Resumes & CVs

Effects of TMS on Different Stages of Motor and Non-Motor Verb Processing in the Primary Motor Cortex

Description
The embodied,cognition hypothesis suggests,that motor and premotor,areas are automatically and necessarily involved in understanding action language, as word conceptual representations are embodied. This transcranial magnetic stimulation (TMS) study
Categories
Published
of 11
5
Categories
Published
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Transcript
  Effects of TMS on Different Stages of Motor and Non-Motor Verb Processing in the Primary Motor Cortex Liuba Papeo 1 , Antonino Vallesi 2 , Alessio Isaja 1 , Raffaella Ida Rumiati 1 * 1 Sector of Cognitive Neuroscience, Scuola Internazionale Superiore di Studi Avanzati (SISSA), Trieste, Italy,  2 Rotman Research Institute - Baycrest, Toronto, Ontario,Canada Abstract The embodied cognition hypothesis suggests that motor and premotor areas are automatically and necessarily involved inunderstanding action language, as word conceptual representations are embodied. This transcranial magnetic stimulation(TMS) study explores the role of the left primary motor cortex in action-verb processing. TMS-induced motor-evokedpotentials from right-hand muscles were recorded as a measure of M1 activity, while participants were asked either to judgeexplicitly whether a verb was action-related (semantic task) or to decide on the number of syllables in a verb (syllabic task).TMS was applied in three different experiments at 170, 350 and 500 ms post-stimulus during both tasks to identify whenthe enhancement of M1 activity occurred during word processing. The delays between stimulus onset and magneticstimulation were consistent with electrophysiological studies, suggesting that word recognition can be differentiated intoearly (within 200 ms) and late (within 400 ms) lexical-semantic stages, and post-conceptual stages. Reaction times andaccuracy were recorded to measure the extent to which the participants’ linguistic performance was affected by theinterference of TMS with M1 activity. No enhancement of M1 activity specific for action verbs was found at 170 and 350 mspost-stimulus, when lexical-semantic processes are presumed to occur (Experiments 1–2). When TMS was applied at 500 mspost-stimulus (Experiment 3), processing action verbs, compared with non-action verbs, increased the M1-activity in thesemantic task and decreased it in the syllabic task. This effect was specific for hand-action verbs and was not observed foraction-verbs related to other body parts. Neither accuracy nor RTs were affected by TMS. These findings suggest that thelexical-semantic processing of action verbs does not automatically activate the M1. This area seems to be rather involved inpost-conceptual processing that follows the retrieval of motor representations, its activity being modulated (facilitated orinhibited), in a top-down manner, by the specific demand of the task. Citation:  Papeo L, Vallesi A, Isaja A, Rumiati RI (2009) Effects of TMS on Different Stages of Motor and Non-Motor Verb Processing in the Primary MotorCortex. PLoS ONE 4(2): e4508. doi:10.1371/journal.pone.0004508 Editor:  Pier Francesco Ferrari, Universita` di Parma, Italy Received  September 30, 2008;  Accepted  January 15, 2009;  Published  February 25, 2009 Copyright:    2009 Papeo et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the srcinal author and source are credited. Funding:  This research was supported by a PRIN to RIR. The funders had no role in study design, data collection and analysis, decision to publish, or preparationof the manuscript. Competing Interests:  The authors have declared that no competing interests exist.* E-mail: rumiati@sissa.it Introduction How are individual words represented in the brain and whatcognitive operations are required in order to understand them? According to the classical cognitive theories, word representationsare abstract and amodal; in other words, they are independent of the sensory and sensorimotor properties of the objects to whichthey refer (e.g. [1,2 ]). Over the last two decades this view has been challenged by evidence to the effect that when people processwords related to actions, motor and premotor areas are activated(see [3] for a review), in addition to the  classical   language-processing areas [4]. For instance, Oliveri  et al.  [5] appliedtranscranial magnetic stimulation (TMS) while their participantsperformed a morphological transformation task with action-related nouns and verbs and found that activity in the left primarymotor cortex (M1) was enhanced regardless of the word’sgrammatical class. Using a fMRI study, Hauk   et al.  [6] showedthat silent reading of action words referring to face-, arm- or leg-related actions, activated areas in the left premotor and primarymotor cortices differentially. Tettamanti  et al.  [7] reportedsomatotopic activation in the left premotor cortex - but not inM1 - during passive listening to sentences implying mouth-, hand-or leg-actions. In both of these imaging studies the motor contentof the linguistic material was argued to have automaticallyrecruited the motor programs of the described actions.These and similar observations have been considered to supportthe embodied cognition hypotheses of language. Exploiting theHebbian model of correlational learning  [8], Pulvermuller [9] proposed that action words which are typically learned in thecontext of action performance are represented in the sensorimotorcircuits associated with the implied action, woven within theperisylvian-language areas. The  simulationist view   holds thatlanguage understanding is achieved via mental simulation of itscontent by activating traces of previous perceptual and motorexperiences in the perceiver’s sensorimotor system [10 – 12 ]. Even though they assume different underlying mechanisms, i.e.associative learning and mental simulation, both proposals predictthat understanding action-language  automatically  entails the motorprograms of the corresponding physical actions ( [3,11]; see [13] for a review). It follows that a conceptual representation, far frombeing abstract and symbolic,  is   in fact sensory and motorinformation.If action-language processing activates motor representationsautomatically, then motor activation should occur even when the PLoS ONE | www.plosone.org 1 February 2009 | Volume 4 | Issue 2 | e4508  participants’ attention is diverted from the motor content of aword. However, in a recent fMRI study in which task requirements were controlled, M1 activation was only observedin a task in which participants had to imagine the content of motorphrases explicitly, but not in a letter-detection task with the sameitems [14]. Motor imagery can be triggered even in the absence of explicit instructions, as a strategy to perform any task withsensorimotor components (e.g. [15 – 17 ]), and it seems to involve M1 particularly when stimuli evoke movements of human bodyparts (e.g. [18 –20 ]). This suggests the possibility that M1 is not an integral part of the network for action-word representation but isrecruited only to accomplish tasks that critically require theretrieval of sensorimotor attributes associated with words.Light can be thrown on the question as to whether motoractivation is automatic for action-language understanding byestablishing the exact time interval during which M1 activityenhances. Word recognition processing is in fact multistage,characterized by lexical, syntactic, semantic and post-conceptualstages, each with its own specific time course. The hypothesis thatM1 is an integral part of word representation implicates that itshould be active during the lexical-semantic access, i.e., within200 ms. Electrophysiological studies seem to support this hypoth-esis [21 – 23]. Pulvermu¨ller  et al.  [23], using magnetoencephalog-raphy (MEG) while participants listened passively to a stream of action words and pseudo-words, reported that a short-lived activityoccurred in frontocentral regions within 200 ms after action wordsappeared. As the technique used has limited spatial resolution, theauthors could only conclude that the processing of action wordswas maintained by ‘‘ different parts of frontocentral cortex, possiblyincluding the prefrontal, premotor and motor areas  ’’ ( [23], p. 889). On theother hand, in TMS studies, where the temporal resolution iscombined with a more precise spatial resolution, language-inducedmodulation of M1 was found to be either an early phenomenon(e.g., arising when listening to an action verb, before itspresentation was over [24] ) or a late phenomenon (500 ms post-word, [5] ).The present study addresses two questions. First, does