You are here: Workshops / Cross-linguistic influence: determining onset and end
Français
English

Cross-linguistic influence: determining onset and end

 

Authors : Pedro Guijarro-Fuentes, Pilar Larraňaga

Title: Cross-linguistic influence: determining onset and end

Discussant: Natascha Müller from the University of Wuppertal has accepted to act as discussant.

 

Summary:

This workshop includes five oral presentations of 20 minutes each with additional 5 minutes discussion after each paper by the authors included below and does not plan to accept unsolicited papers. Prof. N. Müller will attend the workshop and will act as a discussant following the last presentation. The papers to be presented will be reviewed by two external reviewers as well as by the Chairs of the workshop. We do not plan to apply for external funding for the attendants/presenters.

Focusing on different language cognition interface phenomena, the talks, that encompass this workshop, will (a) discuss the manifestation of cross-linguistic influence (CLI), as a possible factor in deviations from universal developmental patterns occurred in a child bilingual grammar, and (b) examine as to when such manifestations start and end. Divergent bilingual populations and methodological approaches are to be used in these studies; all invited speakers will present longitudinal and cross-sectional data using a diverse range of methodologies.

 

Overview abstract:

Much research has been conducted since Hulk & Müller (2000) and Müller & Hulk (2001) postulated that bilinguals acquiring two languages may show structures in one language that seem to be influenced by the other language being acquired. From the research available, we know that cross-linguistic influence is to be found in the pragmatic domain provided that both languages share some superficial properties. The domains of subject-drop and the use of clitics have been identified as domains in which cross-linguistic influence occurs in a number of language combinations (Nicoladis, 2002; Paradis & Navarro, 2003; Genesee et al., 2005; Serratrice et al., 2009; Fernández Fuertes & Liceras, 2010 to name just a few). However, a number of issues has not been addressed as of today, i.e., the age of onset and duration of the cross-linguistic influence. Hence, the present workshop will address several issues. Specifically, the goal is to determine whether there is an age or stage of onset and end of cross-linguistic influence in the linguistic domains of subject and object acquisition. Moreover, it will be attempted to clarify whether language internal or external factors (e.g., specific structural features, language dominance and language proximity) constrain age of onset and end of the cross-linguistic influence as well as the language combination being acquired.

For the above reasons, we restrict the workshop to the phenomenon of subjects and clitics prone to cross-linguistic influence and the combinations of non-pro-drop languages such as English and German in combination with the pro-drop languages Spanish and Italian. In addition to these languages, a combination of Indo-European and non Indo-European 2 languages will also be investigated, i.e., Basque-Spanish/Basque and Romanian/Hungarian. This is the more so interesting for the language combination, since these grammatical domains are diametrically opposite.

In order to discuss the issues above we have invited the following scholars who have already accepted our invitation to participate. The workshop consists of 5 oral presentations (see below for complete list of names, affiliations and email addresses), which will last for 20 minutes each. The workshop will be opened by a 5-minute introduction by the session chairs and closed by a 20 minute discussion lead by the invited discussant, Prof. N. Müller. The summary discussion will lead the audience into the allotted 30 minutes for questions to be addressed to anyone on the panel (the chairs, paper presenters and discussants alike). The total length of the proposed workshop is of 3 hours.

 

List of invited contributors and emails:

 

Abstracts (which appear in order of presentation for the conference) and affiliations

1. Juana Liceras1 and Raquel Fernández-Fuertes2

Title: Subject omission/production in child bilingual English and child bilingual Spanish: The view from linguistic theory

