Cognitive Processing of Grammatical Gender and Number by Russian Native Speakers and Turkic-Russian Bilinguals

The effect of the grammatical structure of L1 on the processing of grammatical categories in L2 was investigated using a visual lexical decision task. Reaction times of Turkic-Russian bilinguals, whose native languages have no grammatical gender, were measured during the processing of grammatical gender and number of Russian nouns preceded by an agreeing, disagreeing, or neutral prime. The research objective was to test the markedness effect hypothesis which states that unmarked members of grammatical oppositions are processed faster and are less dependent on the facilitative prime, as well as the L1 interference hypothesis according to which L1 grammar influences the processing of L2 grammar. The expectations were as follows: 1) L1 Russian speakers would demonstrate default form bias for both gender and number which will manifest itself in a) shorter reaction times to the default member of the corresponding grammatical opposition and b) differential priming patterns for masculine and feminine nouns. 2) If L1 grammar does affect grammar processing in L2, Turkic-Russian bilinguals would only exhibit the default form bias for grammatical number but not gender since only the former is instantiated in their L1. Moreover, the priming effect would be stronger for number than for gender. Experimental design partly replicates that used by Romanova (2013). Two groups of participants took part in a priming experiment with a lexical decision task - Turkic-Russian bilinguals (19 participants) and L1 Russian speakers (34 participants). Predictor variables were target number and gender, and the type of prime. No significant difference in reaction times to masculine and feminine nouns was found in either group. The data therefore provide no evidence of default form bias for grammatical gender. The effect of grammatical number reached statistical significance in both groups but only in the analysis by items with both groups of participants processing singular nouns faster than plural. The effect of grammatical number could be interpreted as evidence of the default form bias for grammatical number but only tentatively since it could also result from the fact that singular nouns are morphologically less complex and, unlike plural ones, are dictionary forms. A firm conclusion cannot be made regarding the effect of L1 grammar on L2 grammar processing. Although the bilingual participants in the experiments demonstrated the default form bias for grammatical number instantiated in their L1 and no default form bias for grammatical gender not instantiated in their L1, so did Russian participants. Since both groups showed similar processing patterns, the difference in gender and number processing cannot be attributed to the bilinguals' L1.

Keywords

Authors

References