Effect of neighbourhood and markedness in bilingual visual word

Commissaire, E.

Laboratoire de Psychologie des Cognitions (EA 4440), Université de Strasbourg, France

How bilinguals deal with cross-language orthographic neighbor words during silent reading has received relatively little attention, despite cross-language lexical competition being a core mechanism assumed by bilingual models (e.g., BIA+, Dijkstra & van Heuven, 2002). Previous studies have investigated this issue by manipulating cross-language neighborhood size (N-size, van Heuven, Dijkstra & Grainger, 1998) or by testing the neighborhood frequency effect with the masked priming paradigm (Dijkstra, Hilberinck-Schulpen, & van Heuven, 2010). Despite some evidence for cross-language lexical competition, our hypothesis is that these effects might have been confounded with orthographic markedness effects reflecting the influence of the degree of language- shared (unmarked) versus specific (marked) orthography within L2 words. We present two experiments that challenge the interpretations made by past studies on this topic.
In the first experiment, French/English bilinguals performed an English (L2) lexical decision task with (non)words belonging to three conditions: 1) large cross-language N-size and unmarked orthography (price), 2) small cross- language N-size and unmarked orthography (drive) and 3) small cross-language N-size and marked orthography (write). We found no neighborhood effect but strong effects of orthographic markedness for both word and nonword processing.
In the second experiment, a similar population performed an L1-to-L2 masked primed lexical decision task with French (L1) primes preceding either English (L2) marked (drap - WRAP) or unmarked neighbor words (drap - TRAP), compared to an unrelated condition (drap- GIFT). Results yielded a cross-language inhibition priming effect in the marked condition only, suggesting that different mechanisms than word lateral inhibition might be at stake.