Statistical word learning in bilinguals: Examining changes in learning behaviour due to differences in language environment

• Applicant: Dr. Tanja Charlotte Roembke, Ph.D.
• Subject Area: General, Cognitive and Mathematical Psychology
• Term: since 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 505754094

Bilingualism—the ability to speak and comprehend more than one language—is very common: As many as 50% of the global population are considered bilingual. While bilingualism is generally considered an advantage, it also comes with unique challenges, such as the need to acquire more than one language sound system. Surprisingly, there has been relatively little research into language acquisition in bilinguals, with many studies explicitly excluding bilinguals from analysis. One important aspect of language acquisition is statistical word learning, the ability to acquire words based on their co-occurrence with semantic concepts alone. Statistical word learning can be assessed in the cross-situational word learning paradigm: In each trial, participants are presented with a novel word and objects. They have to select the object that maps onto the word. Even as every single trial by itself is ambiguous, word-object-mappings can be acquired over time. This is because the target object is always present, whereas competitor objects are randomized across trials. Adults as well as children can acquire words based on such statistical regularities only. In addition, previous research has shown that the structure of the learning environment influences how people acquire words. For example, studies show that bilinguals are less likely to assume that semantic concepts which they already have names for cannot be named by another word, too. During cross-situational word learning, there is some evidence that bilinguals may be more open to accept multiple meanings for words than monolinguals. However, at this point, our understanding of how statistical word learning proceeds in bilinguals is very limited. In this project, we investigate how differences in the language environment that are consistent with bilingualism impact statistical learning of words. Bilingual as well as monolingual adult participants will complete cross-situational word learning experiments that are hosted both in the laboratory as well as online. We will manipulate, among other factors, the number of meanings per word, the frequency of language switching and the sound of to-be-acquired words. We will combine both established behavioral measures (e.g., accuracy, reaction time) as well as more novel ones (e.g., eye movements) to characterize learning continuously and in-the-moment. To summarize, this project will give us insights into how differences in learning environment change how statistical learning occurs. Therefore, it will greatly inform our understanding of basic learning mechanisms as well as our knowledge of bilingual language acquisition.

• DFG Programme: Research Grants
• Co-Investigator: Professor Dr. Iring Koch