Search

We conducted a norming study amongst Dutch children (6-12 years) to investigate whether they would correctly interpret our set of 208 color drawings, developed for longitudinal research on the production and comprehension of transitive sentences, if the transitive action on each drawing was described with an active or a passive sentence. The children provided 93.02% correct answers to our pictures, which indicates that the pictures are clear in terms of how the transitive actions are displayed. There were several factors that contributed to an incorrect interpretation of a transitive action in a picture. (1) Pictures were more difficult to interpret when a passive sentence was used to describe a transitive event compared to an active sentence; (2) children made more errors interpreting pictures when abstract verbs, such as “inhalen” (to overtake) and “vervangen” (to replace), were used to describe a transitive event; (3) children had difficulty interpreting our pictures when particular verbs (“roepen” [to shout] and “volgen” [to follow]) were presented in a passive sentence, while other verbs (“blokkeren” [to block] and “inhalen”) rendered more correct answers when the transitive event was described with a passive sentence; finally, (4) younger children provided more incorrect answers to our pictures than older children, regardless of whether the transitive action in the picture was described with an active or a passive sentence. In particular, the younger children provided significantly more incorrect answers to pictures with the verb “groeten” (to greet) compared to the older children. Because our set contains many verb-noun combinations, the set is very useful to researchers who intend to investigate the acquisition of transitive structures in children in a longitudinal design, not only for studies with typically developing children, but also in studies with children who have a delay in their language development. Keywords: Norming study, picture set, transitive structures, L1 syntax acquisition.

Research on sentence production and comprehension has shown that the abstract structure of sentences can be primed (see Pickering & Ferreira, 2008, for a review). Recently, it has been shown that syntactic structures can also be primed between both languages of a bilingual (Hartsuiker et al., 2004; Schoonbaert et al. 2007; Salamoura & Williams, 2007). Hartsuiker et al. (2004) therefore concluded that syntactic representations can be shared between languages in the bilingual lexicon. In this dissertation, we investigated how similar syntactic structures have to be in order to be shared (Chapters 2 & 3), whether similar structures are shared immediately (Chapter 4) and whether the syntactic preferences of verbs influence these priming effects (Chapter 5). To this aim, we conducted a series of syntactic priming experiments in which the structures of different syntactic constructions were primed within and between both languages of late Dutch-English and Dutch-German bilinguals. We found that structures that differ only in word order (e.g., the shark that is red – de haai die rood is) receive separate, word-order specific syntactic representations. Cross-linguistic priming for such structures can thus only be obtained if the structures in question share a pre-syntactic representation (e.g., a functional and/or conceptual representation). Furthermore, we discovered that all new structures of the second language receive separate representations in an initial stage of acquisition. Finally, we showed that priming effects for dative sentences are modulated by the syntactic preference (verb bias) of the dative verb that is used in the prime. This indicates that the syntactic preferences of verbs is learned and represented in memory.

When learning a language, one needs to acquire the grammatical structures of the new language. Using structural priming, I studied how mental representations of such syntactic structures are formed during the process of learning a language, and how the syntactic representations of one language affect the syntactic representations of another language. Previous production studies suggest that L2 learners might start with item-specific and language-specific syntactic representations, which become abstract and shared between languages over time. I investigated how this transition takes place. My findings suggest that (i) beginning L2 learners of Dutch do indeed start with item-specific representations, which become abstract earlier for less complex structures than for more complex structures. However, (ii) very proficient L2 speakers of Dutch seem to have connected syntactic representations rather than shared representations between languages, and these connected syntactic representations cause cross-linguistic influence. Finally, (iii) a number of comprehension studies suggest that the processing of syntactic structures is not only determined by the nature of the syntactic representations, but also by explicit memory and residual activation of syntactic structures, and implicit learning of syntactic rules, and these factors have a different impact on L2 sentence processing at different stages of language learning.