Session 2E

Aim: This study explores the interaction between Differential Object Marking (DOM) and word order within relative clauses (RC), and compares monolingual Spanish and the bilingual variety of Spanish spoken in Catalonia. Motivation: DOM is a phenomenon by which some DOs are marked explicitly with a grammatical element (Bossong, 1985), the marker ‘a’ in Spanish and Catalan. In Peninsular Spanish, DOM is compulsory with human DOs (Juan ha visto a la niña), optional with animal DOs (Juan ha visto (a) la gata) and ungrammatical with inanimate DOs (*Juan ha visto a la mesa) (RAE, 2009). In Standard Catalan, DOM has a much more restrictive distribution: only personal pronoun DOs show DOM (L’he vist a ell) (IEC, 2016). As for RCs, interestingly, when the subject and the object have the same number and person features, the only way to disambiguate subject (SVO) and object (OVS) RCs introduced with the reduced pronoun que is the ‘a’ marker ((3a) vs. (4a)). On the other hand, when the marker is optional the result is an ambiguous RC ((3b) vs. (4b)). (3) a. La leona que atacó a la cazadora por error. (SVO, human DO) ‘The lion that attacked the hunter by mistake.’ b. La cazadora que atacó (a) la leona por error. (SVO, animal DO) ‘The hunter that attacked the lion by mistake.’ (4) a. La leona que atacó la cazadora por error. (OVS, animal DO) ‘The lion that was attacked by the hunter by mistake.’ b. La cazadora que atacó la leona por error. (OVS, human DO) ‘The hunter that was attacked by the lion by mistake.’ Prolonged contact between languages can lead to crosslinguistic influence (Silva-Corvalan 2014). In the case of DOM between Catalan and Spanish, there are few studies focusing on this topic (Puig-Mayenco et al., 2017; Perpiñán, 2018). Thus, addressing the crosslinguistic influence of DOM within RCs is innovative in two ways: no study has been done on how they interact and this topic is a good test since, as mentioned above, the interpretation of the RC object relies solely on the presence/absence of DOM. Methodology: Thirty-two Spanish-Catalan bilingual speakers and 32 monolingual speakers (age= 19-25) participated in a self-paced reading task with two conditions: order of the RC (SVO/OVS) and type of object (human/animal), as the examples in (3)-(4) show. Two measures were used: accuracy in comprehension questions regarding the interpretation of the RC and reading times (RTs) at different regions. Results: The offline data (comprehension questions) show that both groups are sensitive to DOM, since SVO sentences (which are DOM marked) are interpreted more correctly than OVS sentences (which are not marked). OVS sentences with a postverbal animal subject, which are ambiguous, are usually interpreted as SVO sentences, especially among Catalan-Spanish bilinguals. Perhaps bilinguals are following Catalan patterns, a language with (virtually) no DOM. This is confirmed by RT online data: OVS sentences with a postverbal animal subject, despite their complexity due to the altered order, are processed faster than OVS with a postverbal human subject (particularly by bilinguals), indicating that they are being misinterpreted as SVO. Overall, these findings suggest bilinguals are somewhat influenced by their L1 and that influence varies depending on the semantic properties that govern DOM distribution.

