- Email: [email protected]
Laterality effects on motor awareness
Neuropsychologia 40 (2002) 1379–1386
Laterality effects on motor awareness Elena Daprati∗ , Angela Sirigu Institut des Sciences Cognitives, 67 Bd Pinel, 69675 Bron, France Received 26 January 2000; accepted 28 September 2001
Abstract The ability to perceive self-produced movements and to correctly attribute an action to its proper agent is a natural task and a basic requirement to human social communication. Recent experiments suggest that this apparently simple phenomenon is related to the mechanisms of motor production, raising the question whether recognition of self-produced movement is affected by asymmetries similar to those present in motor skills. In this study, right- and left-handed subjects decided whether a moving hand presented on a screen was the image of their own hand or of that of another person. Two experimental conditions were tested: either subjects saw their own moving hand (subject condition) or they were shown the experimenter’s hand pantomiming their movement (experimenter condition); a glove masked morphological differences between the two hands. Verbal responses and response times were analysed. Results showed that all subjects were more accurate in recognising their preferred hand with respect to their non-preferred hand. Right-handers responded faster than left-handers did, the latter group being especially slowed down in the experimenter condition. On the contrary, in the right-handers group, response times did not differ among conditions. The present data indicate that the ability to recognise self-generated movements is affected by motor dominance, thus suggesting that conscious knowledge of self-produced movements is closely related to the motor system. © 2002 Elsevier Science Ltd. All rights reserved. Keywords: Motor consciousness; Handedness; Motor skills; Manual asymmetry
1. Introduction Over 90% of humans are right-handed, that is they show a clear preference for the use of the right hand in the execution of skilled movements. Although this notion is well-established and seems to be rather consistent across cultures and over time [3,11], the neurobiological correlate of handedness is still a matter of debate. Neuropsychological evidence suggests that this behavioural lateralisation can be related to an asymmetry of the cortical areas controlling the cognitive–motor requirements of skilled movements [14]. As a general rule, movement of one limb is controlled by the motor cortex of the contralateral hemisphere: however, at least in right-handers, a dominant role of the left hemisphere for praxis functions has frequently been invoked. For instance, ipsilateral deficits in tasks involving complex cognitive–motor abilities have been described only following left (but not right) hemisphere damage [14]. Moreover, ideomotor apraxia—i.e. a specific impairment in the performance of complex gestures in the absence of elementary sensory/motor deficits—is known to occur more frequently after left than after right hemisphere damage [12]. ∗
Corresponding author. Tel.: +33-437911227; fax: +33-437911210. E-mail address: [email protected] (E. Daprati).
An alternative interpretation suggests that hand preference could be reflected by functional or structural asymmetries of the primary motor cortex and descending pathways, either in isolation or in association with those suggested for higher order areas. In right-handers, a comparison of the relative degree of activation between the two hemispheres following ipsilateral movements reveals greater activation in the left than in the right hemisphere [19], specifically in the region of the motor cortex [28]. Interestingly, this pattern seems to be reversed in left-handers. In a PET study, Kawashima et al. [18] found that the right premotor cortex was activated by both contralateral and ipsilateral finger movements, whereas left premotor cortex, as well as primary motor cortex and supplementary motor areas, were activated only by contralateral movements. In a recent fMRI study, Dassonville et al. [6] have shown that the volume of activation in contralateral motor cortex is greater during movements of the preferred hand in comparison to movements of the non-preferred hand in both right- and left-handed subjects. A similar observation was obtained using whole-head magnetoencephalography (MEG): Volkmann et al. [30] measured in vivo the hand representation area in right- and left-handed subjects and found a significant expansion of hand motor cortex in the hemisphere corresponding to the preferred hand. The authors
0028-3932/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 8 - 3 9 3 2 ( 0 1 ) 0 0 2 0 3 - 2
1380
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
suggested that such an enlargement could provide the neural substrate for the more efficient processing of motor output of the preferred hand. A recent study using transcranial magnetic stimulation (TMS) supports this hypothesis [29]: measurable differences related to handedness have been found in the excitability of the motor system. In particular, the threshold for eliciting motor evoked potentials (MEPs) varies according to hand dominance: lower TMS thresholds are required for eliciting MEPs in the hemisphere contralateral to the preferred hand. In other words, lower TMS thresholds for one hand are associated with greater dexterity of that hand. Although the matter of motor asymmetry and hemispheric lateralisation has been extensively debated, few studies have investigated this problem at the level of internal motor representations. A recent paper provided some evidence that the principle of contralateral control holds also for mentally simulated movements [23]. Split-brain patients were required to judge handedness of line drawings depicting either right or left hands. Accuracy in the responses was higher when the hand presented was contralateral to the perceiving hemisphere (e.g. a right hand presented to the isolated left hemisphere), whereas patients performed at chance level when the stimulus was ipsilateral to the perceiving hemisphere. Data collected in healthy subjects [16] further confirmed that an advantage exists for each hemisphere when mental simulation involves movements of the contralateral hand compared to the ipsilateral one. Normal subjects required less time to select whether an overhand or an underhand grip was more appropriate in order to grasp a manipulandum when the latter was presented to the hemisphere contralateral to the designated response hand. When a similar paradigm was administered to a right-handed commissurotomised patient [17], a dissociation between the two hemispheres emerged. The left (motor dominant) hemisphere was successful in representing with a high degree of accuracy both distal movements of the contralateral hand (i.e. grasping) and proximal movements of the ipsilateral hand (i.e. reaching). On the contrary, the right hemisphere only appeared capable of representing distal movements of the contralateral hand with modest accuracy, suggesting that processes needed to select upper limb movements may be lateralised to the left hemisphere. The presence of motor asymmetries in both real and mentally simulated movements emerges even more directly in a recent study by Maruff et al. [20]. Fifty subjects (including right- and left-handers) were required to either execute or mentally simulate a visually guided pointing task [26]. Results showed a significant speed reduction when subjects either used or imagined to use their non-dominant hand, thus supporting the hypothesis that real and imagined movements are controlled by similar constraints. However, asymmetries in both speed and error rate were greater for imagined movements as compared to executed movements. This latter result is consistent with recent neuroimaging [13,15,25] and neurophysiological [24] data suggesting that the link between execution and mental representation of movement is very tight, also extending to movement observation. A “mirror” system
has been first described in the monkey’s ventral area 6, where neurones responding to both execution and observation of a given action have been found [24]. In humans, although it has been shown that the two systems, motor execution and motor representation, only partially overlap [10], the fact that similar cortical areas are activated while representing one’s own action and observing someone else’s action implies that the same representation may be shared among individuals. Self-consciousness might rely on the possibility to discriminate between central representations activated from within (self-generated movement) from central representations activated from external agents (movement observation). Such a close link between execution and mental representation of movement raises the question whether the perception of self-performed actions can be influenced by motor asymmetries. Recent experiments suggest that this apparently simple phenomenon relies on a complex mechanism, which requires an efficient on-line control of motor output [27] as well as the ability to update the body schema in response to limb movements [5]. To our knowledge, no study has yet addressed the question whether the capacity to recognise self-produced movements is subject to the same asymmetries present in motor skills. Sirigu et al. [27] required apraxic patients suffering from left parietal lesion and age-matched normal controls to identify the origin of a moving hand in an ambiguous situation. Both the preferred and the non-preferred hand were tested. The results suggested that the ability to correctly attribute a movement to its proper agent is reduced when subjects are asked to recognise the non-preferred hand. In the present study, we further explore this issue in the attempt to assess whether the ability to recognise one’s own movement is subject to asymmetries related to motor dexterity. Toward this end, both right- and left-handed subjects are required to execute hand gestures with their preferred and non-preferred hand and to monitor their execution on a TV screen. According to the experimental condition, the image of the subjects’ hand can be substituted by that of the examiner’s, thus creating an ambiguous environment. The subjects’ goal is to decide whether the presented hand belongs to themselves or not. Indeed, if the ability to recognise one’s own movement is subject to the same asymmetries present in motor skills, recognition of the non-preferred hand should lead to an increased number of errors and/or to longer decision time in both right- and left-handers.
