Comprehension and Acceptability of the English Resultative and Depictive Constructions by Spanish Native Speakers

CELI, MARIA ALEJANDRA

Leiden

Conferencia; Conference of the Student Organization of Linguistics in Europe (ConSOLE) 29th Edition; 2021

Leiden University Centre for Linguistics, Leiden University

Typologically different languages like English and Spanish use completely different linguistic strategies to capture events in the world (Talmy, 2000). The Resultative Construction (ERC) follows the English style for building event representations and thereby has no mirror equivalent in Spanish. This absence has significant implications for its acquisition by Spanish native speakers learning English as foreign language (EFL). The hypotheses of this research are, first, that ERC will be acquired at a later stage than the Depictive Construction, which does exist in both languages; second, ERC has different subtypes that will be learned in a sequence according to their degree of departure from the Spanish pattern. We conducted two experiments -the first one a comprehension task and the second, an acceptability task- in order to test those hypotheses, which were finally backed up by them. English is a satellite framed language in which verbs tend to codify manner, whereas Spanish is a verb framed language in which verbs tend to codify result. English conveys result by means of the English Resultative Construction (ERC), a complex predicate structure with a main verb that codifies manner and a particle (either a Prepositional Phrase or an Adjectival Phrase) that codifies result as a change of state as shown in (1)(1)She hammered the metal flat. Spanish templates do not allow this type of syntax-semantics interface since result is codified in the verb and, in order to express manner, an additional Prepositional Phrase is needed (Paris, 2015) as in (2)(2)Ella aplanó el metal a martillazos.She flattened the metal to hammerings.´She hammered the metal flat’Spanish, however, has an equivalent syntactic structure to the ERC, namely, the Depictive Construction (DC) as in (5), which mirrors the English Depictive Construction (EDC) as shown in (6)(3)He returned the book damaged. ´El devolvió el libro dañado´Semantically, this structure differs from ERC since the two predicates are not related in a causal chain as in the ERC; they only overlap on a given subinterval of time. Therefore, given the absence of the RC in Spanish, our aim is to describe how native speakers of Spanish who learn EFL comprehend ERC in contrast to EDC, and what effects of L1 on L2 can be found in the different acquisition stages, if any. More specifically, the purpose of this study is, first, to evaluate the comprehension of the different types of ERC and EDC by Spanish speakers; second, to determine the correlation between the comprehension of ERCs and EDC with the level of proficiency in L2; and, third, to assess the possible transfer effects from L1 to L2 that might facilitate the acquisition of EDC. We predict that the proficiency level would improve the comprehension of the target structures in general, but this effect would interact with type of ERC. If transfer effects may take place, English structures closer to Spanish (EDC) would be easier to comprehend when compared to ERCs that do not have a parallel construction in Spanish.We conducted two experiments consisting on two independent online questionnaires. Both of them included a proficiency questionnaire and the vocabulary LexTALE test (Lemhöfer, K & Broersma, M., 2011). The first experimentwas a comprehension task designed to measure the interpretation of three types of ERC (Property, Path and Fake Reflexive) and the EDC by native speakers of Spanish with different levels of proficiency in English as L2 (Low, Intermediate, and High). 285 Subject were exposed to 36 sentences and they had to decide the best interpretation of the target sentence out of 4 multiple choice paraphrasis. To test whether the rates for correct responses would change depending on Construction Type and its interaction with proficiency (Vocabulary LexTALEscores), we ran a mixed-effects logistic regression model which turned out significant (χ2 (3) = 68.246, p < 0.0001). Post hoc analysis indicated that response accuracy increased with LexTALE scores for all construction types (z’s < -3.360, p’s < 0.0008). When we compared accuracy within different levels of proficiency, we observed better comprehension performance for EDC than for Property and Fake Reflexives ERC (z’s > 3.58, p’s < 0.001) at low proficiency level, but this was not true for intermediate or high level. Across the range of LexTALE scores, EDC did not differ significantly from Path ERC.In addition, a second experiment was applied in order to investigate whether native speakers of Spanish with different levels of proficiency in English as L2 would differ in the acceptability ratings of grammatical and ungrammatical ERCs. By means of an Acceptability Judgement Task (AJT) ranking from 1-7 Likert scale, 90 subjects rated 96 sentences including target conditions of Property, Fake Reflexive ERC, EDC (half acceptable and half with semantic violations) and fillers. We ran a series of linear mixed-effects regression models on correct and incorrect items separately. Again, we predicted that the closer the structure is to the Spanish DC, the more acceptable they would be rated. Both models turned out significant (χ2 (2)’s > 8.340, p’s < 0.015). Post hoc analysis showed that all three levels of vocabulary proficiency displayed the same pattern: higher acceptability ratings for EDC sentences compared to both Property and Fake Reflexive ERC (T’s > 4.578, p’s < 0.001), with no significant differences between the latter. In addition, acceptability increased with LexTALE scores for Property and Fake Reflexive (t’s < -2.604, p’s < 0.027) ERC, but not for EDC; which may indicate a ceiling effect as it is a structure present in L1. On the other hand, acceptability of incorrect EDC and Property ERC decreased with LexTALE scores (T’s > 2.398, p’s < 0.047, while incorrect Fake Reflexives remained unaffected. This may indicate that Fake Reflexives are far more complex to accept and reject in their semantically violated form than any other ERC and EDC even in high proficient levels. Taken together, results show that the target structures are comprehended differently being the EDC the most easily understood and accepted structure in all proficiency levels (EDC>Path>Property>Fake Reflexive ERC). As vocabulary proficiency increases, comprehension of Path, Property and Fake Reflexive ERC improves. EDCs are easier for Spanish native speakers than any subtype of ERC, which may be due to transfer effects from L1 to L2. In addition, error discrimination improves with vocabulary proficiency for EDC and Property ERC, but not for Fake Reflexives. It seems that Fake Reflexives are syntactically more complex than Property and PathERC, and this difference may play a role in the comprehension and acceptability of this complex structure by Spanish speakers learners of EFL.LINK DE ACCESO A LA PONENCIA: https://view.genial.ly/6007068b7c10780d2ad553b5/presentation-console