Grasping the semantic of actions: a combined behavioral and MEG study

Abstract

There is experimental evidence that the brain systems involved in action execution also play a role in action observation and understanding. Recently, it has been suggested that the sensorimotor system is also involved in language processing. Supporting results are slower response times and weaker motor-related MEG Beta band power suppression in semantic decision tasks on single action verbs labels when the stimulus and the motor response involve the same effector. Attenuated power suppression indicates decreased cortical excitability and consequent decreased readiness to act. The embodied approach forwards that the simultaneous involvement of the sensorimotor system in the processing of the linguistic content and in the planning of the response determines this language-motor interference effect. Here, in a combined behavioral and MEG study we investigated to what extent the processing of actions visually presented (i.e., pictures of actions) and verbally described (i.e., verbs in written words) share common neural mechanisms. The findings demonstrated that, whether an action is experienced visually or verbally, its processing engages the sensorimotor system in a comparable way. These results provide further support to the embodied view of semantic processing, suggesting that this process is independent from the modality of presentation of the stimulus, including language.

Keywords: MEG (magnetoencephalography); beta rhythm; embodiment; language processing; motor responses; semantics; sensorimotor system.

Figure 1

Experimental procedure. Participants were asked to fixate on the center of the screen placed in front of them. Each trial started with the presentation of the stimulus surrounded by a red frame. The stimulus could be either a hand or foot verb, a hand or foot action image, and a pseudoverb or a scrambled image. After 150 ms the frame turned green, and the participants were allowed to respond. Participants were instructed to respond only to words and pictures depicting hand- or foot-related actions. The trial ended when participants provided their responses or after 1,350 ms if no response was given. Stimuli examples: hand action verb (A), hand action image (B), pseudoverb (C), scrambled image (D).

Figure 2

(A) Grand average maps of responses to different categories of stimuli in the 150–350 ms period. The highlighted area represents the ROI used for the analysis. The color scale is the same for each map (0–0.3); for illustrative purposes, maps were thresholded at 50% of the maximum amplitude. (B) Time-frequency representation of the virtual channel obtained from the ROI for each condition. The highlighted area represents the time and frequency interval selected for the analysis. The color scale is the same for each image (−0.3–0.4). (C) Beta band modulation with respect to baseline (−1–0 s) for verbs (upper graph) and images (lower graph) stimuli. Shaded areas indicate the standard error of the mean. Dotted lines indicate the period for AuC calculation, gray dotted line refers to the go signal.