Affiliation: (1) University of Ottawa, (2) University of Valladolid

Email: jliceras(at)uottawa(dot)ca

In bilingual child language acquisition research, a recurrent learnability issue has been to investigate whether and how interlinguistic influence would interact with the non-adult patterns of omission/production of functional categories (Müller, 1998; Döpke, 2000; Yip & Mathews, 2000; Hulk & Müller, 2000; Paradis, 2001; Nicoladis, 2002; Paradis & Navarro, 2003; Genesee et al., 2005; Serratrice et al., 2009; Fernández Fuertes & Liceras, 2010; Liceras et al., 2010; Liceras et al., 2011, among others). In this paper, we analyze the omission/production of subject pronouns in the developing English grammar and the developing Spanish grammar of three English-Spanish simultaneous bilingual children (FerFuLice corpus in CHILDES; Paradis & Navarro, 2003). We base this analysis on: (i) the divide between two different reformulations of the null subject parameter, Alexiadou & Anagnostopoulou‟s (1998) and Sheehan‟s (2006), as pointed out by Martínez (2011); and (ii) Liceras et al.‟s assumptions concerning the role of core syntactic phenomena in interlinguistic influence. We specifically aim at providing an answer to the following questions: (1a) What are the patterns of production/omission of English overt and null subjects that appear in the data of these three bilingual children?; (1b) How do these patterns compare with those of the English monolingual child in Sachs corpus in CHILDES?; 1c) Can we infer from those patterns that the default presence of null subjects in Spanish influences bilingual English so that it displays more null subjects and for a longer period of time than monolingual English?; (2a) What are the patterns of production/omission of Spanish overt and null subjects displayed by the three English-Spanish bilingual children?; (2b) How do these patterns compare with those of a Spanish monolingual child, María, from the Ornat corpus in CHILDES? and (2c) Can we infer from those patterns that the obligatory presence of overt subjects in English influences bilingual Spanish so that it displays more English-like overt subjects and for a longer period of time than monolingual Spanish?

 

2. Ludovica Serratrice* & +Antonella Sorace

Title: Bilingual and monolingual children‟s on-line interpretation of null and overt pronouns in Italian

Affiliations:*University of Manchester, +University of Edinburgh

Email: Ludovica.Serratrice(at)manchester.ac(dot)uk

In the absence of biasing morphological and/or semantic information, null third person subject pronouns in Italian tend to refer back to a topic subject antecedent, while overt third person pronouns privilege a non-topic non-subject antecedent (Carminati, 2002). Off-line data for children‟s interpretation of null and overt third person subject pronouns in Italian have confirmed this pattern of anaphoric dependencies for Italian monolinguals between the ages of 6 and 10 (Sorace, Serratrice, Filiaci & Baldo, 2009). Bilingual English-Italian children in the same study behaved differently depending on whether English or Italian was the language of the community. Younger bilinguals (6-8) in the UK, but not those in Italy, chose significantly more inappropriate topical subject antecedent for an overt pronoun, and so did the older Spanish-Italian bilinguals (8-10) in Spain. The referential properties of null subjects were unaffected by patterns of language exposure and all children, at all ages, overwhelmingly chose a topic subject antecedent.

In the current study we investigated on-line pronominal resolution of null and overt pronouns in monolingual Italian children, English-Italian bilingual children in the UK and Spanish-Italian children in Spain. During an eye-tracking experiment children‟s eye-movements were monitored while they listened to sentences like (1) and (2)

Null subject condition

(1) La nonna salute la nipote in cucina mentre Ø apre con calma la porta.

“The grandma says hello to the grandchild while she calmly opens the door.”

Overt subject condition

(2) Il contadino incontra il prete alla fattoria mentre lui accarezza con curiosita‟ un coniglio.

The farmer meets the priest at the farm while he strokes a rabbit with curiosity.

The results of a mixed effects model run for 10 200ms segments after the onset of the verb for the null pronoun condition showed a significant effect (p <.001) for antecedent between 400ms and 2000ms for all groups with more looks to the subject antecedent. The effects of Age (younger, 6-8; older 8-10) and Group (monolingual, English-Italian bilingual, Spanish-Italian bilingual) were never significant confirming that, regardless of age and language background, all children clearly preferred a topic subject antecedent for null pronouns. 

The mixed effects model for the overt pronoun condition revealed a significant effect of antecedent between 400ms and 1600ms (p<.001) with more looks to a non-subject non-topic antecedent for all children. Between 1200ms and 1400ms there was a significant interaction (p <.01) between antecedent and group with fewer looks for the non-subject non-topic antecedent for English-Italian bilinguals compared to the other two groups (p<.01). These on-line results confirm previous off-line findings for a possible short-lived effect of cross-linguistic influence from English to Italian in the choice of an antecedent for an overt pronoun. In contrast, we found no on-line evidence that Spanish-Italian bilinguals behaved differently from Italian monolinguals in their choice of antecedent for an overt pronoun.

