Speaker
Description
Over the last decades, various models have addressed the mechanisms of conceptual processing in bicultural bilinguals. Among these, the Shared Distributed Asymmetrical Model (Dong et al., 2005) proposes that bilinguals have partially overlapping yet asymmetrically distributed conceptual representations for their two languages: initially, L2 words only map onto L1-based conceptual representations - a phenomenon known as “semantic accent”; then, as proficiency and cultural exposure increase, L2 labels would progressively link more to conceptual representations specific to the semantics and culture of the L2.
This study tested this assumption by running a primed lexical decision task in participants’ L2. Prime-target pairs were of three types: a) semantically-related in L1 but not in L2; b) semantically-related in L2 but not in L1; c) semantically-related in both. Prime-target association was measured in terms of semantic similarity within vector semantic models. Pairs with non-word targets were included as fillers.
A group of Italian monolinguals (control group) and a group of British English-Italian bilinguals with different levels of proficiency and cultural exposure were involved.
As expected, the control group showed slower reaction times for pairs semantically related in L1 but not in L2, compared to the other two conditions, which did not differ significantly. Contrary to expectations, the bilingual group showed a pattern in line with that of monolinguals. We will discuss these results also considering the stages of L2 acquisition and the levels of cultural exposure.
REFERENCES
• Dong, Y. & Gui, S., Macwhinney, B. (2005). Bilingualism: Language and Cognition, 8, 221-228.
