Kathleen Eberhard, Matthias Scheutz, Kathleen Targowski & Jeffrey Spies
University of Notre Dame
A central question in bilingual research concerns whether lexical representations in the two languages are processed independently or interactively. We investigated this question by examining whether the masculine gender associated with the -er ending of words in German is activated when German-English bilinguals read agentive -er nouns in English, which have no grammatical gender. For example, the noun "partner" is grammatically masculine in German, but has no grammatical gender in English. Thus, in German, "Partner" refers to a male while the grammatically feminine form, "Partnerin" refers to a female. The question of interest was whether, relative to monolingual English speakers, German-English bilinguals would exhibit a greater bias to infer the sex of an agentive noun, such as "partner", as male when encountering the word in an English sentence. To test this question, we recorded the eyemovements of German-English bilinguals and monolingual English speakers as they read English sentences with agentive nouns as antecedents of either a masculine or feminine reflexive as shown in (1).
(1) The partner wanted to be by himself/herself after the relationship ended.
Previous studies with native English participants showed longer fixation durations on pronouns whose grammatical gender disagrees with the inferred sex of an antecedent (e.g., Underwood, 1985). Thus, if German-English bilinguals are biased to infer the sex of a noun, such as "partner" as male, then they should exhibit longer fixation durations on "herself" relative to monolingual English speakers. The reflexive's gender was crossed with two other factors involving the agentive -er nouns. One was the nouns' stereotypical sex (neutral; male, e.g., farmer; or female, e.g., babysitter). Thus, if the masculine association with -er nouns in German affects processing -er nouns in English, then relative to monolingual English speakers, German-English bilinguals were expected to show shorter fixation durations on "himself" when the agentive noun is stereotypically female. The second factor was the formal similarity of the agentive nouns in English and German (i.e., cognates, e.g., partner-Partner; semi-cognates, e.g., swimmer-Schwimmer; or non-cognates, e.g., teacher-Lehrer). In addition to native language, the participant factors included the relative level of proficiency in English of the bilingual participants (medium vs. high).
The analyses of the fixation patterns generally support the basic predictions stated above. However, differences were observed as a function of the bilinguals' proficiency in English as well as the formal similarity of the agentive nouns. We will offer a theoretical interpretation of the results that is informed by the simulation results of a computational model.
Reference
Underwood, G. (1985). Eye movements during the comprehension of written language. In A. W. Ellis (Ed.), Progress in the Psychology of Language, pp. 45-71. Hillsdale: Lawrence Erlbaum Associates.