|Abstract:||This study investigated comprehension of complex syntax in Spanish, namely passive and relative clauses, by early English-Spanish bilinguals, also known as heritage language (HL) learners, and by second language (L2) learners. To contribute to current theoretical debates on the role of input and age in L2 and bilingual language acquisition, and to current pedagogical inquires on how instruction can be beneficial for HL learners, this study sought to find out whether age of onset of meaningful exposure (and consequently type of L2/HL input and experience) affected adult L2 learners and heritage speakers’ knowledge of these structures, which are acquired early (by age 3 or 4) but are most frequently used and mastered at a later age during the school-age period. This study also investigated whether language learning experience, for late bilinguals after adolescence as a second language (L2) in a classroom; for early bilinguals, from birth as heritage language (HL) in informal familiar settings, affect their performance in tasks of oral and written modalities. Given that these structures are mastered during the school-age period, this study also focuses on how instruction in Spanish from pre-K to graduate school affect comprehension of these clauses.
Relative clauses are embedded clauses that modify a noun phrase (NP), the head of the main clause, which in this study were either in the subject or in the object position. In Spanish, relative clauses involve a type of wh-movement where a special null wh-word referred to as an operator (Op) in the deep structure moves to the Spec of the CP. When the operator moves, it leaves a trace or a gap within the clause. Although many factors affect processing of relative clauses, it is generally agreed that the linear distance (LD) between the head and the gap accounts for the ease or difficulty processing subject relative and object relative clauses, at least in head-initial languages. The relative clauses in this study were subject and object relative clauses with inanimate NPs and reversible and plausible contexts in which either the subject or the object could perform the action described by the verb in the complementizer phrase (CP). The sentences were designed to control for plausibility to prevent a context bias favoring one interpretation or another. The only way for the learners to determine which NP was the subject or the object in the clause was by processing the inflectional morphology of the verb that agreed with the subject. Only third person singular and plural verb conjugations were used.
In examples (1) and (2) the linear distance in words between the head and the gap is shorter in the subject relative clause than in the object relative clause.
(1) Subject Relative Clause (SR)
The submarineSUBJECT that ____ sankVERB the boatsOBJECT. LD = 1 word
(2) Object Relative Clause (OR)
The submarineOBJECT that the boatsSUBJECT sankVERB _____. LD = 4 words
While English relatives are constructed with one word order, Spanish relatives can be constructed with two, showing different linear distances between the head and the gap. Relative clauses in examples (3) and (4) have the same word order as English relative clauses.
(3) SR El submarinoSUBJECT que _____ hundióVERB los barcosOBJECT. LD = 1 word
(4) OR El submarinoOBJECT que los barcosSUBJECT hundieronVERB _____. LD = 4 words
And examples (5) and (6) show the other possible word order for SR, with Verb-Object inversion, and for OR, with Subject-Verb inversion, respectively. The SR in (5) is extremely rare in the input.
(5) SR El submarinoSUBJECT que los barcosOBJECT _____ hundióVERB. LD = 3 words
(6) OR El submarinoOBJECT que hundieronVERB _____ los barcosSUBJECT. LD = 2 words
Object relatives OR (OSV) were expected to be harder to comprehend than SR (SVO) based on the linear distance between the head and the gap. Furthermore, the ease of comprehension of relative clauses was expected to vary based on the LD in words in the following manner: SR(SVO) >OR(OVS)>SR(SOV>OR(OSV), where ‘>’ means ‘easier than’.
The other structure tested was passive clauses. Passive clauses included verbal and adjectival passive clauses with the copulas in the imperfect tense era and estaba, respectively. See (7) and (8). Comprehension of verbal passive clauses, as opposed to the comprehension of adjectival passive clauses, was predicted to be more difficult for both groups of leaners. First, there is a canonical construction for the verbal passive clause in the past with the copula in the preterite tense fue, with which learners are more familiar. Second, another passive voice construction, the reflexive passive, also known as morphological passive or se-passive, is more frequent in Spanish. Third, the verbal passive clause with the copula in the imperfect tense in HL and L2 acquisition is not common in oral communication. And, last, the imperfect tense is vulnerable to incomplete acquisition in L2 and HL learners’ grammars.
This study was designed with truncated passive clauses with actional verbs and irreversible contexts, meaning that the theme could not perform the action stated by the verb. Instead the theme was always the object of, or was in a state resulting from the action stated by the verb.
(7) La comida estaba servida. Adjectival/ Stative Passive
dinner was (estar.IMPERFECT) served.
‘Dinner was served.’
(8) La comida era servida. Verbal/Eventive Passive
dinner was (ser.IMPERFECT) served.
‘Dinner was being served.’
