features geometry approach offers a much more economical and "clean" description of this process, since it may even account for its scale nature. This study focuses on the description of the process; it does not attempt to explain the irregularity of its application.