The interaction of L1 and L2 processes in the proficient speaker

L1 influences on the processing of subject pronouns in L2 discourse: an eye-tracking study (Roberts, Gullberg, Indefrey)

This study investigates the potential influence of the L1 in the processing of overt subject pronouns in L2 discourse. We test Turkish (pro-drop) and German advanced learners of Dutch. The results show that the learner groups differ from each other off-line. There is a visible L1 effect in interpretation. In contrast, a general learner/bilingualism effect is found on-line. The learner groups both differ from the Dutch but pattern together irrespective of the properties of their L1s.

The Revised Hierarchical Model of bilingual memory and its neural implications (Menenti (MA), Indefrey)

In this research we obtained evidence for L1 (German) word form activation in L2 (Dutch) lexical decision by use of a new cross-language mediated phonological priming paradigm. Based on the behavioural data we conducted a fMRI study and identified a possible neural substrate for L1 lexical activation during L2 processing.

Lexical & functional categories in OV/VO word order patterns (Dikker (MSc), Indefrey)

This research investigated the comprehension of written code-switched sentences in bilingual English/Dutch subjects using the sentence matching paradigm. The results suggest that VP word order following the language of AUX (Jan heeft A NEW BOOK BOUGHT) is easier to process than VP word order following the language of V (Jan heeft BOUGHT A NEW BOOK).

L1 influences on the processing of grammatical progressive aspect in L2 English (Roberts)

In this study we focus on the real-time processing of past simple and past progressive sentences in English by L2 learners whose L1 either encodes past progressive aspect grammatically (French) or only via lexical means (German). An effect of the learners' L1 was found: the French group performed like the English control group, being sensitive on-line to differences in aspect, whereas German L2 learners treated the two sentence types in the same way in on-line processing, even those who could successfully distinguish the past simple and past progressive in an off-line production task.

Effects of sentence context in L2 natural speech comprehension (FitzPatrick, Indefrey)

With written stimulus materials, the N400 effect of semantic incongruity tends to occur later in L2 comprehension than in native comprehension. Using ERP, we investigated semantic integration in natural auditory L2 speech and tested whether a possible delay in semantic processing might be due to bilinguals accessing lexical items from both their L1 and L2. Results show that semantic integration in non-native listening can start on the basis of word initial phonemes and suggest that during auditory L2 sentence processing listeners do not access the L1 lexicon.

The processing of L2 word order (Schneider, Indefrey)

This PhD project investigates the influence of L1 word order on L2 sentence processing in intermediate German and Turkish learners of Dutch. Results suggest that L1 word order has an influence on L2 processing because the two learner groups differed in their processing of L2 sentences. German learners showed a similar reading time pattern as a native Dutch control group, SOV sentences being read more slowly than SVO. Turkish learners showed a reading time pattern that was both different from German learners and L1 speakers of Dutch.

Convergence in established and emerging language systems: Evidence from speech and gesture (Brown, A., Gullberg)

Brown's doctoral dissertation (completed 2007) examines the influence of an L2 (English) on an L1 (Japanese) in the domain of motion event construal. The study examines both speech and gesture and controls for immersion by considering groups matched for formal proficiency but resident either in the L1 community (Japan) or in the L2 community (USA). The overall results suggest that the L2 influences the L1 in both modalities even at moderate levels of proficiency in the L2 regardless of immersion.

The time-course of word-form encoding in second language word production (Bašnaková , Indefrey, Gullberg)

This ERP study shows that lexical phonological retrieval in L2 word production is not delayed compared to L1 word production.

The L2 production and processing of the English present perfect (Roberts)

Using self-paced reading and an off-line form interpretation task, this study investigates German and French L2 learners' production and on-line processing of the present perfect in English. The French learners, whose L1 like English semantically distinguishes the present perfect from the past simple, were better than the German group (whose L1 does not) at correctly producing the present perfect in the appropriate contexts. Furthermore, on-line the French group, unlike the Germans but like the English, were sensitive to mismatches between the temporal adverbial and the present perfect verb. In sum, the learners' performance was affected by how similar their L1 is to English in this respect.

The processing of null and overt pronouns in L2 discourse (Roberts)

This study investigates Turkish-Dutch bilinguals' on-line processing of grammatical, inappropriate and ungrammatical dependencies between pronouns and their antecedents in L2 discourse. To investigate questions of L1 influence and language dominance, the bilingual participants were divided into two dominance groups, Turkish and Dutch, and they were tested in both their languages, and their performance was compared to native speakers of both. For the ungrammatical texts, all groups, whether learner or native, perform in the same way. However, for the grammatical and inappropriate texts, the results show a general bilingualism effect in that all bilinguals performing in their weaker language, whether Turkish or Dutch, pattern together regardless of the properties of the L2s.

Re-structuring of semantic representations in L2 production: Placement events in speech and gestures (Gullberg)

This line of study connects to a large-scale crosslinguistic investigation of placement and removal events (see Annual Report 2005: 83-85, 2006: 69-73), probing differences in the semantics of verbs used to describe such events, but also more far-reaching differences in event construals and linguistic conceptualisation, that is to say, in what information speakers consider as relevant when they talk about placement and removal. This particular study investigates the assumption that target-like surface forms in advanced L2 learners' speech also imply target-like meaning in the domain of placement. A first study establishes that placement events are construed differently in Dutch, German, and French as seen in crosslinguistically different native patterns in speech and gesture. A second study examines the production of Dutch and German learners of French to reveal evidence both of re-structuring in the L2 and of transfer effects.

A related project explores L1 development of placement vocabulary looking at speech and gesture in Dutch and Tamil (with B. Narasimhan).

Expressions for static topological relations in Turkish-Dutch bilingual speakers (Indefrey, Şahin, Gullberg)

This study focuses on cross-linguistic differences in the categorisation of static topological relationships in Turkish and Dutch. Results of an elicitation experiment show that Turkish-dominant and Dutch-dominant bilingual speakers are differentially affected by cross-linguistic differences.

Code-switching in production (Gullberg, Muysken, Indefrey, Hellwig)

This sub-project examines the relationship between postulated linguistic constraints on code-switching and processing. We have developed a paradigm for inducing code-switching experimentally, and collect production data from the same participants on a range of tasks to allow for several types of baseline and validation.