Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Neural correlates of conflict adaptation in word production
Poster Session B, Friday, September 12, 4:30 - 6:00 pm, Field House
Nazbanou Nozari1, Nikhil Lakhani2, Michael Freund3; 1Indiana University, 2University of Pennsylvania, 3Brown University
Introduction. Overriding a prepotent but incorrect response requires cognitive control. One way to identify the neural correlates of such control is by comparing high-conflict and low-conflict conditions. However, a more precise method is to study dynamic changes in control as a function of exposure to prior conflict. This can be captured by studying the congruency sequence effect (CSE) or conflict adaptation. CSE refers to the easier resolution of conflict in high-conflict trials if preceded by another high-conflict vs. a low-conflict trial. New theories of cognitive control attribute CSE to learning. This implies that CSE is long-lasting and survives interference from unrelated tasks. To test this prediction, we interleaved the picture-word interference (PWI) paradigm with another task that involved different representations and examined the neural correlates of CSE in language production. Methods. Twenty-five native speakers of English completed a picture-word interference (PWI) paradigm, interleaved with trials from a non-linguistic visuospatial task, in a 3T Siemens MRI scanner. PWI trials were either C (congruent; low-conflict) or I (incongruent; high-conflict). Congruent trials paired the picture with its correct name (e.g., apple/apple). Incongruent trials paired the picture with a semantically related word (e.g., apple/pear). The preceding PWI trial was also either congruent or incongruent, denoted by small letters c or I, creating four conditions: cC, iC, cI, and iI. Behavioral and neuroimaging data were analyzed from the PWI task. Results. Behaviorally, the typical CSE pattern was obtained, with an interaction between current and preceding congruency, despite the intervening task, pointing to a long-lasting effect. Neurally, the comparison of current incongruent and congruent trials revealed increased activation in frontal and temporal regions, as expected. The CSE analysis revealed greater activity in cI vs. iI trials in left dorsal anterior cingulate cortex (dACC), bilateral inferior frontal gyri (pars opercularis) and insula, bilateral dorsal prefrontal cortices, and left middle and inferior temporal gyri. Conclusions. The neural correlates of conflict adaptation in word production are compatible with conflict models that propose domain-specific conflict among lexical items (in the left posterior middle and inferior temporal gyri) is monitored by central medial frontal regions (e.g., dACC), which recruit and implement control in conjunction with the lateral (inferior and dorsal) prefrontal regions. The two new contributions of this study are that (a) these regions dynamically modulate control over word production, and (b) in line with the prediction of the learning account of control, CSE is long-lasting and resilient against the insertion of irrelevant tasks and representations.
Topic Areas: Control, Selection, and Executive Processes, Language Production