Action verbs and the primary motor cortex: A comparative TMS study of silent reading, frequency judgments, and motor imagery
Introduction
Processing sentences or verbs describing actions has been shown to involve, in addition to “classic” language areas, the motor and premotor cortex. Experimental support for this notion comes from a number of neurophysiological, behavioral, and brain imaging studies in which subjects were presented with action-related words. For example, lexical decisions about action verbs, i.e., to judge whether a verb is a real word or a pseudo-word, have been found to lead to stronger high-frequency EEG activity at recording sites located closely above primary motor (M1) cortex (Pulvermüller, Lutzenberger, & Preissl, 1999). Furthermore, if the processed action words are related to movements of different body parts, then the strongest in-going EEG current is detected close to the cortical representation of the respective body part (Pulvermüller, Harle, & Hummel, 2001). Interestingly, such a somatotopic activation of M1 has also been reported when action words related to face, arm, or leg movements are silently read only (Hauk, Johnsrude, & Pulvermüller, 2004) and even when subjects are presented with action words while they are engaged in a distractor task (Pulvermüller, Shtyrov, & Ilmoniemi, 2005a).
Transcranial magnetic stimulation (TMS) studies lend support to the idea that M1 might be involved in processing action words. Sub-threshold stimulation of the hand area of left M1 leads to a facilitatory effect (i.e., faster response times in a lexical decision task) for arm- compared to leg-action-related words, and the opposite effect has been found for leg-action-related words after stimulation of the leg area (Pulvermüller, Hauk, Nikulin, & Ilmoniemi, 2005b). The excitability of the left M1 hand area (as determined by supra-threshold stimulation and measured by motor evoked potentials, MEPs) is modulated during a transformation task involving action words as compared to non-action words (i.e., producing the singular/plural form of nouns or the 3rd person singular/plural form for verbs; Oliveri et al., 2004). Similarly, listening to hand-action-related sentences decreases the amplitude of MEPs recorded from hand muscles, while listening to sentences related to foot actions modulates the MEPs recorded from foot muscles (Buccino et al., 2005). Furthermore, functional imaging studies revealed that listening to (Tettamanti et al., 2005) and silently reading of (Aziz-Zadeh, Wilson, Rizzolatti, & Iacoboni, 2006) mouth-, hand-, and leg-action-related sentences engage the visuo-motor circuits that subserve action execution and observation in a somatotopic fashion.
Furthermore, the processing of action verbs or action-related sentences has been shown to influence overt motor behaviour. Sentences describing a movement in a certain direction can interfere with responses executed in a different direction, as measured by the action sentence compatibility effect (Glenberg & Kaschak, 2002). In the same way, listening to hand-related action sentences leads to slower reaction times for hand responses compared to foot responses (Buccino et al., 2005). Hand responses to sentences describing manual rotation are faster when both the manual response and the sentence have the same direction of rotation than when the response and the sentence differ in rotation direction. Accordingly, it has been suggested that sentences involving rotations activate a motor program for manual rotation in the listener (Zwaan & Taylor, 2006). Finally, processing of action-related verbs interferes with a concurrent reaching task (Boulenger et al., 2006). These authors also showed that the same action-related verbs could facilitate motor performance when they are processed before movement onset, suggesting that the interaction between action execution and processing of action words possesses a critical temporal dynamic. Taken together, these studies suggest that the “mere reading of action-related words activates the motor homunculus” (see de Lafuente & Romo, 2004).
To date, however, the cause of this M1 activation during action word processing remains to be elucidated. Furthermore, the question arises how M1 contributes specifically to the processing of action words. Some authors argue that it is unclear whether the M1 activation is an integral part of language processing or whether it results from a nonspecific spreading of activation from areas involved in language production to motor areas (Tokimura, Tokimura, Oliviero, Asakura, & Rothwell, 1996). Other authors suggested a specific functional connection between language areas and the hand area of the motor cortex (Meister et al., 2003, Sparing et al., 2007). Two further views have been proposed that are consistent with the idea that language understanding is processed in dedicated cortical areas (e.g., Martin, Haxby, Lalonde, Wiggs, & Ungerleider, 1995; Martin, Wiggs, Ungerleider, & Haxby, 1996), in contrast to the idea that the meaning of a sentence is accessed through amodal mental representations (e.g., Fodor, 2001, Pylyshyn, 1984). The first view, inspired by associationist theories of Hebbian learning, proposes that if a word is frequently presented together with the corresponding visual stimulus and, therefore, acquires meaning, co-activation of neurons in perisylvian and visual cortices leads to the formation of cell assemblies distributed over these perisylvian and temporo-occipital sites (Pulvermüller et al., 1999). Similarly, words which frequently occur in the context of action execution, causing neurons that process the word form and those that process the corresponding action to fire together and thus become linked, will result in word-related overlapping networks of motor and premotor cortex in a somatotopic fashion (Hauk et al., 2004, Pulvermüller et al., 1999, Pulvermüller et al., 2005a, Pulvermüller et al., 2005b). The second view proposes that the comprehension of action sentences relies on “embodied cognition” (e.g., Barsalou, 1999; Feldman & Narayanan, 2004; Gallese & Lakoff, 2005), meaning that sensorimotor representations are similarly accessed when an action is observed (Buccino et al., 2001) or when an action word is processed using the observation–execution–matching system (Aziz-Zadeh et al., 2006, Buccino et al., 2005, Tettamanti et al., 2005).
