Dynamic behavioral and neural effects of cognitive control on language processing

University Of Maryland, College Park, College Park MD

Investigators

Abstract

As people navigate their world, they sometimes face conflicting sources of information due to new rules, cues, or plans of action. In memory, for instance, the presence of familiar but irrelevant information can interfere with a person's ability to recognize an object correctly. While processing language, evidence from discrepant sources of input can result in misinterpretation. In these situations, cognitive control allows people to adjust thoughts and actions on-the-fly to mediate conflicting representations in both linguistic and nonlinguistic contexts. Yet, little is known about whether engaging cognitive control on one task influences or causes performance shifts on another. For example, is the ability to revise incorrect interpretations during language processing affected by prior experience of information-conflict, even in a nonlinguistic context? This project is designed to understand the interplay among multiple cognitive systems, whether the same cognitive control functions operate systematically across conflict in different domains, and to characterize the behavioral and neurobiological mechanisms that underlie their interaction. Specifically, this research investigates the causal relationship between cognitive control and language processing. To understand meaning and intent, people must be able to efficiently process spoken and written language input. Developing and honing these abilities takes considerable practice, especially when reading and listening to language also means dealing with conflicting sources of information. Such information-conflict can hinder a person's ability to make sense of complex material, which can lead to high rates of interpretation error. The work validates, informs, and improves prior training work through rigorous assessment of the underlying causal neurobiological mechanisms. Additionally, the Fellow is fully committed to STEM education and training in cognitive neuroscience. These experiments provide training opportunities for the Fellow and members of her research team. Results are disseminated to target audiences and submitted to academic conferences and journals. Moreover, as a woman, the Fellow has been involved in STEM outreach to underrepresented local groups. In this way, she hopes to increase interest in STEM fields, encouraging diverse and underrepresented groups to engage in scientific research. This project makes important contributions to understanding the human computational system that supports the real-time interpretation and re-interpretation of sentences. The experiments adopt converging eye-tracking and neuroimaging techniques to address a central issue in cognitive science: how language processing is affected by the engagement status of the cognitive control system. These studies represent the first methodical investigation into how nonlinguistic factors shape the time-course of language comprehension in both brain and behavior. Because cognitive control deficits affect patients' memory and language skills alike, elucidating the dynamic interplay between these cognitive systems has major health implications. The results from this research inform an understanding of shared language, memory, and cognitive control functions in the human mind and brain, which can be applied to public knowledge about how various cognitive systems develop typically and atypically, and how they fail following injury to the underlying neurobiological structures. As a result, the PI-team can draw conclusions about the causal nature of certain language and memory processes, which can be disseminated to and used in clinical, educational, and government settings. The first study tests how dynamic cognitive control engagement (turning it on or off depending on previous exposure to conflict in a nonlinguistic task) affects real-time language processing, indexed by fine-grained eye-movement patterns to objects in a scene as listeners carry out spoken instructions. Study 2 takes a neurobiological approach to examine cortical changes during language processing depending on whether cognitive control has been triggered by preceding experience. Study 3 investigates the extent to which ostensibly different memory and language tasks share a common conflict-control mind state by testing whether machine-learning algorithms can accurately classify brain-activation patterns broadly across domains. By pairing eye-tracking and fMRI techniques, this work offers converging evidence for a general-purpose cognitive control system whose deployment can shape how a person interprets language and how specified brain systems respond to linguistic input that is ripe for misanalysis.