motoractivation occur  automatically  even when participants perform a task that barely requires the explicit retrieval of the motor content of the word? Second, which of the different stages of wordrecognition is most likely to activate the left M1? It was possibleto provide answers to these questions using TMS, given theintrinsic characteristics of this technique. In fact, when appliedsupra-threshold to M1 at a given point in time, TMS elicits MEPsfrom body muscles as a direct measure of motor excitability at thattime: the degree of MEP amplitude following TMS is proportion-ate to the level of M1 activity. Moreover, when TMS is applied toa brain area it delays or disrupts the ongoing behavior [25].Therefore the question as to whether M1 activity is necessary foraction-language processing can be answered through an investi-gation of how behavioral performance changes when M1-activityis temporarily altered.In the present study, TMS was applied to the left hand-M1 andMEPs were recorded from hand muscles while participantsperformed a semantic task and a syllabic-segmentation task. Inthe semantic task, participants were instructed to judge whetherthe presented verbs were action-related, which required explicitretrieval of the representation of the described physical actions. Inthe syllabic-segmentation task, participants were asked to indicatethe number of syllables constituting each verb. The syllabicsegmentation primarily entails the sub-lexical features of a word,namely, its orthographical-phonological representation [e.g., 26].The semantic activation is rather automatic in visual wordrecognition [e.g., 27,28], but it might be only implicit when theretrieval of word meaning is not necessary in order to perform thetask, as is in the syllabic segmentation. If M1 activation isautomatic in response to action words, it follows that whenparticipants perform both syllabic segmentation and semanticencoding tasks, MEPs should be greater for action than for non-action verbs.Three separate experiments were set up in which TMS wasapplied at a different point in time after word onset (hereafter,‘‘post-stimulus’’ will be used synonymously). In Experiment 1,TMS was delivered 170 ms post-stimulus, as the findings of Event-related Potential (ERP) studies indicate that the lexical access for visually-presented words occurs between 100 and 200 ms post-stimulus in posterior regions [29], while there is evidence for earlysemantic processes starting prior to 200 ms in anterior regions[21,30]. In Experiment 2, TMS was triggered 350 ms post-stimulus, the time when the brain is thought to encode category-specific attributes of word meaning. In fact, a greater negativity(N400 component) in posterior regions was observed over the300–350 ms latency range for motor words compared to visual orabstract words [31]. In a similar latency range, differences inparietal and frontal positivity (P300- like   ) correlate with more fine-grained aspects of action-word meaning, such as the bodysegments involved in the implied action [30]. In Experiment 3,TMS was applied 500 ms post-stimulus, during the post-conceptual stages of word recognition. Manipulating the delaybetween stimulus onset and magnetic stimulation across experi-ments served to investigate the time-course of M1 activity whenparticipants performed different linguistic tasks. It also provided amethodological control for distracting/alerting effects and acousticand tactile sensations associated with TMS. This control is basedon the assumption that ‘‘non specific effects of TMS will beindependent, whereas the behavioral effects will be highlydependent on the precise interval between the event and thestimulation’’ (see [32] p. 951). This proves to be particularlyappropriate for a single-pulse TMS protocol, where stimulus andpulse are not delivered simultaneously [33]. In addition to MEPs,response accuracy and reaction times (RTs) were collected asmeasures of the participants’ linguistic performance. Thus, besidesidentifying the mental operations that most likely modulate M1(explicit or implicit encoding of motor content), insight was gainedas to  when  M1 is recruited and  what   the nature of its relationship(causal?) is with linguistic performance. Results Experiment 1: measurement of M1 activity during lexical-semantic access Eleven right-handed, native