“Embodied cognition” refers also to mental simulation. Processing action words may result in M1 activation, because action words could trigger motor simulation, a process known to activate M1 cortex somatotopically (e.g., Ehrsson, Geyer, & Naito, 2003; Stippich, Ochmann, & Sartor, 2002). Motor representations may be implicitly triggered during action word processing, because subjects, although not explicitly instructed to imagine themselves or somebody else performing the movements, might implicitly simulate the respective action when processing action words. In fact, the above-mentioned studies did not control for putative motor imagery processing. Modulations of M1 activity contingent upon action word processing may be found even in the absence of specific cognitive demands, e.g., as it happens during silent reading tasks. In silent reading tasks, subjects are not engaged in any additional cognitive operation with the stimuli and are, therefore, free to think about the corresponding action during or after the silent reading. Thus, they might implicitly simulate the movement described in the action word, which in turn may activate M1. The idea that language understanding may trigger mental simulation (motor and visual type) is not new (e.g., Barsalou, 1999; Feldman & Narayanan, 2004; Gallese & Lakoff, 2005; Glenberg & Robertson, 2000).
Taken together, previous studies have shown that reading action words as well as motor imagery may activate the hand area of left M1. To date, however, the relationship between these two observations has not been investigated within the same experiment. Therefore, we have chosen to examine the specific contribution of M1 to action word processing by comparing the effects of single pulse TMS while right-handed subjects performed different cognitive tasks using identical hand-action-related verbs. In particular, we investigated whether sub-threshold TMS of the left M1 hand area differentially modulates task performance. For control, TMS was delivered to the vertex. The silent reading of hand-action-related verbs, which were presented in the infinitive form (e.g., “aufschrauben”, i.e., “to screw”), was kept constant across conditions, whereas the cognitive operation performed during the reading varied: subjects (i) indicated when their silent reading was completed; (ii) mentally simulated the actions, and (iii) estimated the word frequency. Unlike the above-mentioned studies, in which, during silent reading, subjects were not asked to perform any specific cognitive operation other than reading (i.e., they had the time and were free to think about/simulate the respective actions), the purpose of the mental simulation and the frequency task employed in our study was to control the cognitive set. In fact, the question about hand rotation that subjects had to answer in the mental simulation condition prompted subjects to perform motor imagery. In contrast, the question about word frequency led subjects to concentrate on the meaning of the verb and, in addition, prevent them from implicitly performing imagery.
According to a previous TMS study of reading action words (Pulvermüller et al., 2005b), sub-threshold TMS, when applied early (at about 150 ms above M1), should produce a facilitatory effect on action word processing. We reasoned that if reading of action verbs per se triggers M1 activation, a similar (facilitatory) effect of TMS to M1 (TMSM1) should be found for all three tasks, since all tasks involve action verb reading. In contrast, if a specific task component activates M1, then we expected to observe task-dependent differential TMS effects on subjects’ performance. Although based on different assumptions, the two theoretical accounts mentioned above (i.e., embodied cognition theory and associationist theory) lead to identical predictions with respect to the results of our study, since both accounts hold that the language and motor systems share overlapping neuronal representations. The only difference between these two accounts is why M1 activation occurs. Both accounts predict that reading grasp the bottle will activate the motor plan for “grasping the bottle” in the listener either by the observation–execution–matching system (embodied cognition) or as the result of Hebbian learning (associationist theory). Crucially, both accounts predict a facilitatory effect of TMS to M1 for all three tasks. Alternatively, M1 might be involved in semantic processing of action words through motor imagery. If this hypothesis holds true, then we should observe a differential facilitatory effect of hand motor cortex TMS on the imagery task and no effect on the pure silent reading task and the frequency judgments. Such a pattern of results would imply that M1 activation found in previous studies might have resulted from the subjects’ strategy to mentally simulate the movements during the processing of action words (Tomasino, Werner, Weiss, & Fink, 2007).
Section snippets
Participants
Twenty healthy men (mean age 27.4 ± 7.7 years) gave informed consent to participate in the study. They were all right-handed (Edinburgh Inventory test; Oldfield, 1971), native German speakers with comparable levels of education. None of the subjects had a history of neurological or psychiatric disorders. The study was approved by the local ethics committee.