 

3. Marisa Patuto

Title: Cross-linguistic influence in early bilingualism: The acquisition of subjects and the role of language dominance in German-Italian, German-Spanish and French-Italian children

Affiliation: Bergische Universität Wuppertal

Email: mpatuto(at)uni-wuppertal(dot)de, www.sprachwissenschaft.uni-wuppertal.de/patuto

Recent research on early bilingualism highlights the fact that target-deviant subject realizations in null-subject languages are interpretable in terms of cross-linguistic influence in the sense that inappropriate pragmatic decisions about subjectless environments are made (cf. Patuto 2008, in prep.; Schmitz 2007; Serratrice & Sorace 2003; Serratrice, Sorace & Paoli 2004; Serratrice 2007; Sorace & Filiaci 2006 among others). This result is replicated for eight longitudinally studied German-Italian, two German-Spanish and two French-Italian bilinguals. The present work is compatible with the view of cross-linguistic influence proposed by Hulk & Müller (2000) and Müller & Hulk (2000, 2001) according to which interface phenomena are affected by cross-linguistic influence and therefore delayed in acquisition (cf. Paradis & Genesee 1996, Patuto, Repetto & Müller 2011). The principal aim of this contribution is to determine whether cross-linguistic influence is due to syntactic or processing properties (cf. Sorace & Serratrice 2009). In this vein, the empirical investigation demands a syntactic and cognitive interpretation of the longitudinal data which will be evaluated against a predominant impact of language dominance in early bilingualism (cf. Gildersleeve-Neumann, Peña, Davis & Kester 2009). Additionally, the analysis is based on a precise syntactic analysis of the involved target systems. Controversy in the literature on whether strong pronouns and pro occupy the same syntactic position leads to the assumption that the involved null-subject languages may differ syntactically (cf. Cardinaletti 1997, Alexiadou & Anagnostopoulou 1998, Ordóñez & Treviño 1999, Poletto 2000, Carminati 2002, Alonso-Ovalle, Fernández-Solera, Frazier & Clifton 2002, Suñer 2003). Even if Italian and Spanish share the null-subject property, the data support the hypothesis of an underlying syntactic difference. Summarizing, this study relativizes the impact of language dominance and provides empirical evidence that syntactic derivation and language combination are relevant for the occurrence of cross-linguistic influence (cf. Müller & Patuto 2009).

 

4. Veronica Tomescu and Larisa Avram

Title: Romanian clitics and the Interface Hypothesis: The view from two Romanian-Hungarian simultaneous bilinguals

Affiliation: University of Bucharest

Email: larisa.avram(at)g.unibuc(dot)ro

Recent studies addressing bilingual language development have shown that the narrow syntactic properties are acquired early in L1 acquisition and fully in L2 acquisition. Interface phenomena, on the other hand, seem to be a vulnerable domain, whose properties may not be fully acquired in L2 (Sorace and Filiaci 2006 a.o.). Our study investigates the acquisition of accusative clitics in Romanian by two Romanian-Hungarian simultaneous bilinguals (child T., age 1; 4 – 3;0, and child P., age 1;4 – 3;0) on the basis of longitudinal data (Tomescu 2011). The focus will be on the ingredients of clitic doubling constructions: the clitic and the preposition pe (the equivalent of the Spanish a):

(1) Maria l- a văzut pe Ion.

Maria clitic Acc 3rd masc sg has seen pe Ion

„Maria has seen Ion.‟

In particular, the following questions are addressed:

(i) Are accusative clitics acquired early in simultaneous bilingual acquisition?

(ii) Does the acquisition of accusative clitics develop in parallel with the acquisition of pe?

(iii)Are the Romanian data consistent with the Interface Hypothesis?

