Task related modulation of the motor system during language processing

Abstract

Recent neurophysiological and brain imaging studies have shown that the motor system is involved in language processing. However, it is an open question whether this involvement is a necessary requisite to understand language or rather a side effect of distinct cognitive processes underlying it. In order to clarify this issue we carried out three behavioral experiments, using a go-no go paradigm. Italian verbs expressing hand actions, foot actions or an abstract content served as stimuli. Participants used their right hands to respond. In Experiment 1, in which a semantics decision task with an early delivery of the go signal (during processing language material) was used, slower responses were found for hand action-related verbs than for foot action-related verbs. In Experiment 2, using the same task with either an early or a delayed delivery of the go signal (when language material had been already processed), no difference was found between responses to the two verb categories in the delayed delivery condition. In Experiment 3, in which a lexical decision task with an early delivery of the go signal was used, again no difference between the two verb categories was found. The present findings demonstrate that during language processing the modulation of the motor system crucially occurs while performing a semantics decision task, thus supporting the notion that this involvement is a necessary step to understand language rather than a side effect of upstream cognitive processes.

Introduction

The precise neural mechanisms underlying language understanding remain largely debated. According to a classical view, language understanding relies on symbolic, amodal mental representations (e.g., Fodor, 1975). Following this approach, the meaning of language derives from an arbitrary correspondence between abstract symbols and their corresponding extensions in the world. An alternative hypothesis assumes that language understanding relies on ‘embodiement’ (Barsalou, 1999, Gallese and Lakoff, 2005, Glenberg, 1997, Glenberg, in press, Lakoff, 1987, Pulvermüller, 2002, Zwaan, 2004, Zwaan and Taylor, 2006). Based on this view, language understanding implies the involvement of the neural systems used when perceiving or acting.

Focusing on action-related language material, recent neurophysiological data have provided with experimental evidence that the motor system is involved during language processing. In a high resolution electroencephalographic (EEG) study, using leg and face actions-related verbs and pseudo-words in a lexical decision task (i.e., deciding whether a letter string is a word or not), Pulvermüller, Harle, and Hummel (2001) showed that, event-related potentials were differently located for the action-related verbs. While leg-related verbs entailed an activity focus close to the vertex, overlapping with the motor representations of leg movements, the face-related verbs produced a focus above the cortical representation of the face and mouth movements. Recent functional brain imaging studies confirmed this ‘somatotopic’ activation of the motor system during processing of effector-related language material. In an fMRI study Hauk, Johnsrude, and Pulvermüller (2004) found a nearly similar somatotopic organization while using a passive reading task involving face, leg and arm action-related verbs. With the same technique Tettamanti and colleagues (2005) showed that listening to sentences expressing actions performed with the mouth, the hand or the foot, respectively, led to signal increase in different sectors of the premotor cortex, strictly depending on the effector involved in the listened-to action-related sentence. Furthermore, Aziz-Zadeh and colleagues (2006) found in the left premotor cortex common activations both during the observation of actions and the processing of sentences expressing those same actions.

Also consistent with the involvement of the motor system during language processing are the results of a recent transcranial magnetic stimulation (TMS) study, showing that motor-evoked potentials (MEPs) recorded from hand muscles are specifically modulated during listening to sentences expressing hand-related actions, while MEPs recorded from foot/leg muscles are similarly modulated during listening to sentences expressing foot-related actions (Buccino et al., 2005).

Altogether, these data demonstrate that processing verbally presented actions modulates the activity of the motor system and suggests the existence of shared neural substrates for understanding observed actions and action-related language material. These findings may be explained by the functional properties of the ‘mirror neuron’ system, consisting of a particular set of premotor and parietal brain regions, which has been shown to play a role in action understanding (Buccino et al., 2004, Buccino et al., 2006, Rizzolatti and Craighero, 2004). This system is activated not only during the execution of actions but also during action observation and listening to common action-related sounds, both in nonhuman primates and humans (e.g., Aziz-Zadeh et al., 2004, Buccino et al., 2001, Di Pellegrino et al., 1992, Ferrari et al., 2003, Gallese et al., 1996, Keysers et al., 2003, Kholer et al., 2002, Rizzolatti et al., 1996). Furthermore, it appears to be somatotopically organized, as demonstrated by brain imaging studies (Buccino et al., 2001, Sakreida et al., 2005, Wheaton et al., 2004). From these findings, it has been argued that such a system might represent the neural substrate which allowed a gradual transition from manual gestures to speech production (Corballis, 2003, Gentilucci and Corballis, 2006, Rizzolatti and Arbib, 1998), might play a role in building up action concepts (Gallese and Lakoff, 2005, Rizzolatti and Buccino, 2005) and, by extension, might participate to language processing at least as far as action-related language material is concerned (Gallese and Lakoff, 2005, Rizzolatti and Buccino, 2005).

While suggesting a recruitment of the motor system during language processing, the studies reviewed so far do not disentangle whether the involvement of the motor system is a necessary step to understand the presented language material or rather a side effect of distinct upstream processes involved in understanding language (see also Mahon & Caramazza, 2005). To address this issue, we carried out three behavioral experiments, using a go-no go paradigm. Italian verbs expressing hand actions, foot actions or abstract contents served as stimuli. Throughout experiments we varied the linguistic task and the delivery time of the go signal. This allowed us to define the possible functional role of the motor system in different language processing tasks.