This dissertation project examines bilingual speech, specifically subject expression in Catalan-Spanish bilinguals in Minorca, Spain.
Syntactic-theoretical treatments of subject expression abound, most concerned with the formal features that license null subjects. Thus, cross-linguistic differences between null subject languages and non-null subject languages have been attributed to the features of functional projections such as Agr and Tns. While null subjects are licensed by properties of the 'core' or narrow syntax of null subject languages, the appearance of overt subjects in such languages is constrained by the informational context. Thus, in contrast to null subjects, the expression of overt subjects in null subject languages is regulated by properties of the 'peripheral' grammar, specifically, the area where the syntax interfaces with the discourse/pragmatics. The works of Antonella Sorace and her colleagues indicate that while the core syntax remains unchanged in language contact situations, the peripheral grammars (i.e., the syntax interfaces with other modules) are susceptible to inter-lingual effects. For instance, while Spanish-English bilinguals accept null subjects in Spanish (and reject them in English), they demonstrate the emergence of pragmatically-infelicitous overt pronominal subjects in Spanish.
The use of pragmatically-infelicitous overt pronominal subjects in Spanish has been interpreted in two manners: one view identifies this pattern with a simplification process or strategy of cognitive economy attendant to bilingual speech (Sorace 2004); the second view describes it as a convergent outcome, rendering the contact languages more alike (Bullock & Toribio 2004, Toribio 2004). These two approaches make distinct predictions for outcomes in situations of contact between two null subject languages. The simplification approach predicts an increase in overt subjects, as a consequence of a universal process that allows for bilinguals to compensate for the increased processing load of articulating different types of information (syntactic and discursive). In contrast, a convergence account would predict an intermediate rate of overt subjects between the rates observed in the two languages in contact.
This dissertation project expands on previous research by examining Spanish in contact with Catalan in Minorca. The project examines variation between null subject languages, by reference to data from two very closely related languages, and it explores bilingual outcomes, by comparing monolingual and bilingual Spanish data. The aim is to examine the role of language-internal and language-external factors in the emergence of contact-induced forms. To that end, naturalistic Spanish language samples were collected from twelve monolingual Spanish speakers, eleven Spanish L1 bilinguals, and twelve Catalan L1 bilinguals, who were recorded in Spanish; in addition, naturalistic Catalan data were gathered from twelve Catalan-dominant speakers (as there are no monolingual Catalan speakers). All language samples were recorded during an interview in which participants reported on their language history, participated in an ethnolinguistic survey, and responded to questions on language attitudes and ideologies. The recordings ranged from 49.40 minutes (4,369 words) to 99.70 minutes (11,399 words). Data were transcribed and the first 300 relevant tokens produced by each participant were coded and submitted to statistical analysis using Goldvarb X (Sankoff, Tagliamonte & Smith 2006). Apart from the overall rate of subject expression, eleven language-internal variables were considered in the distribution of null, overt pronominal, and overt lexical subjects: person and number, discourse function, distance from previous mention, focus, co-referentiality, clause type, animacy, semantic and syntactic verb type, tense continuity, and verb form ambiguity. In addition, seven language-external variables were considered: age, gender, place of residence, first language, second language proficiency, and first and second language use.
Results indicate that overall rates of overt pronominal subjects are not significantly different in Spanish and Catalan (10.6% and 11.9%). The bilinguals rates in overt pronominal subjects are not significantly different from those in monolingual Spanish (12.8% in the Spanish L1 and 12.6% in the Catalan L1 bilinguals), as predicted by Bullock & Toribio (2004) and Toribio (2004). However, multivariate regressional analyses reveal differences in constraint ranking between Spanish and Catalan. Interestingly, the bilinguals mostly display intermediate positions, i.e., convergent outcomes. Evidence of a bilingual continuum, where the Catalan L1 bilinguals are closer to the Catalan data than the Spanish L1 bilinguals, is also present in the patterns of subject expression, instantiated in constraint ranking differences. Apart from the effect found for L1, proficiency was also returned as significant. Catalan L1 speakers with higher proficiency in Spanish and Spanish L1 speakers with higher proficiency in Catalan display more monolingual-like patterns than those who exhibit more disparate proficiencies across their two languages. (Abstract shortened by UMI.)