Language for action: Motor resonance during the processing of human and robotic voices.

In this fMRI study we evaluated whether the auditory processing of action verbs pronounced by a human or a robotic voice in the imperative mood differently modulates the activation of the mirror neuron system (MNs). The study produced three results. First, the activation pattern found during listening to action verbs was very similar in both the robot and human conditions. Second, the processing of action verbs compared to abstract verbs determined the activation of the fronto-parietal circuit classically involved during the action goal understanding. Third, and most importantly, listening to action verbs compared to abstract verbs produced activation of the anterior part of the supramarginal gyrus (aSMG) regardless of the condition (human and robot) and in the absence of any object name. The supramarginal gyrus is a region considered to underpin hand-object interaction and associated to the processing of affordances. These results suggest that listening to action verbs may trigger the recruitment of motor representations characterizing affordances and action execution, coherently with the predictive nature of motor simulation that not only allows us to re-enact motor knowledge to understand others' actions but also prepares us for the actions we might need to carry out.

Motor features of abstract verbs determine their representations in the motor system.

Embodied cognition theory posits that concept representations inherently rely on sensorimotor experiences that accompany their acquisitions. This is well established through concrete concepts. However, it is debatable whether representations of abstract concepts are based on sensorimotor representations. This study investigated the causal role of associated motor experiences that accompany concept acquisition in the involvement of the motor system in the abstract verb processing. Through two experiments, we examined the action-sentence compatibility effect, in the test phase after an increase in motor features during the learning phase for abstract verbs with low motor features (Experiment 1) or novel words with no conceptual features at all (Experiment 2). After associated motor experiences were added in the word learning phase, action-sentence compatibility effect was found in the semantic processing tasks during the test phase for abstract verbs (Experiment 1a) and novel words (Experiment 2). This was lacking in the word font color judgment task requiring no semantic processing (Experiment 1b). Coupled with our previous study, these findings suggest that motor features formed during word learning could causally affect embodiment in the motor system for abstract verbs, and reactivation of motor experiences in abstract verb processing depends on a given task's demands. Our study supports the view that conceptual representations, even abstract concepts, can be grounded in sensorimotor experiences.

Heterogeneity in abstract verbs: An ERP study.

It has been well documented that different types of nouns and action verbs are associated with behavioral and neural differences. In contrast, abstract verbs (e.g., think, dissolve) are often treated as a homogeneous category. We compared event-related potentials recorded during a syntactic classification task of four verb types; 1) abstract mental, 2) abstract emotional, 3) abstract nonbodily, and 4) concrete. Abstract nonbodily state verbs showed a sustained negativity at frontocentral electrodes and sustained positivity at parietal and occipital electrodes beginning 400 ms post-stimulus onset relative to abstract mental state and concrete verbs. Discrete source localization revealed a right inferior parietal source for all verbs and a distributed source estimation localized sources that distinguished between abstract mental state and abstract nonbodily state verbs to bilateral parietal cortex, left temporal cortex and right ventromedial prefrontal cortex. These findings suggest that different types of abstract verbs are associated with representational differences.

Virtual action and real action have different impacts on comprehension of concrete verbs.

In the last decade, many results have been reported supporting the hypothesis that language has an embodied nature. According to this theory, the sensorimotor system is involved in linguistic processes such as semantic comprehension. One of the cognitive processes emerging from the interplay between action and language is motor simulation. The aim of the present study is to deepen the knowledge about the simulation of action verbs during comprehension in a virtual reality setting. We compared two experimental conditions with different motor tasks: one in which the participants ran in a virtual world by moving the joypad knob with their left hand (virtual action performed with their feet plus real action performed with the hand) and one in which they only watched a video of runners and executed an attentional task by moving the joypad knob with their left hand (no virtual action plus real action performed with the hand). In both conditions, participants had to perform a concomitant go/no-go semantic task, in which they were asked to press a button (with their right hand) when presented with a sentence containing a concrete verb, and to refrain from providing a response when the verb was abstract. Action verbs described actions performed with hand, foot, or mouth. We recorded electromyography (EMG) latencies to measure reaction times of the linguistic task. We wanted to test if the simulation occurs, whether it is triggered by the virtual or the real action, and which effect it produces (facilitation or interference). Results underlined that those who virtually ran in the environment were faster in understanding foot-action verbs; no simulation effect was found for the real action. The present findings are discussed in the light of the embodied language framework, and a hypothesis is provided that integrates our results with those in literature.

Spatiotemporal dynamics during processing of abstract and concrete verbs: an ERP study.

Different accounts have been proposed to explain the nature of concept representations. Embodied accounts claim a key involvement of sensory-motor systems during semantic processing while more traditional accounts posit that concepts are abstract mental entities independent of perceptual and motor brain systems. While the involvement of sensory-motor areas in concrete language processing is supported by a large number of studies, this involvement is far from being established when considering abstract language. The present study addressed abstract and concrete verb processing, by investigating the spatiotemporal dynamics of evoked responses by means of high density EEG while participants performed a semantic decision task. In addition, RTs to the same set of stimuli were collected. In both early and late time intervals, ERP scalp topography significantly differed according to word categories. Concrete verbs showed involvement of parieto-frontal networks for action, according to the implied body effector. In contrast, abstract verbs recruited mostly frontal regions outside the motor system, suggesting a non-motor semantic processing for this category. In addition, differently from what has been reported during action observation, the parietal recruitment related to concrete verbs presentation followed the frontal one. The present findings suggest that action word semantic is grounded in sensory-motor systems, provided a bodily effector is specified, while abstract concepts׳ representation cannot be easily explained by a motor embodiment.