Files
Abstract
This dissertation investigates two morphosyntactic features in Mosquito Coast Spanish (MCS) and Nicaraguan Spanish (NS): subject pronoun expression (SPE) and variable number marking. Both NS and MCS are spoken in Nicaragua; however, NS is a monolingual, non-contact variety and MCS is a contact variety spoken by first language Miskitu speakers.
The analysis of SPE examines the distribution of explicit and implicit third person subject pronouns, while the analysis of variable number marking examines variation in third person plural subject-verb agreement. The study involved the statistical analysis of oral data, which was subjected to a mixed-effects logistic regression to determine what linguistic and social factors motivated the variation of each feature. The participants were 10 NS speakers and 10 MCS speakers for SPE, and 20 MCS speakers for variable number marking.
The results for SPE show NS speakers align with previous research: singular referents, less distinctive TMA, contexts of switch reference, and non-reflexive verbs all motivated the production of explicit pronouns. In the MCS data, less distinctive TMA and contexts of switch reference motivated the production of explicit pronouns, along with participants over the age of 30 and females. Additionally, MCS speakers showed a low rate of explicit pronouns overall (16.5%, compared to 23.9% in NS). The results for variable number marking showed lack of agreement occurred in 23.87% of the MCS data and was motivated by low phonic salience verbs, subject position (distantly preceding the verb and post posed), animate referents, participants over the age of 30, and males. For both features, MCS speakers used salience and L1 influence as strategies to compensate for differences that exist between Spanish and Miskitu regarding SPE and number marking.
The findings provide evidence for incomplete acquisition of Spanish by MCS speakers. The low rate of explicit subject pronouns and variable subject-verb agreement are likely remnants of the rapid language shift to Spanish that took place in the past but have now become fixed features of MCS. Younger speakers, however, are approximating to distributions found in NS, which suggests that these features are currently changing in MCS.
The analysis of SPE examines the distribution of explicit and implicit third person subject pronouns, while the analysis of variable number marking examines variation in third person plural subject-verb agreement. The study involved the statistical analysis of oral data, which was subjected to a mixed-effects logistic regression to determine what linguistic and social factors motivated the variation of each feature. The participants were 10 NS speakers and 10 MCS speakers for SPE, and 20 MCS speakers for variable number marking.
The results for SPE show NS speakers align with previous research: singular referents, less distinctive TMA, contexts of switch reference, and non-reflexive verbs all motivated the production of explicit pronouns. In the MCS data, less distinctive TMA and contexts of switch reference motivated the production of explicit pronouns, along with participants over the age of 30 and females. Additionally, MCS speakers showed a low rate of explicit pronouns overall (16.5%, compared to 23.9% in NS). The results for variable number marking showed lack of agreement occurred in 23.87% of the MCS data and was motivated by low phonic salience verbs, subject position (distantly preceding the verb and post posed), animate referents, participants over the age of 30, and males. For both features, MCS speakers used salience and L1 influence as strategies to compensate for differences that exist between Spanish and Miskitu regarding SPE and number marking.
The findings provide evidence for incomplete acquisition of Spanish by MCS speakers. The low rate of explicit subject pronouns and variable subject-verb agreement are likely remnants of the rapid language shift to Spanish that took place in the past but have now become fixed features of MCS. Younger speakers, however, are approximating to distributions found in NS, which suggests that these features are currently changing in MCS.