2. Methods 2.1. Subjects Eighteen subjects participated to the experiment: nine of them (three men and six women, mean age 23.1 years, range 21–27 years) were right-handers and nine were left-handers (two men and seven women, mean age 23.1 years, range 18–38 years). Mean laterality coefficient, as assessed by the
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
1381
Fig. 1. Schematic representation of the apparatus: details in the text.
Edinburgh Inventory [22], was +0.87 for the right-handers and −0.75 for the left-handers. None of the subjects suffered from neurological or psychiatric pathologies. Vision was normal or corrected to normal. Subjects were naive as to the purpose of the experiment. 2.2. Apparatus A detailed description of the apparatus is given elsewhere [4]. Fig. 1 shows a schematic representation of the system: subjects sat in front of a table on which a mirror was placed (M), at about 35 cm from the subject’s frontal plane and inclined by 30◦ on the vertical plane. Subjects positioned their hand below the mirror, on a comfortable support. A camera (C1) filmed the subjects’ hand and a closed circuit television system reproduced it in order to make it visible to the subjects on the mirror surface. Since no time delay was present, when the hand was placed below the mirror the impression was that of seeing it through a window glass. An identical set-up was located nearby and collected images of the experimenter’s hand (C2). During the experiment, either the image of the subject’s hand or that of the experimenter’s could be presented to the subject. Identical white gloves covered both hands, in order to eliminate morphological differences. In order to get a correct spatial and temporal match between the subject’s and the experimenter’s hand, three TV screens were available in the experimenter’s cabin. The monitors presented the images of both the subject’s and the experimenter’s hand, either individually or superimposed. In order to synchronise movement onset, both subject and experimenter started to move in response to one acoustic signal. All detectable differences in movement onset were controlled and trials in which timing or spatial matching was imprecise were repeated at the end of the session. Two experimenters were sitting in the cabin: alternately, one acted as mime, while the other monitored task performance. One experimenter was strongly right-handed
(laterality coefficient 1.00), the other was almost ambidextrous (laterality coefficient 0.30). 2.3. Procedure Before the experimental session, subjects were shown the gestures they were to perform. Subjects started from the fist-closed posture and moved their hand on the horizontal plane, palm facing the table. Gestures required extension of one or more digits, namely (1) thumb, (2) index, (3) thumb and index, (4) all fingers. The task for the subjects was to produce one gesture in response to a verbal command and to monitor its execution on the screen in front of them. Subjects knew in advance that in half of the trials the hand displayed would be their own one (subject condition), whereas in the other half, it would be that of another person (experimenter condition). In the latter condition, the examiner’s hand always performed the same movement that was required to the subject. Subjects knew in advance that the two conditions would be presented in random order. Indeed, the subjects’ goal was to recognise whether the hand on the screen was their own or not. More details on the procedure are given in Fig. 2. Twenty-four trials were run for both the subject condition and the experimenter condition (six trials for each gesture type). For each subject, both the right and the left hand were tested in separate blocks: order of hand use was counterbalanced among subjects. Verbal responses were recorded: the trial was considered successful when subjects responded “yes” to the image of their own hand and “no” to that of the experimenter’s hand. A voice key system recorded response times. Response times shorter than 100 ms or longer than 4000 ms were considered errors and the corresponding trials were repeated at the end of the session. Due to the dichotomous nature of the analysed variable, the number of correct responses was submitted to non-parameteric tests: namely Mann–Whitney U-test for between groups comparisons and
1382
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
Fig. 2. Experimental procedure. From top to bottom: time = time-scale, screen = events appearing on the screen located in front of the subject, task = instructions given to subjects. “Instruction” refers to the verbal command specifying the type of gesture subjects were required to perform. “Movement” refers to the motor phase, elicited by an acoustic signal (beep). Subjects were allowed an interval of 2 s in order to perform the gesture before vision of the hand was prevented. “Response” refers to the period of 4 s in which subjects were expected to give their answer.
Wilcoxon matched-pairs-tests for within groups comparisons. Response times were analysed by means of three-way ANOVA, group (right-handers versus left-handers), manual preference (preferred versus non-preferred hand) and experimental condition (subject’s versus experimenter’s hand) being the main factors. Newman–Keuls-test was used for post-hoc analysis on response times. Whenever required, a Bonferroni correction for multiple comparisons was applied.
3. Results All subjects well understood the task and errors in the procedure, i.e. execution of a movement different from that requested, anticipation or delay in starting the movement or giving the response were extremely rare (0.31% of trials) and were randomly distributed among conditions. 3.1. Correct responses Independently from their handedness, the 18 subjects produced significantly more correct responses when asked to recognise their preferred hand with respect to the non-preferred hand (correct responses: preferred 38.5, S.D. 3.9, non-preferred 36.2, S.D. 4.0; T = 20.0, Z = 2.85,
P = 0.004). This was true for both the condition in which subjects saw their own hand (subject condition: T = 10.5, Z = 2.4, P = 0.01) and for the condition in which they saw the experimenter’s hand (experimenter condition: T = 27.5, Z = 2.3, P = 0.02, see Table 1 for numerical values). All subjects produced more correct responses in the subject condition, i.e. when they saw their own hand (subject condition 45.0, S.D. 2.2, experimenter condition 29.7, S.D. 7.7; T = 0.0000001, Z = 3.62, P < 0.0003). In other words, errors increased whenever ambiguity was introduced. A closer look to the number of correct responses given by each group of subjects showed that the advantage for the preferred hand was significant only within the right-handers group (correct responses: preferred 39.2, S.D. 3.3, non-preferred 36.7, S.D. 3.6; T = 0.00001, Z = 2.665, P < 0.01). The difference was not significant in the case of left-handers (preferred 37.8, S.D. 4.5, non-preferred 35.7, S.D. 4.5; P = 0.1, Fig. 3 left panel). A direct comparison between the two groups failed to reach significance. As previously stated, all subjects produced more correct responses when they saw their own hand compared to when they saw the experimenter’s hand. The decrease in the number of correct responses in the latter condition was significant for both right-handed subjects (experimenter condition 31.0, S.D. 7.4, subject condition 44.9, S.D. 2.3; T = 0.00001, Z = 2.520, P < 0.01) and left-handed subjects
Table 1 Mean number of correct responses (S.D.) and mean response times (S.D.) for right- and left-handed subjects, and for the entire group Preferred hand Subject’s hand Right-handers Left-handers All subjects Right-handers Left-handers All subjects
Mean 23.2 22.9 23.0 Mean 630.4 676.5 653.4
Non-Preferred hand Experimenter’s hand
Subject’s hand
number of correct responses (S.D.) (maximum of correct responses = 24) (0.8) 16.0 (3.7) 21.7 (1.6) (0.9) 14.9 (4.4) 22.2 (1.9) (0.9) 15.4 (4.0) 21.9 (1.7) response times (S.D.) in ms (153.1) 609.0 (160.1) 607.4 (211.9) (264.1) 742.1 (301.7) 747.5 (192.8) (210.7) 675.5 (244.1) 677.5 (209.4)
Experimenter’s hand 15.0 (3.8) 13.4 (5.0) 14.2 (4.4) 608.2 (183.5) 802.9 (224.9) 705.5 (222.9)
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
1383
Fig. 3. Effect of manual preference on number of correct responses (left panel) and on response times (right panel) given by each group of subjects. Preferred hand (black bars) refers to the right hand for right-handed subjects and to the left hand for left-handed subjects. The opposite for non-preferred hand (dotted bars). Error bars indicate standard errors (preferred hand: continuous line, non-preferred hand: dotted line).
