Volume 51, Issue 2

The study examines which strategies native speakers of German, native speakers of Japanese and advanced German learners of Japanese adopt when linking information in narrative texts (film retellings). The main results are as follows: (a) The L1 German speakers organize temporal shift-relations in order to link information, whereas the protagonist in topic function is often maintained across utterances. In contrast, the Japanese speakers prefer to compare series of events involving each of the main characters by organizing a shift in the entity domain. (b) The L2 strategies cannot be explained by a single feature, as various factors are seen to interact in the L2 speakers' decision making. In particular, the organization of information from the domain of time is directly related to the principles in the source language.

This paper investigates the role of the underlying L1 perspective in the construal of goal-oriented motion and events with a resultant state in L2 production. Previous findings demonstrated that native Czech and German speakers have the same preferences for event construal (holistic perspective). The same holds true for Czech L2 speakers of German. Since L1 Czech and L1 German share the same preferences for motion event encoding, it is impossible to decide whether Czech learners of L2 German chose the holistic perspective because of L1 transfer or because they have restructured their conceptual preferences in the direction of the L2. This question can be resolved by examining data from Czech L2 speakers of English since L1 English, unlike L1 Czech, prefers taking a phasal perspective in motion event construal. The present data show that, for highly proficient L2 speakers, conceptual restructuring in the direction of the L2 is possible, however limited.

In the course of language acquisition learners have to deal with the task of producing narrative texts that are coherent across a range of conceptual domains (<italic>space, time, entities</italic>) -- both within as well as across utterances. The organization of information is analyzed in this study, on the basis of retellings of a silent film, in terms of devices used in the coordination and subordination of events within the narrative sequence. The focus on subordination reflects a core grammatical difference between Italian and French, as Italian is a null-subject language while French is not. The implications of this contrast for information structure include differences in topic management within the sequence of events. The present study investigates in how far Italian-French bilingual speakers acquire the patterns of monolingual speakers of Italian. It compares how early and late bilinguals of these two languages proceed when linking information in narratives in Italian.

This study examined whether resolving ambiguous pronouns in a second language is guided by the L1 preferences of the learners. Given the fact that the typologically closely related languages, German and Dutch, have both been found to use personal pronouns (German <italic>er</italic>, Dutch <italic>hij</italic>; ‘he’) to refer to topical antecedents, and d-pronouns (German <italic>der</italic>, Dutch <italic>die</italic>; ‘he’) for non-topical co-reference (Ellert 2010; Kaiser 2011; Kaiser and Trueswell 2004), it was asked whether Dutch L2 learners of German would exhibit similar preferences when resolving the two pronominal forms in their L2. This was examined with the visual-world eye-tracking paradigm and an off-line referent assignment task. The results showed differences in resolution patterns: the Dutch learners of German showed an overall topic preference across pronouns which became more target-like at higher proficiency levels. This suggests that L2 information organization cannot be merely explained by L1 influences, but needs to take more general L2 learner effects into account.

The present study focuses on the role of grammatical aspect in event construal and its function in encoding the specificity of an event. We investigate whether advanced L2 learners (L1 German) acquire target-like patterns of use of progressive aspect in Dutch, a language in which use of aspect depends on specific situation types. We analyze use of progressive markers and patterns in information selection, relating to specific features of agents or actions in dynamic event scenes. L2 event descriptions are compared with L1 Dutch and L1 German data. The L2 users display the complex situation-dependent patterns of use of aspect in Dutch, but they do not select the aspectual viewpoint (<italic>aan he</italic> construction) to the same extent as native speakers. Moreover, the encoding of specificity of the events (mentioning of specific agent and action features) reflects L1 transfer, as well as target-like performance in specific domains.