“Los murlacos somo[z] inconfundibles”: Variable intervocalic /s/ voicing in the Spanish of Cuenca, Ecuador
Document Type
Paper Presentation
Presenter Language
English
Research Area
Language variation and change; Phonetics & Phonology; Sociolinguistics
Location
MBSC Council Room 306
Start Date
18-10-2024 3:00 PM
End Date
18-11-2024 3:30 PM
Abstract
Previous production studies of intervocalic /s/ voicing ([kasa] ~ [kaza] ‘house’) show a diverse set of linguistic and social conditioning factors depending on dialect. For example, while /s/ voicing is in stable variation in the Spanish of San José, Costa Rica (Author 2016), it is more prevalent among young speakers in Barcelona (Davidson 2020) and Loja, Ecuador (Author 2020). Highland Ecuadorian Spanish provides fruitful ground to study this variation given the dialectal diglossic situation within this macro-dialect (Gómez 2022). The present study systematically examines /s/ voicing in Cuencano Spanish, complementing previous impressionistic studies (Robinson 1979; Calle 2010).
Our data come from sociolinguistic interviews with native Cuencanos from 2013 (15 interviews) and 2023 (24 interviews), including 20 females and 19 males ages 19-73. Sixty tokens of intervocalic /s/ per speaker (20 per word position) were measured for fricative duration and percent voicing. Voicing is analyzed as both a continuous and a categorical dependent variable (‘unvoiced’, ‘partially’, and ‘fully voiced’ following Campos-Astorkiza (2014)). In the analyses, we include the following independent variables: word position, syllable stress, speech rate, preceding/following vowel, and participants’ age and gender, with speaker as a random effect.
Acoustic analysis is ongoing, but preliminary results from 13 interviews demonstrate that voicing in Cuencano is a gradient and variable phenomenon, rather than categorical as Robinson (1979) claims, and is more prevalent than in Lojano. Our results show that duration and voicing are highly correlated but conditioned somewhat differently. Shorter fricative duration is found in word final compared to medial or initial contexts, at higher speech rates, and before an /a/. More voicing is found in word final and medial as opposed to initial context, when /s/ is between unstressed vowels, and at higher speech rates. Examination of individual differences illustrates that while speech rate has the same effect on duration for all speakers, the effect of speech rate on voicing is not uniform among speakers, as previously found for Quiteño (Strycharczuk et al. 2013). The absence of social conditioning of /s/ voicing in Cuencano may suggest stable variation in this variety; however, this hypothesis needs confirmation with the full dataset.
“Los murlacos somo[z] inconfundibles”: Variable intervocalic /s/ voicing in the Spanish of Cuenca, Ecuador
MBSC Council Room 306
Previous production studies of intervocalic /s/ voicing ([kasa] ~ [kaza] ‘house’) show a diverse set of linguistic and social conditioning factors depending on dialect. For example, while /s/ voicing is in stable variation in the Spanish of San José, Costa Rica (Author 2016), it is more prevalent among young speakers in Barcelona (Davidson 2020) and Loja, Ecuador (Author 2020). Highland Ecuadorian Spanish provides fruitful ground to study this variation given the dialectal diglossic situation within this macro-dialect (Gómez 2022). The present study systematically examines /s/ voicing in Cuencano Spanish, complementing previous impressionistic studies (Robinson 1979; Calle 2010).
Our data come from sociolinguistic interviews with native Cuencanos from 2013 (15 interviews) and 2023 (24 interviews), including 20 females and 19 males ages 19-73. Sixty tokens of intervocalic /s/ per speaker (20 per word position) were measured for fricative duration and percent voicing. Voicing is analyzed as both a continuous and a categorical dependent variable (‘unvoiced’, ‘partially’, and ‘fully voiced’ following Campos-Astorkiza (2014)). In the analyses, we include the following independent variables: word position, syllable stress, speech rate, preceding/following vowel, and participants’ age and gender, with speaker as a random effect.
Acoustic analysis is ongoing, but preliminary results from 13 interviews demonstrate that voicing in Cuencano is a gradient and variable phenomenon, rather than categorical as Robinson (1979) claims, and is more prevalent than in Lojano. Our results show that duration and voicing are highly correlated but conditioned somewhat differently. Shorter fricative duration is found in word final compared to medial or initial contexts, at higher speech rates, and before an /a/. More voicing is found in word final and medial as opposed to initial context, when /s/ is between unstressed vowels, and at higher speech rates. Examination of individual differences illustrates that while speech rate has the same effect on duration for all speakers, the effect of speech rate on voicing is not uniform among speakers, as previously found for Quiteño (Strycharczuk et al. 2013). The absence of social conditioning of /s/ voicing in Cuencano may suggest stable variation in this variety; however, this hypothesis needs confirmation with the full dataset.