(experimenter condition 28.3, S.D. 8.3, subject condition 45.1, S.D. 2.2; T = 0.0001, Z = 2.665, P < 0.01). 3.2. Response times When response times are considered, it clearly emerges from Fig. 3 (right panel) that left-handed subjects were on average slower than right-handed subjects (left-handers 742.2 ms, S.D. 219.9; right-handers 613.8 ms, S.D. 141.7). Although no main effect of group was found, a significant interaction emerged between group and experimental condition (F(1, 16) = 4.515, P < 0.0495). Post-hoc analysis showed that left-handers were slower than right-handers in both the subject and the experimenter conditions (subject condition, right-handers: 618.9 ms, S.D. 150.4; left-handers: 712.0 ms, S.D. 198.7, P < 0.001; experimenter condition, right-handers: 608.6 ms, S.D. 135.4; left-handers: 772.5 ms, S.D. 247.9, P < 0.0002). Moreover, a consistent time increase between the “easier” subject condition (712.0 ms, S.D. 198.7) and the more ambiguous experimenter condition (772.5 ms, S.D. 247.9) was present only in the left-handers group. No difference in mean response times was found among right-handers (Fig. 3 right panel, Table 1). Further analyses were run in order to test any evidence for a speed–accuracy trade-off in subjects’ performance. Response times for correct responses were compared to response times for incorrect responses by means of a four-way ANOVA. Group (right-handers versus left-handers) was the between groups factor whereas manual preference
(preferred versus non-preferred hand), experimental condition (subjects versus experimenter’s hand) and response accuracy (correct versus incorrect answer) were within groups factors. No difference between the two groups was found. A main effect of response accuracy emerged (F(1, 7) = 10.18, P < 0.015), showing that average response times were significantly slower when they led to incorrect (858.4 ms, S.D. 189.4) compared to correct responses (675.4 ms, S.D. 189.4). The interaction between experimental condition and accuracy was also significant (F(1, 7) = 5.71, P < 0.0482): namely, response latency significantly increased (926.3 ms, S.D. 258.4) when subjects saw their own hand but were mistaken in recognising it (i.e. they responded “no”). Summing up, the main results of the present study were the following: (1) independently from their handedness, all subjects provided more correct responses when they were asked to recognise their preferred hand. (2) All subjects produced more correct responses when they saw their own hand compared to when they saw the experimenter’s hand, suggesting that errors increased when ambiguity was introduced. (3) Left-handers were on average slower in providing their response than were right-handed subjects. (4) Longer response latencies were not associated to correct responses. 4. Discussion Most people have a preferred hand, whose skilfulness largely exceeds that of the so-called non-preferred hand. Asymmetries related to motor dominance have been de-
1384
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
scribed to affect imagined movements [20]. The present study was designed in order to assess whether this notion extends to the capacity to recognise self-produced movements, or in broader terms, to self-awareness in motor acts. Right and left-handed healthy subjects were required to produce a hand gesture in an ambiguous environment, in which direct vision of their moving limb was prevented and visual feedback provided on a TV screen could be manipulated. Subjects were asked to decide whether they were the agents of the presented movements. Results showed that the number of correct responses given by the subjects varied according to the hand used. Both right- and left-handers were more accurate when asked to identify their preferred hand with respect to their non-preferred hand, suggesting the presence of a close link between motor dominance and motor awareness. We will define here motor awareness as the ability of an individual to both perceive and visually recognise a given movement as his/her own. In other words, as the knowledge enabling subjects to state that they are the source and the owner of a presented gesture. This ability necessarily comes into play in the task used in the present experiment. As previously reported [27], the accomplishment of the present task likely requires several steps. First, in addition to intentional information, subjects must be able to form a reliable representation of the movement they are about to execute. More precisely, subjects should be able to predict the proprioceptive and visual consequences of their act and to fit actual incoming kinaesthetic information into their predictive model. As a result, they will “know” which movement they are executing and whether it matches their intentions. Secondly, this information is compared to the one provided by the visual system, thus making possible to decide whether the movement shown to the subjects belongs to themselves or not. Once the result of this comparison is made conscious, a response can be produced. We recently provided some evidence that brain damage can disrupt this motor recognition mechanism at the level of action representation, thus preventing subjects from having a reliable image of the movement they are executing. More particularly, in apraxic patients suffering from left parietal lesion [27], the ability to either form or update a reliable predictive model of the executed gestures is impaired. Accordingly, when attempting to reproduce a complex hand posture, these patients fail in their execution as well as in recognising their own movement. The present results on healthy subjects support the hypothesis that the ability to visually recognise a movement as one’s own is subject to asymmetries similar to those present in motor skills. It is unlikely that our results may be accounted for by the possible time delay intervening between movement executed and movement observed by the subjects. Indeed, had this been the cue to the correct answer, it should have produced an equal number of correct responses for the preferred and non-preferred hand. The increase in the number of correct responses after presentation of the dominant hand could be accounted for by at
least two non-conflicting interpretations. Either subjects are more trained to attend to their preferred hand during praxis, or self-awareness of motor outputs is richer for the more skilled hand. According to the first interpretation, we can speculate that the better performance for the dominant hand could be a secondary effect of familiarity, subjects having acquired major expertise in visually controlling movements of their preferred hand. On the other hand, it could not be excluded that, when attempting to recognise their own hand, subjects made use of an internal representation of their dominant hand that is “motor” in nature. Evidence has been provided that motor habits can influence our perception of the visual world. Objects that can be grasped and manipulated appear to be processed also in relation to the manual dominance of the subject, thus suggesting the involvement of motor processes (namely motor imagery) in non-motor functions [7]. In the present task, subjects could have found recognition of the preferred hand easier either due to a more familiar visual image or to a richer visual–motor knowledge of their more skilled hand. We favour this second hypothesis, which points to a close relation between the system responsible for motor consciousness and the structures involved in motor representation and movement production. What kinds of mechanisms allow the interplay between consciousness of movement and motor production? Manual preference has an anatomic and functional counterpart in the expansion of hand motor cortex for the preferred hand [1,30] and this phenomenon has been invoked to explain the more efficient processing of the corresponding motor output. Efficiency could improve thanks to an increase in the number of neural elements and/or tighter connectivity: monkey data show both an enlargement and an increased degree of spatial complexity of motor areas opposite to the preferred hand [21]. At the cortical level, this modification may account for better interaction between neuronal assemblies representing muscles that are used in close spatial or temporal contiguity. Analysis of cytoarchitectonic differences between the two hemispheres [1] have shown an increased neuropil compartment in BA 4 in the dominant hemisphere, supporting the hypothesis that hand preference is associated with increased connectivity. At the behavioural level, such improved interaction may be responsible for a more skilled motor repertoire of the preferred hand. In agreement with the above data, the present results further indicate that not only motor production but also motor awareness can benefit from the anatomical and physiological pattern linked to hand preference. Handedness asymmetries seem, thus, to extend beyond the level of primary motor functions and likely emerge also in those structures which are involved in cognitive–motor operations. Two more points emerged from the results of the present experiment. First, the difference between preferred and non-preferred hand was less evident in the left-handers group. Second, left-handers were on average slower in giving their responses than right-handers. The former result could be accounted for by the fact that left-handers
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
are more heterogeneous and less markedly lateralised than right-handers: for instance, on preference measures, i.e. hand or foot usage, left-handers show larger variability and lower lateralisation than right-handed subjects [2]. In addition, in hand selection responses to unimanual tasks, left-handers do not respond with the preferred hand as consistently as do right-handers [8]. This reduced lateralisation emerged also in the sample examined in the present experiment: only one of our left-handed subjects scored between −0.90 and −1.00 at the Edinburgh Inventory, whereas five out of nine right-handed subjects showed such an extreme score. Indeed the absolute scores at the handedness test differed significantly between the two groups (U = 18.00, Z = 1.987, P = 0.046) Accordingly, we can assume that in our sample the reduced asymmetry in hand motor skills present in the left-handers group could be responsible for the reduced handedness effect. Both results are consistent with a previous report by Gentilucci et al. [9] in which right- and left-handed subjects were required to recognise whether a static hand holding an object, with either a congruent or incongruent type of grip, corresponded to the picture of a right or a left hand. Right-handers were faster in judging right hands, whereas no difference was found between preferred and non-preferred hand in the left-handed subjects. Moreover, left-handers were on average slower than right-handers. A similar pattern is also found in another previous report. When asked to decide whether a rotating screwdriver was either screwing or unscrewing [7], healthy subjects required longer response times in those conditions implying motor imagery of the non-dominant hand. Moreover, left-handers were overall slower than right-handers. As suggested by Gentilucci et al. [9], the behaviour of left-handers could be explained by the fact that, when required to covertly represent movement of their own hands, they likely rely on a “pictorial” (i.e. more observation-based) rather than “pragmatic” (i.e. motor experience-based) image of their dominant hand. In agreement with this hypothesis, we can assume here that, on the one hand, left-handers are less affected by asymmetries related to motor dexterity. On the other hand, they likely need further steps in the analysis of the presented image, such as a transition between a visual and a motor-based representation, in order to access those subtle kinematics cues which are required to recognise one’s own movement. In conclusion, we think that the present data demonstrate how the ability to recognise self-generated movements can be affected by motor dominance, suggesting that the system responsible for motor awareness is profoundly nested within the mechanisms controlling motor production.