Italian speakers participated in thisexperiment. They were exposed to separate blocks of verbs,selected through a pilot study (see Methods), and were instructedto judge whether they were action or non-action verbs (semantictask) and to indicate the number of syllables (three or othernumber) composing the verb (syllabic task), through a yes-or-no verbal response. Single-pulse TMS was applied to the left hand-M1 170 ms post-stimulus to elicit MEPs in the first dorsalinterosseus (FDI) muscle of the right hand. Each task wascomposed of two blocks, one using TMS and the other shamstimulation as a control.Previous TMS studies found an increased hand-M1 activityfollowing rather heterogeneous sets of stimuli, including action verbs related to several body effectors, nouns of manipulableobjects [5], or even concrete nouns such as ‘‘house’’ and ‘‘collar’’[34]. Similar findings have been often explained in the context of an evolutionary scenario whereby language is conceptualized as Language and Motor SystemPLoS ONE | www.plosone.org 2 February 2009 | Volume 4 | Issue 2 | e4508  having evolved from manual communication [34 – 36]. On the other hand, evidence exists that M1 is activated by language in asomatotopic fashion, reflecting the different body-effectors of theimplied actions (see [3] for a review). We stimulated the hand-M1,while subjects processed both non action and action verbs. Giventhe uncertainty about the specific involvement of hand-M1 inlanguage, we considered hand-action and non-hand action verbsas separate levels of the verb-category factor. This yielded a2 6 2 6 3 experimental design with within-subjects factors: (i)stimulation condition (TMS to M1 vs. sham), (ii) task (semantic vs. syllabic), (iii) verb category (hand-action vs. non-hand action vs.non action). Table 1, 2 and 3 summarize the results of the RT,accuracy and MEP analyses, respectively, for the three experi-ments. RTs.  The repeated-measures ANOVA revealed a significantmain effect of task   F  (1,10)=6,27,  p =0.03, such as semantic judgments were faster than syllabic judgements (770 6 72 vs.876 6 95 ms; mean 6 sem). Was also significant the effect of category,  F  (2,20)=10.33,  p , 0.001, with action verbs (both handand non-hand related) being processed faster than non action verbs (804 6 82 and 795 6 80 vs. 870 6 88 ms;  p s , 0.001). The task x category interaction was significant,  F  (2,20)=10.33,  p , 0.001,suggesting that the effect of verb category was dependent on thetype of task performed (see Figure 1A). Post-hoc analysis (LSDFisher’s test,  a  # .05) revealed that, in the semantic task,participants responded faster to hand- and non-hand action verbs than non action verbs, (726 6 69 and 716 6 70 vs.867 6 76 ms;  p s , 0.001), with no difference between the twoaction-verb categories (   p . 0.1). Instead, the three verb categoriesdid not differ in the syllabic task (884 6 96 and 875 6 90 vs.872 6 99 ms;  p s . 1). This effect was independent of TMS, as theinteraction between stimulation condition, task and category didnot approach significance,  F  (2,20) , 1, n.s. Thus, the semanticencoding was faster for the action than for the non action verbs.This difference disappeared when performing the task did not relyon the word meaning, as in the syllabic segmentation.  Accuracy.  The effect of task resulted significant, F  (1,10)=5.13,  p , 0.05, semantic judgments being more accuratethan syllabic judgments (0.90 6 0.03 vs. 0.86 6 0.03, meanproportion of correct responses 6 sem). There was a trend for theinteraction between task and category,  F  (2,20)=2.77,  p =0.08 (seeFigure 2A). Post-hoc comparisons showed that the semantic judgments were more accurate on the two categories of action verbs, relative to non action verbs (0.93 6 0.02 and 0.92 6 0.02 vs.0.83 6 0.04;  p s , 0.05), whereas the syllabic task was performedequally well with the three categories (0.86 6 0.03 and 0.84 6 0.03 vs. 0.86 6 0.02,  p s . 0.6). This pattern was consistent with the RTresults, and allows us to rule out the speed-accuracy trade-off as anexplanation for the observed performance. MEPs.  MEPs (mV) recorded from the right FDI muscleduring TMS delivery, were normalized. Mean z-scores of MEPpeak-to-peak amplitude were subjected to a 2 6 3 repeated-measures ANOVA with task and category as factors. No effector interaction approached significance (all  p s . 