The aim of this paper is to present how highly proficient non-native Swedish L1 speakers of French and Spanish L2 use left and right periphery utterances in French and Spanish. Syntactic peripheries have drawn attention in linguistic research (Beeching and Detges, 2014), but less in the field of SLA. Romance languages have, if not the opposite kind of preferences than Swedish, at least very different patterns, in the organization of the elements preceding or following the core sentence. The components in the beginning of an utterance belong to the left periphery (LP), and consequently, at the end of the utterance, there are components belonging to the right periphery (RP). These are distributed according to the preferred use in a specific language. Spanish and French are considered to be languages with high degree of LP while Swedish has a tendency to put more information in the RP (Fant, Bartning & Österberg, forthcoming; Fant, 2019). The question whether acquisition of complex linguistic structures is attainable or not at very advanced levels of L2 use has been pointed out in several previous studies addressing the mastery of the syntax-discourse interface and discourse complexity (Bartning, Forsberg Lundell & Hancock, 2018). Not only morpho-syntactic features are acquired late, but also pragmatic structures e.g. formulaic speech. Research questions: • To what extent do French and Spanish L2 users acquire the patterns of peripheralization used by the natives in spoken language? • To what extent do French and Spanish L2 users transfer the Swedish patterns of peripheralization in spoken language? • What similarities/differences are found between French and Spanish L2 users regarding the patterns of peripheralization? The corpus, "High proficiency in the user of the L2” (Hyltenstam, Bartning and Fant, 2018, recruited through convenience sampling), consists of semi-structured interviews (genre: self-presentation) with ten native speakers of Spanish and French, and, correspondingly, ten nonnative speakers in Spanish and French (L1: Swedish), and five with Swedish natives (“Swedia”), in total 55 interviews/ 110 759 words. The peripheries were analyzed from a syntactic-pragmatic perspective and 14 categories were defined independently and intersubjectively until consensus was reached. First, the three L1 were compared (frequencies and length of peripheries), secondly, the advanced L2 speakers were compared to the native speakers and to each other in a GLM univariate analysis. Different discourse-organizational preferences were observed with significant differences at the group level confirming a strong tendency to peripherize to the left in Spanish and French and to the right in Swedish. French and Spanish L2 users display an almost total alignment with native use, very different from Swedish L1 users. Hardly any cross-linguistic transfer was to be found. Differences interpreted as interlanguage features were found, concerning pragmatic functions as mitigating and intensifying, an overuse of pronouns and a noticeably lower use of very long sequences by the French L2 speakers. BARTNING, I., FORSBERG LUNDELL, F., & HANCOCK, V. (2012), On the role of contextual factors for morpho-syntactic stabilization in high level L2 French, Studies in SLA, 34, 243-267. BEECHING, K. & DETGES, U. (eds) (2014), Functions at the left and right periphery: Cross-linguistic investigations of language use and language change, Brill. FANT, L., BARTNING, I. & ÖSTERBERG, R. (forthcoming). The left and right periphery in native and non-native speech – a comparative study between French L1/L2, Spanish L1/L2 and Swedish L1. IRAL. FANT, L. (2019), Las periferias sintácticas en el habla coloquial de español L1 y L2 y de sueco L1: un estudio tipológico y adquisicional. Normas Revista de Estudios Lingüísticos Hispánicos. HYLTENSTAM, K., BARTNING, I., & FANT, L. (eds) (2018), High level proficiency in second languages and multilingual contexts. Cambridge University Press.

The human brain receives tremendous amounts of data every minute. To cope with this cognitive burden efficiently, the brain employs a myriad of mechanisms, including prediction. Relevant to the current study, native speakers use determiners to predict the gender of upcoming nouns (e.g., Lew-Williams & Fernald, 2007, 2010, Hopp, 2012, 2016; Dussias, Valdés, Guzzardo, & Gerfen, 2013), but adult L2 learners exhibit persistent difficulties wit (e.g., Lew-Williams & Fernald, 2010; Grüter, Lew-Williams, & Fernald, 2012; Hopp, 2012, 2016; Dussias et al., 2013; Halberstadt, Valdés, & Dussias, 2018). Some accounts attribute the L2 learners’ struggle to their weak knowledge of noun gender assignment (Lexical Gender Learning Hypothesis, Grüter, Lew-Williams, & Fernald, 2012), whereas others ascribe it to how the L2 parser adapts features of L2 words (Parser-as-Language Acquisition Device Hypothesis, Dekydtspotter & Renaud, 2009, 2014). Importantly, L2 learners’ knowledge of noun gender assignment increases with proficiency (Hopp, 2016), and their ability to predict nouns improves with higher proficiency (Hopp, 2012), transparent noun suffixes (Halberstadt et al., 2018), and more gender cues (semantic and syntactic priming, Fowler & Jackson, 2017). Previous studies examine prediction of nouns using gender cues of determiners, but are unable to determine the role of the type and number of cues on predictive processing. To this end, we examine the role of knowledge of noun gender assignment, noun gender transparency, number of gender cues, and L2 proficiency on the use of gender suffixes in determiners and nouns to predict adjectives. Spanish monolinguals and intermediate and advanced adult English learners of Spanish completed a visual-world paradigm eye-tracking task. In this task, they read two adjectives on the screen (nuevo ‘newMASC.SG’; nueva ‘newFEM.SG’), listened to sentences with gendered and genderless determiners (Dice que (elMASC.SG/su) carro nuevo cuesta mucho dinero ‘He/She says that (theMASC.SG/his-her) newMASC.SG carMASC.SG costs a lot of money’) and indicated which adjective they heard. Participants also completed a language background questionnaire, a Spanish proficiency test, and a vocabulary test to assess knowledge of noun gender assignment. The results reveal that all groups accurately predicted the adjectives. However, the intermediate learners relied less on determiners and adjectives with opaque gender than the advanced learners and the natives. These findings support the Lexical Gender Learning Hypothesis (Grüter et al., 2012), and advance our understanding of how the human brain employs prediction to facilitate information processing in a L2.

