- Email: [email protected]
Contribution of action to perception of self-orientation in humans
Neuroscience Letters 349 (2003) 99–102 www.elsevier.com/locate/neulet
Contribution of action to perception of self-orientation in humans Luc Tremblaya,*, Digby Elliottb a
Department of Health and Human Performance, University of Houston, 104 Garrison Gym, 3855 Holman, Houston, TX 77204, USA b Department of Kinesiology, McMaster University, 1280 Main St. W., Hamilton, ON L8S 4K1, Canada Received 4 December 2002; received in revised form 11 June 2003; accepted 30 June 2003
Abstract In this study, we evaluated the effect of action on the perception of an egocentric illusion. Eighteen participants were asked to indicate the perceived morphological horizon under two backward body tilts from upright in the median plane (i.e. pitch) using five different response modes. The response modes varied in the degree of motor and cognitive involvement. Differences in perception of the morphological horizon between the two body tilts were significant only when proximal limb control was not involved. These results suggest that motor involvement and frame of reference may both be important in visual-vestibular illusions. q 2003 Published by Elsevier Ireland Ltd. Keywords: Human perception; Morphological horizon; Perception-action coupling; Dorsal-ventral streams; Visual-vestibular illusion
The typical procedure used to investigate sensorimotor interactions has been to manipulate stimulus variables and to observe their impact on perceptual-motor behaviours. In the last decade, however, researchers have accumulated a large body of evidence to indicate that the type of motor or cognitive response a person makes also influences an individual’s perceptual judgements. In this context, a number of recent studies have shown that susceptibility to certain visual illusions depends on the response requirements associated with the perceptual decision. For example, Aglioti et al. [1] used grip size to evaluate the perceived diameter of poker chip disks surrounded by TitchenerEbbinghaus circles. When simply estimating the size of the poker chip with their grip size, participants’ judgements were affected by the size of the surrounding circles. However, during a reach and grasp movement, the same grip component was found to be unaffected by the illusion. Similar dissociations between cognitive and motor judgements have been reported for the Mu¨ller-Lyer illusion [6]. Milner and Goodale [15] have used this type of finding, as well as clinical and animal evidence, to forward a neurophysiological model of visual processing. They suggested that there are two distinct visual streams. These streams are associated with different types of perceptual judgements and response requirements. *
Corresponding author. Tel.: þ 1-419-946-0200; fax: þ1-416-971-2118. E-mail addresses: [email protected] (L. Tremblay), [email protected] (D. Elliott) . 0304-3940/03/$ - see front matter q 2003 Published by Elsevier Ireland Ltd. doi:10.1016/S0304-3940(03)00797-3
The ventral stream, which projects from the visual cortex to the inferotemporal cortex, is specialized for form perception and object recognition. The response mode associated with the ventral stream is usually cognitive and language-based. Moreover, judgements of form usually require an ‘object-centred’ view of the stimulus (i.e. frame of reference centred on the object) as well as the encoding of the object’s intrinsic properties (e.g. its use). In ‘objectcentred’ perceptual decision-making individuals adopt an allocentric frame of reference. In the case of visual illusions (e.g. Ref. [1]), an allocentric frame of reference entails consideration of the target stimulus in the context of the objects around it (e.g. adjacent circles in the TitchenerEbbinghaus illusion). The dorsal stream extends from the visual cortex to the superior parietal areas. This stream is associated with more action-based perceptual judgements, which require the control of limb and body movements (e.g. moving to a location in space, intercepting a moving object). The dorsal stream provides a ‘viewer-centred’ perspective of the stimulus. This is sometimes referred to as ‘egocentric coding’. Thus, when participants actually reach for an object, they adopt a perspective in which the encoding of the object is independent of the visual surround. Presumably, this makes visual-motor behaviours associated with the dorsal stream uninfluenced by context-induced illusions. In the majority of the illusion studies designed to examine the dissociation between perception and action, the response requirements and the frame of reference required for a decision covary (see Ref. [8] for a review).