0.2). Although Table 1.  Mean RTs (ms) in all experimental conditions of Experiments 1–3. Semantic task Syllabic task Hand-act Non-hand act Non-act Hand-act Non-hand act Non-act Experiment 1 TMS 730 726 853 948 905 909Sham 722 705 882 818 844 836Experiment 2 TMS 781 845  962 1156 1166  1105Sham 774 848  886 1065 1101  1111Experiment 3 TMS 627 683 713 843 856 862Sham 560 612 683 767 814 762Tabled mean RTs (ms) following the semantic and the syllabic processing of the hand-action ( Hand-act  ), the non-hand action ( Non-hand act  ) and the non action ( Non-act  ) verbs, during TMS and sham stimulation, in Experiments 1–3. The regions in bold type showed the only significant differences between TMS and sham (Experiment2).doi:10.1371/journal.pone.0004508.t001 Table 2.  Mean Accuracy (proportion of correct responses) in all experimental conditions of Experiments 1–3. Semantic task Syllabic task Hand-act Non-hand act Non-act Hand-act Non-hand act Non-act Experiment 1 TMS 0.93 0.93 0.83 0.87 0.84 0.87Sham 0.94 0.91 0.84 0.86 0.86 0.86Experiment 2 TMS 0.97 0.88 0.88 0.89 0.86 0.88Sham 0.99 0.92 0.88 0.93 0.90 0.93Experiment 3 TMS 0.94 0.90 0.84 0.83 0.84 0.82Sham 0.93 0.90 0.90 0.84 0.84 0.82Tabled mean accuracy (proportion of correct responses) following the semantic and the syllabic processing of the hand-action ( Hand-act  ), the non-hand action ( Non-hand act  ) and the non action ( Non-act  ) verbs, during TMS and sham stimulation, in Experiments 1–3. A difference between TMS and sham stimulation was neverobserved.doi:10.1371/journal.pone.0004508.t002 Language and Motor SystemPLoS ONE | www.plosone.org 3 February 2009 | Volume 4 | Issue 2 | e4508  action verbs enjoyed a temporal advantage on non action verbs, insemantic processing, the resulting early lexical-semantic access didnot elicit a specific enhancement of M1 activity. Experiment 2: measuring M1 activity during semantic-attribute processing  A total of 14 right-handed, native Italian speakers took part inExperiment 2. The experimental design and statistical analysiswere identical to those used in Experiment 1, with the soledifference that, here, TMS was applied 350 ms post-stimulus. RTs.  The effect of task was significant,  F  (1, 13)=13.53,  p =0.003, as was the effect of category,  F  (2,26)=12.00,  p =0.001.RTs in the syllabic task were slower than those in the semantic task (1117 6 104 vs. 849 6 66 ms), and hand-action verbs wereprocessed faster than non-hand action verbs and non action verbs (944 6 82 vs. 989 6 83 and 1016 6 89 ms;  p s , 0.01). The task x category interaction resulted significant,  F  (2, 26)=10.4,  p , 0.01(see Figure 1B). In the semantic task, hand-action and non-handaction verbs were processed faster than non action verbs (778 6 61and 846 6 63 vs. 924 6 73;  p s , 0.01). The difference between thetwo action-verb categories was also significant, with hand-verbsbeing judged faster than non-hand action verbs (   p , 0.01). Thethree categories did not differ in the syllabic task (1111 6 105 and1133 6 104 vs. 1108 6 105;  p s . 0.2). A significant TMS x task xcategory interaction was also found,  F  (2, 26)=4,27  p =0.02(Table 1). In the semantic task, TMS further delayed theparticipants’ performance on non action verbs compared withthe sham condition (   p =0.02). No effect of TMS was observed inthe semantic task for the two action-verb categories (   p s . 0.1).Conversely, in the syllabic task TMS slowed down responses toboth action-verb categories compared with the sham condition(   p , 0.05). As for Experiment 1, participants processed action verbsfaster than non-action verbs in the semantic but not in the syllabictask. In addition, the three-way interaction revealed that TMSdelivery to M1 inhibited participants’ responses when theyperformed the semantic task with non action verbs, and whenthey performed the syllabic task with action verbs.  Accuracy.  