Acknowledgements The authors wish to thank Virginie Czernecki for her precious help in collecting the data. This work was supported by CNRS.
1385
References [1] Amunts K, Schlaug G, Steinmetz H, Dabringhaus A, Roland PE, Zilles K. Asymmetry in the human motor cortex and handedness. Neuroimage 1996;4:216–22. [2] Borod JC, Caron HS, Koff E. Left-handers and right-handers compared on performance and preference measures of lateral dominance. British Journal of Psychology 1984;75:177–86. [3] Coren S, Porac C. Fifty centuries of right-handedness: the historical record. Science 1977;198:631–2. [4] Daprati E, Franck N, Georgieff N, Proust J, Pacherie E, Dalery J, et al. Looking for the agent. An investigation into consciousness of action and self-consciousness in schizophrenic patients. Cognition 1997;65:71–86. [5] Daprati E, Sirigu A, Pradat-Dihel P, Franck N, Jeannerod M. Recognition of self-produced movement in a case of severe neglect. Neurocase 2000;6:477–86. [6] Dassonville P, Zhu XH, Ugurbil K, Kim SG, Ashe J. Functional activation in motor cortex reflects the direction and the degree of handedness. Proceedings of the National Academy of the Sciences USA 1997;94:14015–8. [7] de’Sperati C, Stucchi N. Recognizing the motion of a graspable object is guided by handedness. Neuroreport 1997;8:2761–5. [8] Gabbard C, Iteya M, Rabb C. A lateralized comparison of handedness and object proximity. Canadian Journal of Experimental Psychology 1997;51:176–80. [9] Gentilucci M, Daprati E, Gangitano M. Right-handers and left-handers have different representations of their own hand. Cognitive Brain Research 1998;6:185–92. [10] Gerardin E, Sirigu A, Lehericy S, Poline JB, Gaymard B, Marsault C, et al. Partially overlapping neural networks for real and imagined hand movements. Cerebral Cortex 2000;10:1093–4. [11] Gilbert AN, Wysocki CS. Hand preference and age in the United States. Neuropsychologia 1992;30:601–8. [12] Gonzales-Rothi LJ, Heilman KM, editors. Apraxia: the neuropsychology of action. New York: Lawrence Erlbaum, 1996. [13] Grezes J, Costes N, Decety J. The effects of learning and intention on the neural network involved in the perception of meaningless actions. Brain 1999;122:1875–87. [14] Haaland KY, Harrington DL, Clinical implications of ipsilateral motor deficits after unilateral hemispheric damage. In: Bruno A, Chollet F, Vellas BJ, Albarede JL, editors. Facts and research in gerontology: stroke in the elderly. New York: Springer, 1996: pp. 101–14. [15] Iacoboni M, Woods RP, Brass M, Bekkering H, Mazziotta JC, Rizzolatti G. Cortical mechanisms of human imitation. Science 1999;286:2526–8. [16] Johnson SH. Cerebral organization of motor imagery: Contralateral control of grip selection in mentally represented prehension. Psychological Science 1998;9:219–22. [17] Johnson SH, Corballis PM, Gazzaniga MS. Within grasp but out of reach: Evidence for a double dissociation between imagined hand and arm movements in the left cerebral hemisphere. Neuropsychologia 2001;39:36–50. [18] Kawashima R, Inoue K, Sato K, Fukuda H. Functional asymmetry of cortical motor control in left-handed subjects. Neuroreport 1997;8:1729–32. [19] Kim SG, Ashe J, Hendrich K, Ellerman JM, Merkle H, Ugurbil K, et al. Functional magnetic resonance imaging of motor cortex: hemispheric asymmetry and handedness. Science 1993;261: 615–7. [20] Maruff P, Wilson PH, De Fazio J, Cerritelli B, Hedt A, Currie J. Asymmetries between dominant and non-dominant hands in real and imagined motor task performance. Neuropsychologia 1999;37:379– 84.
1386
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
[21] Nudo RJ, Jenkins WM, Merzenich MM, Prejean T, Grenda R. Neurophysiological correlates of hand preference in primary motor cortex of adult squirrel monkeys. Journal of Neuroscience 1992;12:2918–47. [22] Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 1971;9:97–113. [23] Parsons LM, Gabrieli JD, Phelps EA, Gazzaniga MS. Cerebrally lateralized mental representations of hand shape and movement. Journal of Neuroscience 1998;18:6539–48. [24] Rizzolatti G, Fadiga L, Gallese V, Fogassi L. Premotor cortex and the recognition of motor actions. Cognitive Brain Research 1996;3:131– 41. [25] Rizzolatti G, Fadiga L, Matelli M, Bettinardi V, Paulesu E, Perani D, et al. Localization of grasp representations in humans by PET: 1. Observation versus execution. Experimental Brain Research 1996;111:246–52.
[26] Sirigu A, Duhamel JR, Cohen L, Pillon B, Dubois B, Agid Y. The mental representation of hand movements after parietal cortex damage. Science 1996;273:1564–7. [27] Sirigu A, Daprati E, Pradat-Dihel P, Franck N, Jeannerod M. Perception of self-generated movement following left parietal lesion. Brain 1999;122:1867–74. [28] Taniguchi M, Yoshimine T, Cheyne D, Kato A, Kihara T, Ninomiya H, et al. Neuromagnetic fields preceding unilateral movements in dextrals and sinistrals. Neuroreport 1998;9:1497–502. [29] Triggs WJ, Calvanio R, Levine M. Transcranial magnetic stimulation reveals a hemispheric asymmetry correlate of intermanual differences in motor performance. Neuropsychologia 1997;35: 1355–63. [30] Volkmann J, Schnitzler A, Witte OW, Freund HJ. Handedness and asymmetry of hand representation in human motor cortex. Journal of Neurophysiology 1998;79:2149–54.
