Comerse las eses: A selective bibliographic survey of /s/ aspiration and deletion in dialects of Spanish.

Abstract

Comerse las eses or eating the s's is a term used by native speakers of Spanish describing the aspiration ( (s) $>$ (h)) or deletion ( (s) $>$ o of /s/. Experts identify /s/ aspiration and deletion as among the most complicated features of Spanish. This dissertation critically surveys Spanish /s/ in terms of these two phenomena. It aims to provide at least a partial explanation of Spanish /s/ reduction by considering data from phonetics, synchronic phonology, diachronic linguistics, dialectology (Peninsular and New World), and sociolinguistics. The survey identifies several factors that may be relevant to retention versus reduction of Spanish /s/: tongue position (apical versus laminal), place of articulation (ranging from alveolar to palatal), and position within the syllable (initial versus final). Laminal (s) (consistently alveolar) is more likely to reduce than apical (s) (which tends to be postalveolar). Syllable-final /s/ is far more likely to reduce than syllable-initial. Chapter 2 investigates possible phonetic sources underlying these factors. Intensity may be relevant to apical retention: non-experimental evidence suggests that apicals may be more intense than laminals and hence possibly more salient perceptually. If so, apicals may be more resistant than laminals to reduction. Acoustic-perceptual factors may also be relevant to the tendency for syllable-final /s/ to reduce. Experimental evidence supports the notion that phonetic information is weaker in syllable codas than in onsets. Chapter 3 considers diachronic evidence. These data provide further support for the claim that apical (s) has been resistant to reduction in that, when the apical does reduce, it appears to pass first through a laminal stage prior to reduction. Chapter 4 reviews dialectal studies that further delineate the apical-laminal issue and also reveals the extent of retention, aspiration, and deletion and their phonetic contexts (e.g., preconsonantal reduction is more prevalent than prevocalic or prepausal reduction). Recent sociolinguistic research treats the effects of age, gender, profession, education, and socioeconomic levels on /s/ realizations. Chapter 5 concludes the dissertation with an assessment of current understanding of these phenomena and recommendations for future research.