Compensatory lengthening in the Spanish of Havana, Cuba: Acoustic analyses of word-internal, post-nuclear /l/ and /r/

Kristin M Carlson, Purdue University

Abstract

Given the geographic, demographic, and historical importance of Cuba vis-á-vis the dissemination of language and culture throughout the Hispanic Caribbean, one would naturally anticipate a larger corpus of scientifically-noteworthy linguistic publications on Cuban Spanish, which is far from the actual case. Moreover, the gemination of an onset positionally subsequent to the deletion of a syllable-final liquid (generally termed liquid gemination in the literature) has been repeatedly claimed yet remarkably unsubstantiated as a pervasive characteristic of Cuban Spanish, particularly of the western dialect region (cf. Alfaraz (2000, 2007, 2008), Casanellas and Alamo (1985), Choy López (1985, 1988, 1989), Costa Sánchez (1987), Darias Concepción (2001, 2005), Dohotaru (2002, 2007), Figueroa Esteva and Dohotaru (1994), García Pérez (2006), García Riverón (1991), Haden and Matluck (1973, 1974, 1977), Isbaˇşescu (1965, 1968), Lamb (1968), Lévina (1970), Montero Bernal (1990, 2002, 2007a, b), Ringer Uber (1986), Ruiz Hernández (1978), Sosa (1974), Terrell (1976), Trista and Valdés (1978), Valdés Acosta (1980), and Vera Riverón (2000)). As a result, in the interest of supplementing all antecedent work concerning the allophony of final liquids as well as affording a more descriptively-precise account of the allophony of word-internal, post-nuclear /l/ and /[Special character omitted]/ in Cuban Spanish in addition to expressly addressing the need for empirical data-collection and analysis processes, the present investigation was specifically designed and implemented to acoustically investigate the phenomenon of gemination as it is purported to occur in the Spanish of the region of Havana, Cuba: more specifically, (1) to acoustically examine the qualitative and quantitative patternings of post-nuclear /l/ and /[Special character omitted]/ within the word; and (2) to statistically evaluate the relationship between gemination and eight independent variables: gender, age group, educational level, morphological significance of the syllable closed by the liquid phoneme, position of the syllable closed by the liquid phoneme relative to stress placement, voicing specification, manner of articulation, and place of articulation of the surface realization of the immediately-following onset. Qualitative acoustic analyses of the 1,895 tokens of word-internal, post-nuclear /l/ (n=469) and /[Special character omitted]/ (n=1,426) extracted from the corpus of informal data demonstrated seven allophones for the lateral liquid phoneme (Ø, [l], [d[Special character omitted]], [g[Special character omitted]], [w], [b[Special character omitted]], and [m]) and eleven for the simple vibrant phoneme (Ø, [g[Special character omitted]], [l], [d[Special character omitted]], [[Special character omitted]], [[Special character omitted][Special character omitted]], [j], [r], [g], [b[Special character omitted]], and [h]). Because deletion of both liquid segments accounted for the greatest number of allophonic cases in each respective data set, further quantitative acoustic analyses vis-á-vis four predetermined durational categories (i.e. full gemination, partial gemination, a 0-49.9% increase in the length of the immediately-subsequent onset, and a decrease in the length of the immediately-subsequent onset) revealed that for the post-nuclear lateral liquid a 0-49.9% durational increase was preferred irrespective of any of the eight independent variables under investigation; for the simple vibrant, a 0-49.9% length increase was also preferred irrespective of gender, educational level, morphemic status, and stress placement. However, a decrease in the length of the subsequent onset was the favored articulation by age group 3 (i.e. 30-39 years of age), according to a [-voice] specification, a fricative manner of articulation, and a dental, labiodental, velar, or glottal manner of articulation of the immediately-following consonantal segment. Statistical analyses of the four temporal categories vis-á-vis each of the aforementioned internal and external variables revealed a number of interesting results overall and specifically relative to gemination (or any durational change, for that matter). Statistical significance was found between the following dependent and independent variables for /-l/ → Ø / V_.C tokens: (1) partial gemination and male gender, age group 2 (i.e. 20-29 years of age), educational level 4 (i.e. attended or completed university or beyond), pretonic stress placement, a nasal manner of articulation, and a labial place of articulation of the immediately-following consonantal segment; (2) a 0-49.9% increase in the length of the subsequent onset and age group 2, educational level 4, pretonic stress placement, a [+voice] specification, a stop or nasal manner of articulation, and a velar or labial place of articulation of the immediately-following consonantal segment; and (3) a decrease in the length of the subsequent onset and age group 2, educational level 4, a [+voice] specification, a fricative or nasal manner of articulation, and a velar place of articulation of the immediately-following consonantal segment. No statistical significance was found between any of the examined durational categories and morphological significance. A statistically-significant relationship was found between the following dependent and independent variables for /-[Special character omitted]/ → Ø / V_.C tokens: (1) full gemination and educational level 4 and a [+voice] specification of the immediately-following onset; (2) partial gemination and age group 2, educational level 4, word-internal position, a [+voice] specification, a nasal manner of articulation, and a labial or alveolar place of articulation of the immediately-following consonantal segment; (3) a 0-49.9% increase in the length of the immediately-following onset and male gender, age group 2, educational level 4, word-internal position, a [+voice] specification, a nasal manner of articulation, and a labial or alveolar place of articulation of the subsequent onset; and (4) a decrease in the length of the immediately-following onset and age group 2, educational level 4, word-internal position, a fricative manner of articulation, and an alveolar or velar place of articulation of the immediately-following consonantal segment. No statistical significance was found between any of the examined durational categories and stress placement.

Degree

Ph.D.

Advisors

Hammond, Purdue University.

Subject Area

Linguistics|Latin American Studies|Acoustics

COinS