Laterality effects on motor awareness Elena Daprati∗ , Angela Sirigu Institut des Sciences Cognitives, 67 Bd Pinel, 69675 Bron, France Received 26 January 2000; accepted 28 September 2001
Abstract The ability to perceive self-produced movements and to correctly attribute an action to its proper agent is a natural task and a basic requirement to human social communication. Recent experiments suggest that this apparently simple phenomenon is related to the mechanisms of motor production, raising the question whether recognition of self-produced movement is affected by asymmetries similar to those present in motor skills. In this study, right- and left-handed subjects decided whether a moving hand presented on a screen was the image of their own hand or of that of another person. Two experimental conditions were tested: either subjects saw their own moving hand (subject condition) or they were shown the experimenter’s hand pantomiming their movement (experimenter condition); a glove masked morphological differences between the two hands. Verbal responses and response times were analysed. Results showed that all subjects were more accurate in recognising their preferred hand with respect to their non-preferred hand. Right-handers responded faster than left-handers did, the latter group being especially slowed down in the experimenter condition. On the contrary, in the right-handers group, response times did not differ among conditions. The present data indicate that the ability to recognise self-generated movements is affected by motor dominance, thus suggesting that conscious knowledge of self-produced movements is closely related to the motor system. © 2002 Elsevier Science Ltd. All rights reserved. Keywords: Motor consciousness; Handedness; Motor skills; Manual asymmetry
1. Introduction Over 90% of humans are right-handed, that is they show a clear preference for the use of the right hand in the execution of skilled movements. Although this notion is well-established and seems to be rather consistent across cultures and over time [3,11], the neurobiological correlate of handedness is still a matter of debate. Neuropsychological evidence suggests that this behavioural lateralisation can be related to an asymmetry of the cortical areas controlling the cognitive–motor requirements of skilled movements [14]. As a general rule, movement of one limb is controlled by the motor cortex of the contralateral hemisphere: however, at least in right-handers, a dominant role of the left hemisphere for praxis functions has frequently been invoked. For instance, ipsilateral deficits in tasks involving complex cognitive–motor abilities have been described only following left (but not right) hemisphere damage [14]. Moreover, ideomotor apraxia—i.e. a specific impairment in the performance of complex gestures in the absence of elementary sensory/motor deficits—is known to occur more frequently after left than after right hemisphere damage [12]. ∗
Corresponding author. Tel.: +33-437911227; fax: +33-437911210. E-mail address: [email protected] (E. Daprati).
An alternative interpretation suggests that hand preference could be reflected by functional or structural asymmetries of the primary motor cortex and descending pathways, either in isolation or in association with those suggested for higher order areas. In right-handers, a comparison of the relative degree of activation between the two hemispheres following ipsilateral movements reveals greater activation in the left than in the right hemisphere [19], specifically in the region of the motor cortex [28]. Interestingly, this pattern seems to be reversed in left-handers. In a PET study, Kawashima et al. [18] found that the right premotor cortex was activated by both contralateral and ipsilateral finger movements, whereas left premotor cortex, as well as primary motor cortex and supplementary motor areas, were activated only by contralateral movements. In a recent fMRI study, Dassonville et al. [6] have shown that the volume of activation in contralateral motor cortex is greater during movements of the preferred hand in comparison to movements of the non-preferred hand in both right- and left-handed subjects. A similar observation was obtained using whole-head magnetoencephalography (MEG): Volkmann et al. [30] measured in vivo the hand representation area in right- and left-handed subjects and found a significant expansion of hand motor cortex in the hemisphere corresponding to the preferred hand. The authors
0028-3932/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 8 - 3 9 3 2 ( 0 1 ) 0 0 2 0 3 - 2
1380
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
suggested that such an enlargement could provide the neural substrate for the more efficient processing of motor output of the preferred hand. A recent study using transcranial magnetic stimulation (TMS) supports this hypothesis [29]: measurable differences related to handedness have been found in the excitability of the motor system. In particular, the threshold for eliciting motor evoked potentials (MEPs) varies according to hand dominance: lower TMS thresholds are required for eliciting MEPs in the hemisphere contralateral to the preferred hand. In other words, lower TMS thresholds for one hand are associated with greater dexterity of that hand. Although the matter of motor asymmetry and hemispheric lateralisation has been extensively debated, few studies have investigated this problem at the level of internal motor representations. A recent paper provided some evidence that the principle of contralateral control holds also for mentally simulated movements [23]. Split-brain patients were required to judge handedness of line drawings depicting either right or left hands. Accuracy in the responses was higher when the hand presented was contralateral to the perceiving hemisphere (e.g. a right hand presented to the isolated left hemisphere), whereas patients performed at chance level when the stimulus was ipsilateral to the perceiving hemisphere. Data collected in healthy subjects [16] further confirmed that an advantage exists for each hemisphere when mental simulation involves movements of the contralateral hand compared to the ipsilateral one. Normal subjects required less time to select whether an overhand or an underhand grip was more appropriate in order to grasp a manipulandum when the latter was presented to the hemisphere contralateral to the designated response hand. When a similar paradigm was administered to a right-handed commissurotomised patient [17], a dissociation between the two hemispheres emerged. The left (motor dominant) hemisphere was successful in representing with a high degree of accuracy both distal movements of the contralateral hand (i.e. grasping) and proximal movements of the ipsilateral hand (i.e. reaching). On the contrary, the right hemisphere only appeared capable of representing distal movements of the contralateral hand with modest accuracy, suggesting that processes needed to select upper limb movements may be lateralised to the left hemisphere. The presence of motor asymmetries in both real and mentally simulated movements emerges even more directly in a recent study by Maruff et al. [20]. Fifty subjects (including right- and left-handers) were required to either execute or mentally simulate a visually guided pointing task [26]. Results showed a significant speed reduction when subjects either used or imagined to use their non-dominant hand, thus supporting the hypothesis that real and imagined movements are controlled by similar constraints. However, asymmetries in both speed and error rate were greater for imagined movements as compared to executed movements. This latter result is consistent with recent neuroimaging [13,15,25] and neurophysiological [24] data suggesting that the link between execution and mental representation of movement is very tight, also extending to movement observation. A “mirror” system
has been first described in the monkey’s ventral area 6, where neurones responding to both execution and observation of a given action have been found [24]. In humans, although it has been shown that the two systems, motor execution and motor representation, only partially overlap [10], the fact that similar cortical areas are activated while representing one’s own action and observing someone else’s action implies that the same representation may be shared among individuals. Self-consciousness might rely on the possibility to discriminate between central representations activated from within (self-generated movement) from central representations activated from external agents (movement observation). Such a close link between execution and mental representation of movement raises the question whether the perception of self-performed actions can be influenced by motor asymmetries. Recent experiments suggest that this apparently simple phenomenon relies on a complex mechanism, which requires an efficient on-line control of motor output [27] as well as the ability to update the body schema in response to limb movements [5]. To our knowledge, no study has yet addressed the question whether the capacity to recognise self-produced movements is subject to the same asymmetries present in motor skills. Sirigu et al. [27] required apraxic patients suffering from left parietal lesion and age-matched normal controls to identify the origin of a moving hand in an ambiguous situation. Both the preferred and the non-preferred hand were tested. The results suggested that the ability to correctly attribute a movement to its proper agent is reduced when subjects are asked to recognise the non-preferred hand. In the present study, we further explore this issue in the attempt to assess whether the ability to recognise one’s own movement is subject to asymmetries related to motor dexterity. Toward this end, both right- and left-handed subjects are required to execute hand gestures with their preferred and non-preferred hand and to monitor their execution on a TV screen. According to the experimental condition, the image of the subjects’ hand can be substituted by that of the examiner’s, thus creating an ambiguous environment. The subjects’ goal is to decide whether the presented hand belongs to themselves or not. Indeed, if the ability to recognise one’s own movement is subject to the same asymmetries present in motor skills, recognition of the non-preferred hand should lead to an increased number of errors and/or to longer decision time in both right- and left-handers.
