Sá-Leite, Luna, Fraga & Comesaña

A meta-analysis about the cross-linguistic Gender Congruency effect in bilinguals

Ana Rita Sá-Leite (Cognitive Processes & Behaviour Research Group, Department of Social Psychology, Basic Psychology & Methodology, University of Santiago de Compostela, Spain), Karlos Luna (Human Memory Research group, CIPsi, School of Psychology, University of Minho, Portugal), Isabel Fraga (Cognitive Processes & Behaviour Research Group, Department of Social Psychology, Basic Psychology & Methodology, University of Santiago de Compostela, Spain), & Montserrat Comesaña (Centro de Enseñanza Superior Cardenal Cisneros, Universidad Complutense de Madrid, Spain; Psycholinguistics Research group, CIPsi, School of Psychology, University of Minho, Portugal)

The study of gender representation and processing in bilingual populations contrasts two perspectives: the integrated vs. the autonomous view (Klassen, 2016). According to the first one, bilinguals have a unique gender system in which cross-linguistic interactions may occur. More specifically, the activation of a gender value in a target language can prompt the activation of nouns of the same gender value in the other language and hamper the activation of nouns of a different gender value. Conversely, according to the autonomous view, each language has its own gender system. Thus, the fact that two translations share the same gender value is irrelevant for the grammatical representation and processing of nouns of a second language (L2). To test both perspectives, authors have classically examined the Gender Congruency (GC) effect. The GC effect consists in a facilitation on the naming or translation of L2 nouns when they share the same gender value with their first language (L1) translations. Yet, conclusions from the reviewed studies suggest that this effect is quite “slippery” and may vary as a function of multiple variables. More specifically, the syntactic structure produced to translate or name the target (bare nouns vs. noun phrases) may affect the results, as the requirement of gender agreement in noun phrases (e.g., “la casa” [the house] vs. “el armario” [the closet]) might activate gender to a greater extent than when agreement does not have to be fulfilled (“e.g., “casa” or “armario”). In addition, modulations in the effect are expected as a function of the phonological gender transparency of the languages at issue (i.e., the presence of phonological gender cues related to the ending letter of nouns [e.g., “–a” in feminine nouns such as “casa” in Spanish]), the degree of L2 proficiency, and certain task requirements (naming vs. translation). The aim of this meta-analysis was to test the robustness of the GC effect. We considered twenty-five experiments from 11 different studies. The results show a small but significant GC effect regardless of the aforementioned variables and thus support is given to a bilingual gender-integrated view.

References

Klassen, Rachel. 2016. The representation of asymmetric grammatical gender systems in the bilingual mental lexicon. Probus, 28.

Back to Conference Programme