The ANOVA revealed a significant main effect of TMS,  F  (1, 13)=6,83,  p =0.02, with participants being lessaccurate during TMS than sham stimulation (0.90 6 0.02 vs.0.92 6 0.02) . However, this factor did not interact with any otherfactor in the design. The main effect of category was alsosignificant,  F  (2, 26)=10,59,  p , 0.001, such as processing hand-action verbs was more accurate than processing hand-action andnon-action verbs (0.95 6 0.01 vs. 0.89 6 0.02 and 0.89 6 0.02;  p s , 0.001). The ANOVA also indicated a significant task xcategory interaction,  F  (2, 26)=4,71,  p =0.02 (see Figure 2B).Particularly, participants performed semantic judgements moreaccurately on hand-action verbs than on non-hand action andnon-action verbs (0.98 6 0.007 vs. 0.90 6 0.02 and 0.88 6 0.03;  p s=0.001). No difference was observed between action and non-action verbs in the syllabic task (0.91 6 0.02 vs. 0.88 6 0.02 and0.90 6 0.02;  p s . 0.1). This ruled out the speed-accuracy trade-off effect as an explanation for the RT results, confirming that thesemantic task was more difficult when non action verbs wereinvolved, and that this difference disappeared when processing phonological aspects of verbs. MEPs.  The ANOVA of mean MEP peak-to-peak amplitudesrevealed only a trend for the effect of category,  F  (2, 26)=3,26,  p =0.05. The MEP amplitude was the greatest for non-action verbs. However, the lack of interaction between task and category,[  F  (2, 26) , 1, n.s.] did not support any obvious conclusionregarding a specific involvement of left M1 in word processing. Experiment 3: measuring M1 activity during post-conceptual processing Experiment 3 involved 11 new participants. The procedureswere identical to that of Experiments 1–2, the sole difference being that the delay between stimulus onset and TMS delivery was500 ms. RTs.  The RT analysis revealed a significant main effect of task,  F  (1, 10)=13.95,  p , 0.01, with the semantic task being fasterthan the syllabic (646 6 103 vs. 817 6 135 ms). The effect of category was also significant  F  (2, 20)=6.03,  p , 0.01. Again, hand-action verbs were processed faster than non-hand and non action verbs (699 6 115 vs. 742 6 125 and 755 6 116 ms;  p s , 0.05).Basically the pattern of the interaction between task andcategory was comparable to Experiments 1–2 (see Figure 1C),although it did not approached significance,  F  (2,20) , 1, n.s.  Accuracy.  The ANOVA revealed a significant main effect of task,  F  (1,10)=5,27,  p =0.04: the semantic task was performedbetter than the syllabic task (0.90 6 0.02 vs. 0.83 6 0.004).Descriptively, the pattern of participant’s performance in thetwo tasks, with the three verb-categories was consistent toExperiments 1–2 (see Figure 2C), but no interaction approachedsignificance (all  p s . 0.1). MEPs.  The ANOVA showed a significant task x categoryinteraction,  F  (1, 10)=8.872,  p =0.01. Post-hoc analyses showedthat the semantic processing of hand-action verbs elicited greatermotor activation compared with non action verbs (   p =0.03),whereas the MEP amplitude for non-hand action did not differfrom that for non-action verbs (   p . 0.1). In the syllabic task, thedifference in the level of M1 excitability after processing hand-action verbs was significantly smaller relative to non-hand actionand non action verbs (   p s , 0.03). Again, there was no difference inthe MEP amplitude between non-hand action and non action Table 3.  Means of normalized (sem) MEP peak-to-peak amplitudes in all experimental conditions of Experiments 1–3. Semantic task Syllabic task Hand-act Non-hand act Non-act Hand-act Non-hand act Non-act Experiment 1  2 0.05 (0.05) 0.06 (0.08)  2 0.02 (0.06) 0.03 (0.07)  2 0.09 (0.06)  2 0.003 (0.03)Experiment 2 0.04 (0.05)  2 0.17 (0.07) 0.06 (0.04)  2 0.04 (0.08)  2 0.09 (0.09) 0.06 (0.04)Experiment 3  0.12 (0.04)  0.07 (0.08)  2 0.07 (0.04)  2 0.19 (0.07)  0.01 (0.07) 0.05 (0.05)Tabled mean normalized MEP amplitude following the semantic and the syllabic processing of the hand-action ( Hand-act  ), the non-hand action ( Non-hand act  ) and thenon action ( Non-act  ) verbs in Experiments 1–3. The regions in bold type showed the facilitation in the semantic task and the inhibition in the syllabic task, for hand-action verbs only, as compared to non-action verbs (Experiment 3). A similar dissociation between tasks was not observed for the other verb categories in any of thethree experiments.doi:10.1371/journal.pone.0004508.t003 Language and Motor SystemPLoS ONE | www.plosone.org 4 February 2009 | Volume 4 | Issue 2 | e4508   verbs (   p =0.6). A difference was observed in the MEP amplitudebetween the semantic and syllabic processing of hand-action verbs(   p =0.001): M1 activity significantly increased and decreaseddepending on whether the same verbs were processed semanticallyand syllabically, respectively. A similar difference between taskswas not observed for the other two categories (   p s . 