Abstract
Does sentence comprehension related to faces modulate activity in the fusiform face area (FFA) and does sentence comprehension related to places modulate activity in the parahippocampal place area (PPA)? We investigated this question in an fMRI experiment. Participants listened to sentences describing faces, places, or objects, with the latter serving as a control condition. In a separate run, we localized the FFA and PPA in each participant using a perceptual task. We observed a significant interaction between the region of interest (FFA vs. PPA) and sentence type (face vs. place). Activity in the left FFA was modulated by face sentences and in the left PPA was modulated by place sentences. Surprisingly, activation in each region of interest was reduced when listening to sentences requiring semantic analysis related to that region's domain specificity. No modulation was found in the corresponding right hemisphere ROIs. We conclude that processing sentences may involve inhibition of some visual processing areas in a content-specific manner. Furthermore, our data indicate that this semantic-based modulation is restricted to the left hemisphere. We discuss how these results may constrain neural models of embodied semantics.
Acknowledgements
This study was supported by a NIH grant (P01 NS40813) and the Klaus Tschira Foundation. CJF was supported by an Emmy-Noether grant from the German Research Foundation (DFG).
We thank Stephen M. Wilson and Miriam R. L. Petruck for their assistance with this study and members of Mark D'Esposito's laboratory at UC Berkeley for providing stimuli for the localizer task.
Notes
1In addition to the study by Buccino et al. (2005), which shows inhibition, a study by Pulvermuller's group may also indicate inhibition. Pulvermuller et al. had participants perform a vocal lexical decision task while TMS was applied to different regions of motor cortex (Pulvermuller, Hauk, Nikulin, & Ilmoniemi, Citation2005). RTs were facilitated in an effector-specific manner; e.g., judgments of hand-related words were faster when TMS was applied over the hand area compared to the foot area. They interpreted these results as consistent with the notion that activation of motor cortex supports effector-specific lexical retrieval. However, if the TMS pulses are considered to add noise, then one could argue that the effector-specific improvements in RT might be due to the transient disruption of potentially competing signals from motor cortex. For example, judging that the letters “POKE” spell a hand-action word might be faster if a representation of the current posture of the hand is disrupted. Three points suggest that this alternative interpretation is more parsimonious. First, the effects of single-pulse TMS are usually viewed as disruptive to normal processing, rather than facilitatory. Second, a main effect on RT was associated with hand area TMS, a result attributed to the spread of interference effects to the mouth area. Third, this interpretation is consistent with the more direct assay provided by Buccino et al. (2005).