Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Action language and motor imagery recruit a shared motor network: an fMRI study
Poster Session E, Friday, October 2, 11:00 am - 1:00 pm, Wangari Maathai
Marcela Perrone-Bertolotti1, William Dupont2, Richard Palluel-Germain1, Florent Lebon3, Carol Madden-Lombardi2, Laurent Vercueil4, Emilie Cousin1,5; 1Univ. Grenoble Alpes, CNRS, LPNC UMR 5105, 38000, Grenoble, France, 2Université de Bourgogne Franche-Comté, UFR STAPS, INSERM U1093, Cognition, Action et Plasticité Sensorimotrice (CAPS), Dijon, France, 3Université Claude Bernard Lyon 1, INSERM U1093 (CAPS), Lyon, France, 4Service d’Exploration Fonctionnelle du Système Nerveux, CHU Grenoble Alpes, Grenoble, France, 5Université Grenoble Alpes, INSERM US17, CHU Grenoble Alpes, CNRS UAR 3552, IRMaGe, 38000, Grenoble, France
Introduction: Motor imagery and action language have both been proposed to recruit motor representations, consistent with embodied cognition theories suggesting that understanding action-related language partly relies on motor simulations. Neuroimaging studies have reported overlapping activations between action language and motor imagery within premotor and parietal regions, suggesting shared motor representations. However, it remains unclear whether these overlaps reflect a truly shared motor network or neighboring but functionally distinct processes. Moreover, few studies have directly compared action language and motor imagery within the same framework while anchoring both processes to overt motor execution. This fMRI study aimed to characterize the neural substrates of action language and motor imagery, quantify their overlap using conjunction analyses, and determine whether shared activations correspond to core motor regions identified through overt motor execution. Methods: Twenty-nine healthy right-handed adults (18–40 years) participated in a 3T fMRI study conducted at the IRMaGe MRI facility (Grenoble, France). Participants completed three tasks: (1) kinesthetic motor imagery of hand and foot actions, (2) semantic categorization of action-related sentences (action language), and (3) overt motor execution used as a functional localizer. A non-action language condition involving abstract sentences served as a linguistic control. Functional MRI data were analyzed using a general linear model implemented in SPM12. Whole-brain contrasts were computed for each condition relative to baseline, as well as direct contrasts between action language and motor imagery. Conjunction analyses based on the minimum statistic approach identified shared activations across conditions. Statistical inference used family-wise error correction (p < .05). Exploratory analyses examined effector-specific representations for hand- and foot-related actions. Results: Motor imagery, action language, and overt motor execution each robustly activated a bilateral motor network including primary motor cortex, premotor cortices, supplementary motor area, and parietal regions. In contrast, non-action language did not significantly activate motor regions. Direct comparisons between action language and motor imagery revealed no significant differences, indicating highly similar activation profiles across the two conditions. Conjunction analyses demonstrated significant shared activations between action language and motor imagery within premotor, primary motor, and parietal regions involved in action representation. Importantly, these shared activations substantially overlapped with regions identified during overt motor execution, suggesting recruitment of core motor circuits rather than adjacent associative areas. No significant overlap was observed between non-action language and either motor imagery or overt execution. Exploratory analyses further revealed shared hand- and foot-related activations across action language and motor imagery conditions. Conclusion: These findings provide converging univariate evidence that action language comprehension and motor imagery rely on a shared motor network substantially overlapping with regions engaged during overt action execution. The results support embodied cognition accounts proposing that action-related language processing involves motor simulations similar to those recruited during imagery and motor planning. Ongoing multivariate analyses are examining whether action language and motor imagery rely on shared or distinct representational codes within this motor network.
Topic Areas: Meaning: Lexical Semantics,