Subject pronoun expression (SPE) is one of the most studied variables in Spanish sociolinguistics, in the production of both monolingual and bilingual speakers. In the case of U.S. Spanish-English bilinguals, some studies suggest that the almost categorical use of English overt subject personal pronouns (SPPs) boosts the SPP rate in the Spanish of these speakers, whereas some other studies do not support such claim. Aiming to shed light on this subject, the present analysis of SPE in Phoenix is a variationist contribution to the literature on bilingual SPE in the U.S. Although the factors favoring SPP occurrence in the Spanish of Phoenix coincide for the most part with those attested across Spanish varieties, compared to other U.S.-bilingual Spanish varieties accounted for, it shows the lowest rate of overt SPPs: 17.8%, even lower than monolingual Mexican varieties. In addition, Spanish-dominant speakers in Phoenix favor overt SPPs, whereas English-dominant bilinguals disfavor them. When separate regression analyses were run for each proficiency group, it was found that both share virtually the same constraints contributing to the presence of the overt variant. These results do not support the contact claim. Since the Phoenix bilinguals pattern with other U.S. Southwest communities with regard to the aforementioned trends, it is suggested that proximity to the border may neutralize the SPE English effect in the Spanish of these communities, in contrast to communities that lack such regional traits, like New York and Florida.

Objectives and Research Questions: This study explores the effects of bilingualism on the production of subject personal pronouns (SPPs) in speakers of two null-subject languages, Quechua and Spanish. The paper also seeks to determine if these effects can be explained by general bilingual accounts, such as the Interface Hypothesis (IH), or by contact-specific accounts. Methodology: This is a sociolinguistic variationist study; therefore, the data were collected with sociolinguistic interviews. Data and Analysis: The data consist of transcriptions of audio recordings of eight Spanish monolingual and eight Quechua-Spanish bilingual speakers of Huancayo (Peru). The data were analyzed by using the statistical software SPSS 23.0 and Goldvarb X. Findings: The IH predicts that the overt SPP rate of the bilinguals should be higher than that of the monolinguals and that the pragmatic switch reference constraint should be difficult for the Quechua first language speakers to master. The results show, however, that their rates are similar, and that switch reference was the most robust predictor for the bilinguals. This study’s results suggest that indirect transfer from the mandatory Quechua switch reference subordinating particle – pti is taking place. Originality: This is one of the first variationist studies examining the IH predictions regarding SPP production in bilinguals speaking an indigenous American null-subject language alongside Spanish. In addition, this is the first study to show, through statistical analyses, the contact-specific effect the other language can exert on a particular constraint in the subject pronoun expression of the bilinguals. Significance: The results of the study suggest that even subtle transfer in situations of language contact can be accurately explained by contact-specific accounts.

This study is one of the first variationist accounts of subject pronoun expression (SPE) in Andean Spanish, in particular, in the variety spoken in Huancayo, Peru. The data analyzed consist of sociolinguistic interviews of sixteen participants, equally divided by sex.The results show that the overt subject personal pronoun (SPP) rate in Huancayo Spanish is the lowest one attested: 16.2%, even lower than that of Lima, the capital (16.8%), which situates Andean Spanish among conservative varieties regarding SPE.In order to determine if Huancaínos are following Limeños as a model in their SPE use and therefore sharing the same constraints, I conducted two separate regression analyses: one for Huancayo and one for the Limeño acrolect using Goldvarb X. The results show that Lima's dialect is sensitive to two factor groups that are neutralized in Huancayo, which suggests that Huancaínos do not take Limeños as a role model for their overt SPP production.

Abstract This variationist study analyzes the linguistic and social factors constraining the alternation of resumptive pronouns (RPs) and gaps in direct object (DO) relative clauses (RCs) in the Peruvian Limeño variety. Using a number of mixed-effects (logistic regression) models in Rbrul, results reveal that the set of linguistic constraints favoring pronominal DO resumption does not coincide with those reported to promote subject and oblique RP presence in previous studies. Furthermore, when compared to their subject and oblique counterparts, DO RPs are constrained by a higher number of factors of syntactic, semantic, processing and pragmatic nature. I suggest that this sensitivity to a broader set of constraints is crucial in explaining why DO RPs are more frequent in RCs than subject and oblique RPs. With respect to the social factors analyzed, this study shows a lack of effect of gender, age and education on the speaker’s choice for the resumptive variant.

No abstract