Unlike adjectival passives, verbal passive clauses trigger a reanalysis of the sentence to organize the thematic roles. To comprehend adjectival/stative passives with participles, learners had to realize that a description of a state or final result followed the copula estar. To comprehend verbal passive clauses, learners had to know that the first noun phrase was not the agent, but the object. The reanalysis is triggered at the past participle in English (Mack, Meltzer-Asscher, Barbieri, & Thompson, 2013). In this dissertation it is assumed that Spanish reanalysis involves the copula. If the copula was ser, the clause was a verbal passive not an adjectival passive. When processing the copula, learners had to integrate the meaning of the imperfect form era, which refers to an ongoing or habitual action in the past. L2 and HL learners are familiar with passive voice structures and they have been shown to be familiar with the canonical passive clause with fue in Spanish. Thus, comprehension of these passive clauses required not only knowledge of its syntax, but also of the complementary distribution of ser and estar. Acquiring the complementary distribution of the copulas is a hard task for L2 learners and HL learners as they need to acquire not only their irregular inflectional morphology but also the contrasts between each copula regarding syntax, semantics and pragmatics, a contrast that is not present in English.
116 participants completed an aural and a written version of a Picture Matching Task created to test comprehension of these clauses. They also completed an aural and written version of a Grammaticality Judgment Task to test basic grammar knowledge needed to comprehend these structures, and their linguistic history was recorded in a bilingual language questionnaire. Their proficiency was measured with a written test (DELE) and an oral narrative coded to calculate the moving average type-token ratio (MATTR), mean length of utterance (MLU) and fluency in words per minute.
Results showed that SR(SVO) and OR(OSV) were both easily comprehended by both groups of learners. The fact that SR(SVO) and OR(OSV) show the same word order as their English counterparts might explain the result. A study with speakers of another L1 with different word order (i.e. Chinese, Korean, Turkish) would be needed to confirm this possibility with more certainty. Contrary to the prediction based on the LD, comprehension of SR(SOV) was low, which is explained by how infrequent they are in the input. OR (OVS) were interpreted as SR (SVO) by L2 and HL learners of lower proficiency. As written DELE proficiency increased so did correct interpretation of OR(OVS) for HL and L2 learners. The prediction based on the LD was not borne out and highlights the importance of extra-linguistic factors in the comprehension of these clauses. Frequency of the structure affects comprehension of relative clauses, and so does written and oral proficiency.
Results for passive clauses showed that L2 and HL learners were more accurate with adjectival passive clauses than with verbal passive clauses. These difficulties in the comprehension of verbal passive clauses show that learners have not yet acquired the full spectrum of copula uses and interpreted era in the verbal passive clause as estaba in an adjectival passive clause, a simpler construction.
Besides comprehension of these structures, this study investigated how age of meaningful exposure to Spanish, modality of task, and instruction in the HL/L2 language from pre-K to graduate school affected comprehension of these clauses.
Results from the comprehension of passive and relative clauses show age of meaningful exposure to Spanish as a robust predictor of successful language acquisition. HL learners were significantly more accurate than L2 learners in the comprehension of passive clauses and significantly more accurate in the comprehension of the object relative clause OR-OVS. This indicates that early exposure to Spanish confers HL learners an advantage in the comprehension of clauses with complex syntax.
Regarding task modality, L2 and HL learners’ linguistic experience influences their abilities with language tasks because each experience emphasizes different language skills (Bowles, 2011b; Montrul, 2016). L2 learners’ meaningful exposure to Spanish in the classroom helps them develop metalinguistic or explicit knowledge of the language, and hence they tend to perform well in tasks that target metalinguistic knowledge. HL learners also have limited meaningful exposure to the language, but their exposure is mostly aural in informal settings with family and community. They do not have significant experience writing or reading the HL, and, typically do not have metalinguistic knowledge of the HL (Montrul, 2008b, 2016). Because their experience with the language is mostly aural, they tend to perform better in aural tasks that tap on implicit knowledge, or on intuitive information (Bowles, 2011a; Montrul, Foote, & Perpiñán, 2008a; Montrul & Perpiñán, 2011). Results showed that modality affected comprehension of both groups of learners but with different clauses. L2 learners performed significantly better in the written modality with relative clauses, while HL learners performed significantly better in the aural modality with passive clauses.
This thesis also focused on how instruction interacted with age of meaningful exposure in the acquisition of complex syntax. Results showed that instruction significantly accounted for L2 learners’ comprehension of passive clauses, but not for comprehension of relative clauses. Although instruction did not account for HL learners’ comprehension of any of the clauses, an important finding of this dissertation is that instruction in Spanish could be more fruitful for HL learners if it starts in elementary school. Early instruction and language use at an early age conferred advantages to HL learners in the comprehension of verbal passive clauses with era. The data suggested that parental Spanish use with learners (input), combined with learners’ Spanish language use with parents (output), and early frequent Spanish use (between the ages of 6 and 10) positively influenced comprehension of verbal passive clauses with era. Moreover, the data suggested that early bilingual education could be a contributing factor too.