2. Methods 2.1. Subjects Eighteen subjects participated to the experiment: nine of them (three men and six women, mean age 23.1 years, range 21–27 years) were right-handers and nine were left-handers (two men and seven women, mean age 23.1 years, range 18–38 years). Mean laterality coefficient, as assessed by the
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
1381
Fig. 1. Schematic representation of the apparatus: details in the text.
Edinburgh Inventory [22], was +0.87 for the right-handers and −0.75 for the left-handers. None of the subjects suffered from neurological or psychiatric pathologies. Vision was normal or corrected to normal. Subjects were naive as to the purpose of the experiment. 2.2. Apparatus A detailed description of the apparatus is given elsewhere [4]. Fig. 1 shows a schematic representation of the system: subjects sat in front of a table on which a mirror was placed (M), at about 35 cm from the subject’s frontal plane and inclined by 30◦ on the vertical plane. Subjects positioned their hand below the mirror, on a comfortable support. A camera (C1) filmed the subjects’ hand and a closed circuit television system reproduced it in order to make it visible to the subjects on the mirror surface. Since no time delay was present, when the hand was placed below the mirror the impression was that of seeing it through a window glass. An identical set-up was located nearby and collected images of the experimenter’s hand (C2). During the experiment, either the image of the subject’s hand or that of the experimenter’s could be presented to the subject. Identical white gloves covered both hands, in order to eliminate morphological differences. In order to get a correct spatial and temporal match between the subject’s and the experimenter’s hand, three TV screens were available in the experimenter’s cabin. The monitors presented the images of both the subject’s and the experimenter’s hand, either individually or superimposed. In order to synchronise movement onset, both subject and experimenter started to move in response to one acoustic signal. All detectable differences in movement onset were controlled and trials in which timing or spatial matching was imprecise were repeated at the end of the session. Two experimenters were sitting in the cabin: alternately, one acted as mime, while the other monitored task performance. One experimenter was strongly right-handed
(laterality coefficient 1.00), the other was almost ambidextrous (laterality coefficient 0.30). 2.3. Procedure Before the experimental session, subjects were shown the gestures they were to perform. Subjects started from the fist-closed posture and moved their hand on the horizontal plane, palm facing the table. Gestures required extension of one or more digits, namely (1) thumb, (2) index, (3) thumb and index, (4) all fingers. The task for the subjects was to produce one gesture in response to a verbal command and to monitor its execution on the screen in front of them. Subjects knew in advance that in half of the trials the hand displayed would be their own one (subject condition), whereas in the other half, it would be that of another person (experimenter condition). In the latter condition, the examiner’s hand always performed the same movement that was required to the subject. Subjects knew in advance that the two conditions would be presented in random order. Indeed, the subjects’ goal was to recognise whether the hand on the screen was their own or not. More details on the procedure are given in Fig. 2. Twenty-four trials were run for both the subject condition and the experimenter condition (six trials for each gesture type). For each subject, both the right and the left hand were tested in separate blocks: order of hand use was counterbalanced among subjects. Verbal responses were recorded: the trial was considered successful when subjects responded “yes” to the image of their own hand and “no” to that of the experimenter’s hand. A voice key system recorded response times. Response times shorter than 100 ms or longer than 4000 ms were considered errors and the corresponding trials were repeated at the end of the session. Due to the dichotomous nature of the analysed variable, the number of correct responses was submitted to non-parameteric tests: namely Mann–Whitney U-test for between groups comparisons and
1382
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
Fig. 2. Experimental procedure. From top to bottom: time = time-scale, screen = events appearing on the screen located in front of the subject, task = instructions given to subjects. “Instruction” refers to the verbal command specifying the type of gesture subjects were required to perform. “Movement” refers to the motor phase, elicited by an acoustic signal (beep). Subjects were allowed an interval of 2 s in order to perform the gesture before vision of the hand was prevented. “Response” refers to the period of 4 s in which subjects were expected to give their answer.
Wilcoxon matched-pairs-tests for within groups comparisons. Response times were analysed by means of three-way ANOVA, group (right-handers versus left-handers), manual preference (preferred versus non-preferred hand) and experimental condition (subject’s versus experimenter’s hand) being the main factors. Newman–Keuls-test was used for post-hoc analysis on response times. Whenever required, a Bonferroni correction for multiple comparisons was applied.
3. Results All subjects well understood the task and errors in the procedure, i.e. execution of a movement different from that requested, anticipation or delay in starting the movement or giving the response were extremely rare (0.31% of trials) and were randomly distributed among conditions. 3.1. Correct responses Independently from their handedness, the 18 subjects produced significantly more correct responses when asked to recognise their preferred hand with respect to the non-preferred hand (correct responses: preferred 38.5, S.D. 3.9, non-preferred 36.2, S.D. 4.0; T = 20.0, Z = 2.85,
P = 0.004). This was true for both the condition in which subjects saw their own hand (subject condition: T = 10.5, Z = 2.4, P = 0.01) and for the condition in which they saw the experimenter’s hand (experimenter condition: T = 27.5, Z = 2.3, P = 0.02, see Table 1 for numerical values). All subjects produced more correct responses in the subject condition, i.e. when they saw their own hand (subject condition 45.0, S.D. 2.2, experimenter condition 29.7, S.D. 7.7; T = 0.0000001, Z = 3.62, P < 0.0003). In other words, errors increased whenever ambiguity was introduced. A closer look to the number of correct responses given by each group of subjects showed that the advantage for the preferred hand was significant only within the right-handers group (correct responses: preferred 39.2, S.D. 3.3, non-preferred 36.7, S.D. 3.6; T = 0.00001, Z = 2.665, P < 0.01). The difference was not significant in the case of left-handers (preferred 37.8, S.D. 4.5, non-preferred 35.7, S.D. 4.5; P = 0.1, Fig. 3 left panel). A direct comparison between the two groups failed to reach significance. As previously stated, all subjects produced more correct responses when they saw their own hand compared to when they saw the experimenter’s hand. The decrease in the number of correct responses in the latter condition was significant for both right-handed subjects (experimenter condition 31.0, S.D. 7.4, subject condition 44.9, S.D. 2.3; T = 0.00001, Z = 2.520, P < 0.01) and left-handed subjects
Table 1 Mean number of correct responses (S.D.) and mean response times (S.D.) for right- and left-handed subjects, and for the entire group Preferred hand Subject’s hand Right-handers Left-handers All subjects Right-handers Left-handers All subjects
Mean 23.2 22.9 23.0 Mean 630.4 676.5 653.4
Non-Preferred hand Experimenter’s hand
Subject’s hand
number of correct responses (S.D.) (maximum of correct responses = 24) (0.8) 16.0 (3.7) 21.7 (1.6) (0.9) 14.9 (4.4) 22.2 (1.9) (0.9) 15.4 (4.0) 21.9 (1.7) response times (S.D.) in ms (153.1) 609.0 (160.1) 607.4 (211.9) (264.1) 742.1 (301.7) 747.5 (192.8) (210.7) 675.5 (244.1) 677.5 (209.4)
Experimenter’s hand 15.0 (3.8) 13.4 (5.0) 14.2 (4.4) 608.2 (183.5) 802.9 (224.9) 705.5 (222.9)
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
1383
Fig. 3. Effect of manual preference on number of correct responses (left panel) and on response times (right panel) given by each group of subjects. Preferred hand (black bars) refers to the right hand for right-handed subjects and to the left hand for left-handed subjects. The opposite for non-preferred hand (dotted bars). Error bars indicate standard errors (preferred hand: continuous line, non-preferred hand: dotted line).