Although L1 syntactic transfer robustly surfaces in off-line comprehension among adult L2 learners (e.g. Grüter & Conradie, 2006; Rankin, 2014) as well as in cross-linguistic syntactic priming (e.g. Nitschke et al., 2014), it remains unclear whether and when L1 effects emerge during L2 sentence processing (Clahsen & Felser, 2018). Previous studies found no evidence of online L1 influence (Rankin et al., 2019) or L1 effects only at lower proficiency levels (Hopp, 2017b) or in language-mixing contexts (Hopp, 2017a; Jacob, 2009). To examine the temporal dynamics of L1 activation in L2 sentence processing, we conducted a visual-world eye-tracking experiment combined with cross-linguistic syntactic priming. 28 L1-German and 28 L1-Japanese intermediate-to-advanced learners of English listened to English subject and object wh-questions (targets; (1a&b)), and selected the target picture in a display presenting both interpretations (see Figure 1). Critically, English subject-questions (1a) are ambiguous between subject and object-questions when mapped onto German verb-second syntax (1c), yet cannot be accommodated by Japanese SOV syntax (3). To prime subject or object interpretations, each target question was preceded by unambiguous wh-questions in the respective L1 (primes; (2 & 3)). (1) a. Which animal pushes the cat? (English subject question) b. Which animal does the cat push? (English object question) c. Welches Tier schubst die Katze? (German subject/object question) which animalNOM/ACC push-3sg the catACC/NOM (2) a. Wer schubst die Katze? (German subject question) whoNOM push-3sg the cat b. Wen schubst die Katze? (German object question) whoACC push-3sg the cat (3) a. dare ga neko o oshiteimasuka (Japanese subject question) whoNOM catACC push Q b. neko ha dare o oshiteimasuka (Japanese object question) whoACC catNOM push Q GLMERs on participants' final interpretations of English questions revealed a Group-by-Question-Type interaction, yet no cross-linguistic priming effects (Table 1). L1-German learners had lower accuracy on English subject questions (M_Subj = 93.7%; M_Obj = 98.1%), consistent with L1 influence, while L1-Japanese learners scored lower on English object questions (M_Subj = 95.7%; M_Obj = 70.6%). By contrast, eye-movement patterns were similar across groups, but showed significant cross-linguistic priming effects, albeit only for object-questions. In the L1 Japanese group, the priming effect was delayed compared to the L1 German group. Based on the eye-tracking data, we conclude that processing of wh-questions seems cross-linguistically governed by parsing principles favouring subject-questions, and can be affected by cross-linguistic priming of the dispreferred object-questions. As seen in the interpretation data, L1 transfer only surfaces late in off-line interpretation, and it is not modulated by cross-linguistic priming. These findings suggest that the L1 grammar only gets selected late, namely, when participants assign a global interpretation to the parse; yet, it does not guide incremental parsing. We discuss the results in the context of current approaches to L2 sentence processing. Selected References Clahsen, H., & Felser, C. (2018). Some notes on the shallow structure hypothesis. Studies in Second Language Acquisition, 40(3), 693-706. Hopp, H. (2017a). Cross-linguistic lexical and syntactic co-activation in L2 sentence processing. Linguistic Approaches to Bilingualism, 7(1), 96-130. Nitschke, S., Serratrice, L., & Kidd, E. (2014). The effect of linguistic nativeness on structural priming in comprehension. Language, Cognition and Neuroscience, 29(5), 525-542. Rankin, T. (2014). Variational Learning in L2: The Transfer of L1 Syntax and Parsing Strategies in the Interpretation of wh-questions by L1 German learners of L2 English. Linguistic Approaches to Bilingualism, 4(4), 432-461. Rankin, T., Grüter, T., & Hopp, H. (2019). Investigating co-activation of L1 syntax during processing of wh-questions: eye-tracking evidence from L1 German-L2 English. In R. Slabakova, J. Corbet, L. Dominguez, A., Dudley, & A.J.F. Wallington (Eds), Explorations in second language acquisition and processing (pp. 154-170). Newcastle: Cambridge Scholars.