Romanian accusative clitics have been analysed as a purely syntactic phenomenon in several studies (Dobrovie-Sorin 1994 a.o.). According to the Interface Hypothesis, clitics should be acquired similarly to clitics in Romanian as L1. Similar results have been reported, for example, for early clitics in Croatian-Italian simultaneous bilinguals (Kraš and Miličević 2011). The preposition pe has been argued to be more than a mere case marker; it has been analysed as a differential object marker (DOM) (Bossong 1998, Farkas and Heusinger 2003) or a speaker-oriented topicality marker (Avram and Coene 2009). The Interface Hypothesis predicts that the acquisition of pe by bilinguals should be delayed, since it is an interface phenomenon. Clitics and pe would develop at different speed, with pe lagging behind. The longitudinal data investigated reveal that the developmental path of accusative clitics with Romanian-Hungarian simultaneous bilinguals is similar to the one reported for monolingual Romanians (Avram and Coene 2004). The preposition pe, however, tends to be overused at a rate and in contexts which are different from the one reported for L1 (Avram 2011). Our findings are consistent with the predictions of the Interface Hypothesis. Interestingly, however, the difference between clitics as a purely syntactic phenomenon and pe as an interface phenomenon interferes with an important property of Hungarian; it lacks accusative clitics but it has DOM (Kamper 2006). As expected, DOM takes different forms in the two languages. The findings will also be examined in terms of the way in which language specific properties in bilingual settings may interfere with the distinction between the acquisition of narrow syntax and that of interface components.

 

5. Pedro Guijarro-Fuentes and Pilar Larrañaga

Title: Clitic drop in Basque Spanish: adult profiles and early acquisition

Affiliation: University of Plymouth

Email: p.guijarro-fuentes(at)plymouth.ac(dot)uk

The variety of Spanish spoken in the Basque Country allows for clitic drop in referential contexts. Note that Basque has no clitics but allows null subjects and objects by virtue of the rich verb agreement (Hualde 2003; Oyharçabal 2003; Ortiz de Urbina 2003). Larrañaga & Guijarro-Fuentes (2012) show that clitic-drop is a very idiosyncratic issue in adult language and very much depends on the individual speaker. The two children exposed to this variety of Spanish and studied by Larrañaga & Guijarro-Fuentes are balanced bilinguals and converge with the target language at some point in development, namely at around the age of three. Interestingly, the rate of clitic drop in the later age correlates with the rate of clitic drop of the interlocutors, the father and the nanny respectively, from which the researchers may assume that adult interlocutors spend a considerable amount of time with the children. The objective of the present paper is to test (a) whether one unbalanced Basque-Spanish bilingual child (data studied from the age of around two to four years of age) whose MLU lies well behind the MLU of the bilingual children studied, shows the same path of acquisition and (b) determine in which stage the unbalanced bilingual attains target-like performance. Moreover, we aim to determine which linguistic factors constrain the use of clitics in the different adult performance profiles in order to tease apart child performances that are explainable by the input received, from those which could be attributable to cross-linguistic influence.

22.07.2013   10:30-12:30

Chair: Pilar Larrañaga

 10:30 - 11:00 Pilar LARRAÑAGA
Introduction
11:00 - 11:30 Juana LICERAS et al.
Subject omission/production in child bilingual English and child bilingual Spanish: The view from linguistic theory
> read abstract...
11:30 - 12:00 João COSTA et al.
Clitic production by Portuguese and Capeverdean children: omission in bilingualism
> read abstract...
12:00 - 12:30 Anna GAVARRÓ
Objects and (2)L1 crosslinguistic variation, and the import of pragmatics
> read abstract...

22.07.2013   14:00-16:00

Chair:

14:00 - 14:30 Larisa AVRAM et al.
Romanian clitics and the Interface Hypothesis: The view from two Romanian-Hungarian simultaneous bilinguals
> read abstract...
14:30 - 15:00 Pedro GUIJARRO-FUENTES et al.
Clitic drop in Basque Spanish: adult profiles and early acquisition
> read abstract...
15:00 - 16:00 Pedro GUIJARRO-FUENTES et al.
Discussion