Hiatus Resolution in Spanish: An Experimental Study
Open Access
- Author:
- Souza, Benjamin J.
- Graduate Program:
- Spanish
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- December 16, 2009
- Committee Members:
- Henry Gerfen, Dissertation Advisor/Co-Advisor
Henry Gerfen, Committee Chair/Co-Chair
Paola Eulalia Dussias, Committee Member
John Lipski, Committee Member
Barry Richard Page Jr., Committee Member - Keywords:
- linguistics
Spanish
phonology
hiatus
diphthong
laboratory
phonetics - Abstract:
- In Spanish, adjacent vowels across and within word boundaries are either in hiatus or form a diphthong. Generally, when either of the unstressed high vowels /i/ and /u/ appears next to any of the other vowels /e/, /a/, or /o/ the result is a diphthong (i.e., puerta ‘door’ > [pwer.ta], miel ‘honey’ > [mjel], and so on). All other combinations of the vowels /i, e, a, o, u/ (along with their orthographically-accented counterparts where possible, /í, é, á, ó, ú/) typically remain in separate syllables and are said to be in hiatus (i.e., teatro ‘theater’ > [te.a.tro], poeta ‘poet’ > [po.e.ta], país ‘country’ > [pa.ís], etc.). When speaking, however, vowel combinations that are supposed to be in hiatus often are not, such that teatro [te.a.tro], for example, is reported to be pronounced [tea.tro] or even [tja.tro]; se aman [se.a.man] might become [sea.man] or [sja.man], and so on— a phenomenon referred to as hiatus resolution. Previous accounts of hiatus resolution have claimed that there are a number of different possible phonetic outcomes of hiatus resolution with most authors finding that hiatus resolution typically results in a) the formation of a diphthong, b) vowel deletion, or c) maintenance of hiatus (Jenkins 1999, Aguilar 2003, Alba 2006, Díaz-Campos et al. 2006). Some of these same accounts have also suggested that there are a number of factors that predict when hiatus resolution is likely to happen. These include which specific vowels are found in the VV sequence, and which vowel(s) (if any) receive stress (Jenkins 1999). Other factors have been investigated as possible catalysts for either hiatus resolution or maintenance including dialectal differences (Lipski 1994, Jenkins 1999, Hualde 2005, Alba 2006, Díaz-Campos et al. 2006, Michnowicz 2007, among others), individual speaker differences (e.g., Quilis 1981, Hualde et al. 2002), and differences in socioeconomic levels (e.g., Matluck 1995, Díaz-Campos et al. 2006), speech rate, and carefulness of speech. It is readily agreed upon that the faster one speaks and the less care with which one speaks the more likely vowel sequences in hiatus are resolved (e.g., Navarro-Tomás 1968b, Harris 1969, Jenkins 1999, Alba 2006). An additional factor that has been reported to affect hiatus resolution is frequency. Although it has received less attention in the traditional hiatus/diphthong literature, recently some authors have found that more frequent vowel sequences, words, and/or word strings tend to show both a higher propensity to resolve hiatus, as well as to show a higher degree of vowel resolution (Jenkins 1999, Aguilar 2003, Alba 2006, Díaz-Campos et al. 2006). Both Jenkins (1999) and Alba (2006) found that the more frequent word strings in their corpora showed greater tendency to resolve hiatus and to resolve it to a greater degree (i.e., more resolution, or even more unlike the original vowels in hiatus). These findings claim that hiatus resolution is not brought on only by fast and/or careless speech but that frequency might also be playing an important role in predicting what (and to what degree) vowel sequences, words, or word strings will undergo hiatus resolution. A 2x2 experimental investigation of 36 Mexican Spanish speakers was conducted to investigate via word-naming tasks the role of frequency (high vs. low) and speech rate (fast vs. slow) in the resolution of within-word [ea] sequences in Spanish. Words and vowel sequences were measured for overall word duration and VV sequence duration. Formant data was also measured for the vowel sequences to determine the nature of the resolution (diphthong-like, or not). The results reveal that fast speech rate significantly modulated both temporal measures as well as formant measures. The results also reveal a significant modulation of temporal and formant measures for the high frequency words. Phonetic environment also significantly affected the modulation of the vowel sequences. The nature of the resolution in the fast speech condition and in the high frequency words was away from diphthongization and toward vowel coalescence. The results suggest that frequency is, indeed, and important factor in the resolution of [ea] sequences in Spanish, and lend support to a usage-based model of phonology which posits that high-usage items are likely to undergo more phonological modifications than their low-usage counterparts.