A PERCEPTUAL AND ACOUSTIC STUDY OF SYLLABLE-FINAL AND WORD-FINAL -S AND -N IN PUERTO RICAN SPANISH

Diane Uber

Ph.D. Dissertation, 1981

Abstract

This study examines the production and perception of syllable-final and word-final -s and -n in Puerto Rican Spanish. A review of the literature in Chapter II discusses the research which has been carried out to date on the deletion and weakening of -s and -n in all dialects of Spanish for which these processes have been noted. It is found that there have been many claims that when these sounds are deleted, there is always some other phonetic modification present which maintains the grammatical or semantic distinctions marked by [s] or [n] in other dialects of Spanish. Chapter III presents the results of a perception test which show that if such phonetic distinctions are present, they are not always perceived by native speakers of Puerto Rican Spanish when utterances are taken out of context. Morphological, syntactic, semantic, and discourse context are shown to be very important for perception of utterances involving -s and -n. The results of the perception test show that weakened variants of -s and -n, in addition to deletion of -s and -n, can cause perceptual problems for listeners. Chapter IV compares perception and production of utterances involving -s and -n for one subject-informant. It is found that she uses in her own speech many of the same phonetic realizations as those which she identified incorrectly on the perception test. Chapter V presents the results of a spectrographic study of the items used for the perception test. These acoustic displays provide physical evidence for the auditory perceptions of this author and help explain the responses of the subjects in some cases. Chapter VI discusses implications for phonological theory. The underlying representations for -s and -n in syllable-final and word- final positions are considered to be /h/ and /(eta)/, respectively for(, ) Puerto Rican Spanish. A set of unordered rules, compatible with the theory of Natural Generative Phonology, will produce the phonetic output. It is shown that Puerto Rican Spanish may be moving toward a system with no underlying segments for -s and -n, in which listeners must rely on context for correct perception.