This study explores the principle of linguistic relativity (Whorf, 1956), by examining how grammatical gender influences cognitive processes - categorisation (Experiment 1) and memory (Experiment 2) - in Ukrainian-Russian simultaneous bilinguals. Both languages feature distinct three-gendered systems, allowing us to test the claim that speakers of three-gendered languages are less influenced by grammatical gender compared to those of two-gendered languages (Sera et al., 2002). In this talk, we present two experiments to investigate this. Additionally, all experiments were conducted in an all-English environment to assess whether the grammatical gender effects of bilinguals’ L1s persist when tested in a genderless L2. 

Both experiments used two types of stimuli: (T1) nouns with aligning grammatical gender in Ukrainian and Russian, and (T2) nouns with misaligning grammatical gender in the two L1s. All nouns were conceptually neutral and paired with either a gendered character (Experiment 1) or a male/female name (Experiment 2). Stimuli were assigned to congruent or incongruent conditions based on the alignment of grammatical gender with the biological sex of the paired stimulus.

In Experiment 1, participants rated the similarity of object-character pairs (e.g., notebook-ballerina) on a Likert scale. In T1 stimuli, Ukrainian-Russian bilinguals were expected to rate congruent pairs as more similar, compared to the incongruent ones. As for T2, we expected to find effects of the more proficient L1 on the bilinguals’ ratings. Aligning with our predictions, bilinguals showed higher ratings for pairs where the grammatical gender of the object was congruent with the biological sex of the character for both types of stimuli. In Experiment 2, participants completed an object-name memory task where they memorized pairs of objects and prototypical male or female names (e.g., chair-Patrick). Bilinguals showed a grammatical gender effect for T1 nouns, with higher recall accuracy for congruent pairs. However, L1 proficiency did not affect recall for T2 nouns.

Together, these findings suggest that grammatical gender influences cognitive processes in bilinguals even in languages with three genders and when the aforementioned gendered languages are not required (i.e., all-English context). However, the effects vary when grammatical genders mismatch in 2L1s, providing more insights into how two gendered grammatical systems interact in a bilingual mind.