0.1). Thus, theenhancement of M1 activity occurred only when participantsexplicitly encoded the content of the hand-action verbs, but notwhen they encoded their phonology. In the latter condition, M1activity resulted to be rather inhibited. Mean normalized MEPamplitudes for all the conditions of Experiments 1–3 are listed inTable 3. Between-subjects analysis MEP data from all the three experiments were subjected to an ANOVA with factors, 2 task and 3 category manipulated withinsubjects, and 3 timing of TMS delivery as a between-subjectsfactor. This analysis was performed in order to investigate thetime-course of M1 activity associated with each verb categoryduring their semantic and syllabic processing. The three-wayinteraction between task, category and TMS timing approachedsignificance,  F  (4,66)=2,1656,  p =0.08 (see Figure 3). Post-hoccomparisons revealed a different pattern of M1-activity for hand- vs. non-hand action verbs, when compared with non action verbs. Hand-action verbs.  At the first two timings of TMS delivery(i.e., 170 and 350 ms post-stimulus), MEP amplitude for hand-action verbs was not different from that for non action verbs, ineither task (all  p s . 0.3). Moreover, at these latencies, MEPamplitude following the semantic and the syllabic processing of hand-action verbs did not differ (   p . 0.3). When recorded at500 ms post-stimulus, the difference between MEPs for hand-action verbs and non-action verbs approached significance in thesemantic task (   p =0.07), and reached significance in the syllabictask (   p =0.03). Moreover, M1 activity for hand-action verbsresulted greater in the semantic than in the syllabic task (   p , 0.01).Confirming those from the individual experiments, these findingssuggest that M1 activity is modulated by hand-action verbprocessing only during post-conceptual stages of wordrecognition (500 ms post-stimulus), with the direction of themodulation (increase or decrease) depending on the task-demand. Non-hand action verbs.  The pattern of M1-activityfollowing non-hand action verbs proved to be different from thatof hand-action verbs. The MEP amplitude associated with this verb category did not differ from that of non-action verbs in eithertask condition, and at any time interval (all  p s . 0.2), except at350 ms. At this latency, the MEP amplitude decreased for non-hand action verbs relative to non action verbs in the semantic task only (   p =0.02). In the same condition, it was significantly smallereven when compared with the MEP amplitude associated withhand-action verbs (   p =0.03). A difference between the two action- verb categories was also observed at 500 ms: here, MEP amplitudefor non-hand action was greater than that for hand-action verbs inthe syllabic task (   p =0.02). This pattern suggests that the semanticand the syllabic processing of the non-hand action verbs did notelicit motor facilitation and inhibition, respectively, whencompared with non-action verbs. No difference was observed inMEP amplitude when the same non-hand action verbs weresubjected to the two tasks (   p =0.5), so that the dissociation in MEP Figure 1. Mean RTs (ms) as a function of the tasks (semanticand syllabic)for theverb categories (hand-action, ‘‘ hand   ’’; non-hand action, ‘‘ non-hand   ’’; and non action, ‘‘ non-act  ’’).  Verticalbars denote the Standard Error of the mean.  (A)  Experiment 1: both thehand-action and the non-hand action verbs were processed faster thanthe non action verbs in the semantic task; no difference was observedbetween the verb categories in the syllabic task.  (B)  Experiment 2: bothaction-verb categories were processed faster than the non action verbswith an advantage of the hand-action over the non-hand action verbs,in the semantic task. RTs for the three categories did not differ in thesyllabic task.  (C)  Experiment 3: the pattern of performance wasconsistent with that of Experiments 1–2, although the interaction didnot approach significance.doi:10.1371/journal.pone.0004508.g001Language and Motor SystemPLoS ONE | www.plosone.org 5 February 2009 | Volume 4 | Issue 2 | e4508
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x