Spanish Modal Infinitival Constructions in English-Spanish Bilingual Grammars

The aim of this thesis is twofold: first, to determine whether a group of Spanish-native adult speakers and six groups of English-Spanish bilinguals are able to differentiate modal periphrases (MP), as in debo/puedo comer (I must/can eat), from nominal infinitival clauses (NIC), as in prefiero/deseo comer (I prefer/wish to eat). These two structures are superficially similar (inflected verb + infinitive) and have modal meaning, but differ in their internal structure (Gómez Torrego, 1999; RAE-ASALE, 2011; Vivanco, 2019). The second aim is to determine whether there is a pattern of inclusion/exclusion of the English infinitival marker to in intra-sentential English-Spanish code-switching in AUX/MOD-V and V-V sequences which depict MPs, NICs, and the periphrastic future (PF). To achieve the first goal, we designed a written Grammaticality Judgment Task (GJT) where participants were asked to judge sentences containing three syntactic processes that differentiate MPs from NICs. This GJT was administered to 96 participants: (i) 10 Heritage Spanish speakers, (ii) 10 Heritage English speakers, (iii) 14 L1Spa-L2Eng Immigrant speakers, (iv) 10 L1Eng-L2Spa Immigrant speakers, (v) 22 L1Spa-L2Eng Students, and (vi) 10 L1Eng-L2Spa Students. A Control group (CG) of 20 native Spanish speakers, born, raised, and residing in Spain, was also recruited. The results of the first task showed positive overall performance (84% accuracy). There were no significant differences between the groups, but Grammaticality and the three syntactic processes played a significant role in determining Accuracy. Our participants' performance was better with grammatical than with ungrammatical items, except for the Heritage English group, who were more accurate than the rest when judging ungrammatical MPs. As for the three syntactic processes, 'That' sentences yielded a significantly higher level of accurate answers, followed by 'What' and 'Demonstrative' sentences. Despite the significant role played by Grammaticality and the syntactic processes, the high rate of accurate responses and the lack of significant differences between the overall performance of an ample spectrum of English-Spanish bilinguals could indicate the acquisition of MPs and NICs, characterized by their lack of saliency, their variety of grammatical constraints, and their semantic complexity. To investigate the second goal, we designed a written Two-alternative Forced-choice task (2AFC), where participants were asked to choose one sentence within pairs of Spanish-English code-switched infinitival constructions (MP, NIC, and PF) with and without the English infinitival marker to (Spanish a in PF) between the inflected verb and the infinitive. This task was administered to the 76 participants in the bilingual groups but not to the native Spanish Control group. We also compared the experimental data with the corresponding patterns observed in a longitudinal corpus of two Spanish heritage siblings (4 and 8 years old). The results of the second task showed a significant pattern of exclusion of the infinitival to in MPs and of inclusion of to/a in the PF. With respect to NICs, the probability of including to was above 0.5, at a significant distance from the other two. There were no significant differences between the groups of bilinguals, and they were more likely to exclude to when the inflected verb was in Spanish. These results were partially in accordance with the patterns detected in the naturalistic data: the exclusion of to in MPs, the inclusion of a in the PF, and the exclusion of to in NICs when the inflected verb was in Spanish. There were no code-switched infinitival constructions with an English inflected verb in the corpus. These results contribute new data to discerning potential patterns that govern AUX/MOD-V and V-V code-switched sequences and provide a new perspective on the debate concerning switches between functional and lexical categories.