Lee Osterhout - US grants

The proposed research will investigate the cognitive and neurological processes underlying language comprehension. Event-related brain potentials (ERPs) will be recorded while subjects read or listen to sentences containing specific deviations from well-formedness. Two approaches are taken in examining ERP effects associated with language comprehension. The first approach focuses on the ERP components themselves, in an attempt to learn more about the cognitive events underlying these effects. Recent work has indicated that syntactic and semantic anomaly elicit distinct ERP effects (the P600 and N400 effects, respectively). The proposed research investigates the hypothesis that these effects are elicited as a function of the linguistic level of the anomaly. The second approach focuses on the process of comprehension. How does the comprehender rapidly derive a single, coherent interpretation of the language input? How quickly are relevant types of information used, and which types of information have precedence over other types of information? The proposed research will use ERP responses to linguistic anomalies to investigate the use of information during language comprehension. Specifically, these experiments will examine the use of information in resolving ambiguity at the syntactic and coreferential levels of analysis. Finally, how similar are the processes that underlie the comprehension of written and spoken language? The proposed research will investigate this question by directly contrasting the ERP response to written and spoken sentences. Although the proposed experiments involve normal adult subjects, one long-term goal is to study language comprehension in children and adults with language disabilities (e.g., the dyslexias and aphasias). The studies proposed here, by furthering our knowledge concerning the cognitive and neural bases of comprehension in normal comprehenders, provide a critical starting point for subsequent ERP investigations of language pathologies.

Global changes in transportation and communication have placed a premium on the ability to speak two or more languages. However, learning a third language (L3) can interfere with the learner's proficiency in a previously acquired second (L2) language, a phenomenon referred to as language attrition. Fundamental questions about why languages attrite, how they attrite, and what aspects of language are most susceptible to attrition remain unanswered. This proposal seeks to better understand how the learning of a new language (L3) can interfere with the learner's proficiency with a previously acquired non-native language (L2), in the context of real-world language learning among college students. Dr. Lee Osterhout, from the University of Washington, will perform three experiments using a high density EEG-based brain activity measure to longitudinally investigate changes in learners' ability to process sentences in their L3 (Italian) and in their L2 (a similar language, Spanish, or a less similar language, such as Japanese). Because the brain activity measure is differentially sensitive to aspects of linguistic meaning and grammar, the results will provide a specific understanding of what linguistic knowledge (for both the L2 and L3) has been gained or lost, and how the acquisition of one language influences the loss of another.

Given the importance of multi-lingual skills in the modern, inter-connected world, a better understanding of the causes of language attrition will have broad impact. Findings from this project may lead to ways to improve and customize instructional pedagogy with respect to language learning, and to protect previously acquired foreign languages. This project will generate a large data set with broad relevance. Potentially, this data set could be used to study the spatial origins of language-sensitive brain responses, test new mathematical and computational tools for high-density EEG-based brain source imaging, and address the fundamental question of how brain activity, at the temporal scale of milliseconds and from multiple brain regions, changes as a result of learning and "unlearning" outside of the laboratory.