(experimenter condition 28.3, S.D. 8.3, subject condition 45.1, S.D. 2.2; T = 0.0001, Z = 2.665, P < 0.01). 3.2. Response times When response times are considered, it clearly emerges from Fig. 3 (right panel) that left-handed subjects were on average slower than right-handed subjects (left-handers 742.2 ms, S.D. 219.9; right-handers 613.8 ms, S.D. 141.7). Although no main effect of group was found, a significant interaction emerged between group and experimental condition (F(1, 16) = 4.515, P < 0.0495). Post-hoc analysis showed that left-handers were slower than right-handers in both the subject and the experimenter conditions (subject condition, right-handers: 618.9 ms, S.D. 150.4; left-handers: 712.0 ms, S.D. 198.7, P < 0.001; experimenter condition, right-handers: 608.6 ms, S.D. 135.4; left-handers: 772.5 ms, S.D. 247.9, P < 0.0002). Moreover, a consistent time increase between the “easier” subject condition (712.0 ms, S.D. 198.7) and the more ambiguous experimenter condition (772.5 ms, S.D. 247.9) was present only in the left-handers group. No difference in mean response times was found among right-handers (Fig. 3 right panel, Table 1). Further analyses were run in order to test any evidence for a speed–accuracy trade-off in subjects’ performance. Response times for correct responses were compared to response times for incorrect responses by means of a four-way ANOVA. Group (right-handers versus left-handers) was the between groups factor whereas manual preference
(preferred versus non-preferred hand), experimental condition (subjects versus experimenter’s hand) and response accuracy (correct versus incorrect answer) were within groups factors. No difference between the two groups was found. A main effect of response accuracy emerged (F(1, 7) = 10.18, P < 0.015), showing that average response times were significantly slower when they led to incorrect (858.4 ms, S.D. 189.4) compared to correct responses (675.4 ms, S.D. 189.4). The interaction between experimental condition and accuracy was also significant (F(1, 7) = 5.71, P < 0.0482): namely, response latency significantly increased (926.3 ms, S.D. 258.4) when subjects saw their own hand but were mistaken in recognising it (i.e. they responded “no”). Summing up, the main results of the present study were the following: (1) independently from their handedness, all subjects provided more correct responses when they were asked to recognise their preferred hand. (2) All subjects produced more correct responses when they saw their own hand compared to when they saw the experimenter’s hand, suggesting that errors increased when ambiguity was introduced. (3) Left-handers were on average slower in providing their response than were right-handed subjects. (4) Longer response latencies were not associated to correct responses. 4. Discussion Most people have a preferred hand, whose skilfulness largely exceeds that of the so-called non-preferred hand. Asymmetries related to motor dominance have been de-
1384
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
scribed to affect imagined movements [20]. The present study was designed in order to assess whether this notion extends to the capacity to recognise self-produced movements, or in broader terms, to self-awareness in motor acts. Right and left-handed healthy subjects were required to produce a hand gesture in an ambiguous environment, in which direct vision of their moving limb was prevented and visual feedback provided on a TV screen could be manipulated. Subjects were asked to decide whether they were the agents of the presented movements. Results showed that the number of correct responses given by the subjects varied according to the hand used. Both right- and left-handers were more accurate when asked to identify their preferred hand with respect to their non-preferred hand, suggesting the presence of a close link between motor dominance and motor awareness. We will define here motor awareness as the ability of an individual to both perceive and visually recognise a given movement as his/her own. In other words, as the knowledge enabling subjects to state that they are the source and the owner of a presented gesture. This ability necessarily comes into play in the task used in the present experiment. As previously reported [27], the accomplishment of the present task likely requires several steps. First, in addition to intentional information, subjects must be able to form a reliable representation of the movement they are about to execute. More precisely, subjects should be able to predict the proprioceptive and visual consequences of their act and to fit actual incoming kinaesthetic information into their predictive model. As a result, they will “know” which movement they are executing and whether it matches their intentions. Secondly, this information is compared to the one provided by the visual system, thus making possible to decide whether the movement shown to the subjects belongs to themselves or not. Once the result of this comparison is made conscious, a response can be produced. We recently provided some evidence that brain damage can disrupt this motor recognition mechanism at the level of action representation, thus preventing subjects from having a reliable image of the movement they are executing. More particularly, in apraxic patients suffering from left parietal lesion [27], the ability to either form or update a reliable predictive model of the executed gestures is impaired. Accordingly, when attempting to reproduce a complex hand posture, these patients fail in their execution as well as in recognising their own movement. The present results on healthy subjects support the hypothesis that the ability to visually recognise a movement as one’s own is subject to asymmetries similar to those present in motor skills. It is unlikely that our results may be accounted for by the possible time delay intervening between movement executed and movement observed by the subjects. Indeed, had this been the cue to the correct answer, it should have produced an equal number of correct responses for the preferred and non-preferred hand. The increase in the number of correct responses after presentation of the dominant hand could be accounted for by at
least two non-conflicting interpretations. Either subjects are more trained to attend to their preferred hand during praxis, or self-awareness of motor outputs is richer for the more skilled hand. According to the first interpretation, we can speculate that the better performance for the dominant hand could be a secondary effect of familiarity, subjects having acquired major expertise in visually controlling movements of their preferred hand. On the other hand, it could not be excluded that, when attempting to recognise their own hand, subjects made use of an internal representation of their dominant hand that is “motor” in nature. Evidence has been provided that motor habits can influence our perception of the visual world. Objects that can be grasped and manipulated appear to be processed also in relation to the manual dominance of the subject, thus suggesting the involvement of motor processes (namely motor imagery) in non-motor functions [7]. In the present task, subjects could have found recognition of the preferred hand easier either due to a more familiar visual image or to a richer visual–motor knowledge of their more skilled hand. We favour this second hypothesis, which points to a close relation between the system responsible for motor consciousness and the structures involved in motor representation and movement production. What kinds of mechanisms allow the interplay between consciousness of movement and motor production? Manual preference has an anatomic and functional counterpart in the expansion of hand motor cortex for the preferred hand [1,30] and this phenomenon has been invoked to explain the more efficient processing of the corresponding motor output. Efficiency could improve thanks to an increase in the number of neural elements and/or tighter connectivity: monkey data show both an enlargement and an increased degree of spatial complexity of motor areas opposite to the preferred hand [21]. At the cortical level, this modification may account for better interaction between neuronal assemblies representing muscles that are used in close spatial or temporal contiguity. Analysis of cytoarchitectonic differences between the two hemispheres [1] have shown an increased neuropil compartment in BA 4 in the dominant hemisphere, supporting the hypothesis that hand preference is associated with increased connectivity. At the behavioural level, such improved interaction may be responsible for a more skilled motor repertoire of the preferred hand. In agreement with the above data, the present results further indicate that not only motor production but also motor awareness can benefit from the anatomical and physiological pattern linked to hand preference. Handedness asymmetries seem, thus, to extend beyond the level of primary motor functions and likely emerge also in those structures which are involved in cognitive–motor operations. Two more points emerged from the results of the present experiment. First, the difference between preferred and non-preferred hand was less evident in the left-handers group. Second, left-handers were on average slower in giving their responses than right-handers. The former result could be accounted for by the fact that left-handers
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
are more heterogeneous and less markedly lateralised than right-handers: for instance, on preference measures, i.e. hand or foot usage, left-handers show larger variability and lower lateralisation than right-handed subjects [2]. In addition, in hand selection responses to unimanual tasks, left-handers do not respond with the preferred hand as consistently as do right-handers [8]. This reduced lateralisation emerged also in the sample examined in the present experiment: only one of our left-handed subjects scored between −0.90 and −1.00 at the Edinburgh Inventory, whereas five out of nine right-handed subjects showed such an extreme score. Indeed the absolute scores at the handedness test differed significantly between the two groups (U = 18.00, Z = 1.987, P = 0.046) Accordingly, we can assume that in our sample the reduced asymmetry in hand motor skills present in the left-handers group could be responsible for the reduced handedness effect. Both results are consistent with a previous report by Gentilucci et al. [9] in which right- and left-handed subjects were required to recognise whether a static hand holding an object, with either a congruent or incongruent type of grip, corresponded to the picture of a right or a left hand. Right-handers were faster in judging right hands, whereas no difference was found between preferred and non-preferred hand in the left-handed subjects. Moreover, left-handers were on average slower than right-handers. A similar pattern is also found in another previous report. When asked to decide whether a rotating screwdriver was either screwing or unscrewing [7], healthy subjects required longer response times in those conditions implying motor imagery of the non-dominant hand. Moreover, left-handers were overall slower than right-handers. As suggested by Gentilucci et al. [9], the behaviour of left-handers could be explained by the fact that, when required to covertly represent movement of their own hands, they likely rely on a “pictorial” (i.e. more observation-based) rather than “pragmatic” (i.e. motor experience-based) image of their dominant hand. In agreement with this hypothesis, we can assume here that, on the one hand, left-handers are less affected by asymmetries related to motor dexterity. On the other hand, they likely need further steps in the analysis of the presented image, such as a transition between a visual and a motor-based representation, in order to access those subtle kinematics cues which are required to recognise one’s own movement. In conclusion, we think that the present data demonstrate how the ability to recognise self-generated movements can be affected by motor dominance, suggesting that the system responsible for motor awareness is profoundly nested within the mechanisms controlling motor production.