Procedure
Throughout the experiment, subjects sat in a comfortable reclining armchair in front of a computer screen at a distance of 57 cm. Stimuli were
Subjects’ judgments
In the I task, consistent with the results of the pilot study, “rotation-related” verbs elicited significantly more “Yes” (71.7%) than “No” (28.3%) responses to the “presence of rotational wrist/hand movements” question (t(12) = 9.0, p < 0.001), whereas the opposite pattern was found for the “non-rotation-related” verbs (more “No” [86.1%] than “Yes” [13.9%] responses, t(12) = −7.8, p < 0.001). In the FJ task, consistent with the CELEX database, verbs with a high-written frequency elicited significantly
Task-dependent modulation of reaction times after TMSM1: facilitation
In this TMS study, we investigated the nature of the previously reported primary motor cortex (M1) involvement in the processing of action words (e.g., Hauk et al., 2004). To examine whether reading of action words per se or rather certain task components modulate M1 activity, we applied single-pulse TMS above the left M1 (TMSM1) or, for control, the vertex (TMSvertex), while subjects performed three different tasks (silent reading, motor imagery, and frequency judgment). The main finding of
Acknowledgments
We thank our volunteers and our colleagues from the Cognitive Neurology Group, especially Dr. Dennis A. Nowak, for valuable suggestions on the experimental design; Dr. Corrado Corradi Dell’Acqua, for help with the data analysis; and Oliver Haumann, for his expert technical assistance. The Alexander von Humboldt foundation is gratefully acknowledged for supporting B.T. Further support was provided by the DFG KFO 112 (TP1, GRF).
References (56)
- et al.
Congruent embodied representations for visually presented actions and linguistic phrases describing actions
Current Biology
(2006) - et al.
Parietal magnetic stimulation delays visuomotor mental rotation at increased processing demands
Neuroimage
(2002) - et al.
Listening to action-related sentences modulates the activity of the motor system: a combined TMS and behavioral study
Cognitive Brain Research
(2005) - et al.
Language abilities of motor cortex
Neuron
(2004) - et al.
Sparing of verbs and preserved, but ineffectual reading in a patient with impaired word production
Cortex
(1995) - et al.
Embodied meaning in a neural theory of language
Brain and Language
(2004) - et al.
Task instructions influence the cognitive strategies involved in line bisection judgements: evidence from modulated neural mechanisms revealed by fMRI
Neuropsychologia
(2002) - et al.
Cerebellar involvement in verb generation: an fMRI study
Neurosci. Lett.
(2006) - et al.
Symbol grounding and meaning: A comparison of high-dimensional and embodied theories of meaning
Journal of Memory and Language
(2000) - et al.
Somatotopic representation of action words in human motor and premotor cortex
Neuron
(2004)
Reactivation of physical motor information in the memory of action events
Brain Research
Motor cortex hand area and speech: implications for the development of language
Neuropsychologia
The assessment and analysis of handedness: The Edinburgh Inventory
Neuropsychologia
Early semantic context integration and lexical access as revealed by event-related brain potentials
Biological Psychology
Walking or talking? Behavioral and neurophysiological correlates of action verb processing
Brain and Language
Action comprehension in aphasia: linguistic and non-linguistic deficits and their lesion correlates
Neuropsychologia
Somatotopic mapping of the human primary sensorimotor cortex during motor imagery and motor execution by functional magnetic resonance imaging
Neuroscience Letters
Delay in simple reaction time after focal transcranial magnetic stimulation of the human brain occurs at the final motor output stage
Brain Research
The CELEX lexical database (CD-ROM)
Perceptual symbol systems
Behavioral Brain Science
Cross-talk between language processes and overt motor behavior in the first 200 ms of processing
Journal of Cognitive Neuroscience
Focal transcranial magnetic stimulation and response bias in a forced-choice task
Journal of Neurology, Neurosurgery, and Psychiatry
Action observation activates premotor and parietal areas in a somatotopic manner: an fMRI study
European Journal of Neuroscience
Functional relevance of cross-modal plasticity in blind humans
Nature
Imagery of voluntary movement of fingers, toes, and tongue activates corresponding body-part specific motor representations
Journal of Neurophysiology
The involvement of primary motor cortex in mental rotation revealed by transcranial magnetic stimulation
European Journal of Neuroscience
Speech listening specifically modulates the excitability of tongue muscles: a TMS study
European Journal of Neuroscience.
On the benefits of not trying: brain activity and connectivity reflecting the interactions of explicit and implicit sequence learning
Cerebral Cortex
Cited by (119)
Electrophysiological alterations during action semantic processing in Parkinson's disease
2024, Neurobiology of AgingThe brain stimulation of DLPFC regulates choice preference in intertemporal choice self-other differences
2023, Behavioural Brain ResearchDoes the involvement of motor cortex in embodied language comprehension stand on solid ground? A p-curve analysis and test for excess significance of the TMS and tDCS evidence
2022, Neuroscience and Biobehavioral Reviews