Acknowledgements The authors wish to thank Virginie Czernecki for her precious help in collecting the data. This work was supported by CNRS.
1385
References [1] Amunts K, Schlaug G, Steinmetz H, Dabringhaus A, Roland PE, Zilles K. Asymmetry in the human motor cortex and handedness. Neuroimage 1996;4:216–22. [2] Borod JC, Caron HS, Koff E. Left-handers and right-handers compared on performance and preference measures of lateral dominance. British Journal of Psychology 1984;75:177–86. [3] Coren S, Porac C. Fifty centuries of right-handedness: the historical record. Science 1977;198:631–2. [4] Daprati E, Franck N, Georgieff N, Proust J, Pacherie E, Dalery J, et al. Looking for the agent. An investigation into consciousness of action and self-consciousness in schizophrenic patients. Cognition 1997;65:71–86. [5] Daprati E, Sirigu A, Pradat-Dihel P, Franck N, Jeannerod M. Recognition of self-produced movement in a case of severe neglect. Neurocase 2000;6:477–86. [6] Dassonville P, Zhu XH, Ugurbil K, Kim SG, Ashe J. Functional activation in motor cortex reflects the direction and the degree of handedness. Proceedings of the National Academy of the Sciences USA 1997;94:14015–8. [7] de’Sperati C, Stucchi N. Recognizing the motion of a graspable object is guided by handedness. Neuroreport 1997;8:2761–5. [8] Gabbard C, Iteya M, Rabb C. A lateralized comparison of handedness and object proximity. Canadian Journal of Experimental Psychology 1997;51:176–80. [9] Gentilucci M, Daprati E, Gangitano M. Right-handers and left-handers have different representations of their own hand. Cognitive Brain Research 1998;6:185–92. [10] Gerardin E, Sirigu A, Lehericy S, Poline JB, Gaymard B, Marsault C, et al. Partially overlapping neural networks for real and imagined hand movements. Cerebral Cortex 2000;10:1093–4. [11] Gilbert AN, Wysocki CS. Hand preference and age in the United States. Neuropsychologia 1992;30:601–8. [12] Gonzales-Rothi LJ, Heilman KM, editors. Apraxia: the neuropsychology of action. New York: Lawrence Erlbaum, 1996. [13] Grezes J, Costes N, Decety J. The effects of learning and intention on the neural network involved in the perception of meaningless actions. Brain 1999;122:1875–87. [14] Haaland KY, Harrington DL, Clinical implications of ipsilateral motor deficits after unilateral hemispheric damage. In: Bruno A, Chollet F, Vellas BJ, Albarede JL, editors. Facts and research in gerontology: stroke in the elderly. New York: Springer, 1996: pp. 101–14. [15] Iacoboni M, Woods RP, Brass M, Bekkering H, Mazziotta JC, Rizzolatti G. Cortical mechanisms of human imitation. Science 1999;286:2526–8. [16] Johnson SH. Cerebral organization of motor imagery: Contralateral control of grip selection in mentally represented prehension. Psychological Science 1998;9:219–22. [17] Johnson SH, Corballis PM, Gazzaniga MS. Within grasp but out of reach: Evidence for a double dissociation between imagined hand and arm movements in the left cerebral hemisphere. Neuropsychologia 2001;39:36–50. [18] Kawashima R, Inoue K, Sato K, Fukuda H. Functional asymmetry of cortical motor control in left-handed subjects. Neuroreport 1997;8:1729–32. [19] Kim SG, Ashe J, Hendrich K, Ellerman JM, Merkle H, Ugurbil K, et al. Functional magnetic resonance imaging of motor cortex: hemispheric asymmetry and handedness. Science 1993;261: 615–7. [20] Maruff P, Wilson PH, De Fazio J, Cerritelli B, Hedt A, Currie J. Asymmetries between dominant and non-dominant hands in real and imagined motor task performance. Neuropsychologia 1999;37:379– 84.
1386
E. Daprati, A. Sirigu / Neuropsychologia 40 (2002) 1379–1386
[21] Nudo RJ, Jenkins WM, Merzenich MM, Prejean T, Grenda R. Neurophysiological correlates of hand preference in primary motor cortex of adult squirrel monkeys. Journal of Neuroscience 1992;12:2918–47. [22] Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 1971;9:97–113. [23] Parsons LM, Gabrieli JD, Phelps EA, Gazzaniga MS. Cerebrally lateralized mental representations of hand shape and movement. Journal of Neuroscience 1998;18:6539–48. [24] Rizzolatti G, Fadiga L, Gallese V, Fogassi L. Premotor cortex and the recognition of motor actions. Cognitive Brain Research 1996;3:131– 41. [25] Rizzolatti G, Fadiga L, Matelli M, Bettinardi V, Paulesu E, Perani D, et al. Localization of grasp representations in humans by PET: 1. Observation versus execution. Experimental Brain Research 1996;111:246–52.
[26] Sirigu A, Duhamel JR, Cohen L, Pillon B, Dubois B, Agid Y. The mental representation of hand movements after parietal cortex damage. Science 1996;273:1564–7. [27] Sirigu A, Daprati E, Pradat-Dihel P, Franck N, Jeannerod M. Perception of self-generated movement following left parietal lesion. Brain 1999;122:1867–74. [28] Taniguchi M, Yoshimine T, Cheyne D, Kato A, Kihara T, Ninomiya H, et al. Neuromagnetic fields preceding unilateral movements in dextrals and sinistrals. Neuroreport 1998;9:1497–502. [29] Triggs WJ, Calvanio R, Levine M. Transcranial magnetic stimulation reveals a hemispheric asymmetry correlate of intermanual differences in motor performance. Neuropsychologia 1997;35: 1355–63. [30] Volkmann J, Schnitzler A, Witte OW, Freund HJ. Handedness and asymmetry of hand representation in human motor cortex. Journal of Neurophysiology 1998;79:2149–54.