100
L. Tremblay, D. Elliott / Neuroscience Letters 349 (2003) 99–102
The purpose of this study was to examine the influence of response mode on the manifestation of an egocentric illusion induced by body tilt. Specifically, we measured the morphological horizon (MH) [4]. This illusion is egocentric in nature as it involves a single visual stimulus in an otherwise dark environment. The MH is associated with vestibulo-ocular interactions, which can be altered when the magnitude and/or direction of inertial forces are manipulated [5,17] (see also Refs. [12,13]). These vestibulo-ocular biases are associated with, but are not completely explained by, a different resting eye orientation [2,3]. For example, a backward body pitch (i.e. supine position) leads to a footward bias in the perceived MH [7]. This illusion is unconscious and usually examined with verbal judgements.1 Eighteen naive members of the McMaster University community (nine females and nine males) took part in the experiment (mean age 25.3 years, SD 3.9 years) in exchange for a small financial compensation (i.e. $10). Each participant provided informed consent according to the guidelines of the ethics board of McMaster University. The apparatus included a tilting device to incline the participant and a device to record the perceived MH. The tilting device was installed in a dark room and consisted of a bed that could rotate around its transverse axis. The participant was secured to the bed by means of a waist harness (KBoum). Spine board straps (Ferno) and Velcro straps were used to stabilize the shoulders and the feet, respectively. Head movements were restricted by an adjustable cervical immobilization collar (Wizloc). The bed was equipped with a 8 cm thick foam mattress for comfort. The measurement device consisted of an arc with a radius of 53.5 cm, 4.5 cm wide and painted black. In the centre of the arc, a smooth round manipulandum was equipped with a potentiometer and a laser pointer (see Fig. 1). The potentiometer was linked through an Analog-to-Digital converter (Dataq Instruments – Model 220) to a computer (PC compatible, AMD450MHz processor), which acquired data with the Windaq program (Dataq Instruments) at 200 Hz (accuracy of 0.08 deg). This device was used to obtain 1 The perceived morphological horizon should not be confused with visually perceived eye-level as the latter is related to the perceived horizon independent of body orientation. In order to examine response mode, we employed a task that involved a visual stimulus manipulated by the participant to provide her/his response instead of using a visual target per se. 2 The specific sequence order adopted in this study was designed to reduce the opportunity for protocol to protocol carry-over effects. In past work, we have demonstrated that there is no reduction in judgement bias across trials, or blocks of trials for either the switch response mode or the wrist response mode when no feedback is provided about performance [19]. We were concerned, however, that the work against gravity associated with a whole arm movement (e.g. fourth and fifth response modes) might provide participants with enough information to modify their performance. The absence of any influence of block, as well as the reversal in the direction of bias between the fourth and the fifth response mode, however, indicates that this was not the case.
Fig. 1. Drawing of the arc-manipulandum part of the straight-ahead measurement device installed on the inverting bed.
perceptual judgements with five response modes. The participants were able to adjust the position of the laser beam (or just their limb depending on the response mode) at the ‘perceived straight-ahead’ (i.e. perceived MH) in a selfpaced manner without gravitational influences other than their own limb weight. This could be done by using the manipulandum alone (grip diameter 7.3 cm) or when the device was equipped with a counterbalanced aiming rod (42 cm in length) installed parallel to the orientation of the laser. In addition, it was possible to engage a switch-controlled DC motor against the manipulandum, which moved the laser at 4.55 deg/s and allowed either the experimenter or the participant to adjust the position of the laser at the perceived MH. The link between the motor and the manipulandum was manipulated according to the experimental task. This arc-manipulandum was installed on a 7 cm by 107 cm piece of plywood, which could be moved to align the shaft of the potentiometer with the participant’s eyes in the frontal and the saggital planes. This device was on the right-hand side of the participant. First, the participant was explained the experimental task and then secured to the bed. She/he was asked to make perceptual judgements of her/his ‘straight-ahead at eyelevel’. This was further described as ‘a point 90 deg relative to head orientation at eye-level’, and the explanation was visually supported with a stick figure on a chalkboard. The participant was instructed to keep her/his eyes still during each trial. The experimental conditions included perceptual decisions from five response modes performed in two body orientations. The DC motor linked to the laser was controlled by the experimenter in the first response mode (Verbal) and by the participant in the second response mode (Switch). The third response mode involved the use of the smooth manipulandum alone (Wrist). For the last response modes, the aiming rod and its counterweight were installed on the manipulandum. In the fourth response mode, the laser was shut off and the participant was required to aim her/his index finger at the MH by manipulating the rod (Arm w/o Laser). The fifth response mode was similar to the fourth except that the laser was turned on (Arm). The presentation of the response modes was sequential (i.e. from Verbal to Arm)2 and half of the participants performed in the 15 deg orientation first while the other half started with the 75 deg
L. Tremblay, D. Elliott / Neuroscience Letters 349 (2003) 99–102
orientation. Whole body pitch, backward from the anatomical upright, was manipulated prior to each block of ten trials. As vestibulo-ocular biases take time to stabilize (see Ref. [13]), there was a waiting delay of 90 s between the time of body pitch and the first perceptual response. Each trial was performed as follows. The direction of the laser was oriented footward and outside the participant’s field of view before each trial even if it was shut off (i.e. Arm w/o Laser). Then, the experimenter gave a signal that allowed the participants to perform the trial. The participants were instructed to keep their eyes still during the trial and to position the laser and/or finger at ‘the perceived straight-ahead’, giving a verbal signal to the experimenter when the final position was achieved. This signal was used by the experimenter to insert a marker in the data collection program and determine the perceived MH (i.e. signed angle value). Afterwards, the manipulandum/aiming rod was moved footward, outside the field of view before the next trial. The perceived MH was calculated for each block of five trials, for each orientation (i.e. 15 deg and 75 deg) and in each response mode. This led to a 5 Response Mode (Verbal, Switch, Wrist, Arm w/o Laser, Arm) £ 2 Position (15 deg, 75 deg) £ 2 Block (trials 1 – 5, trials 6 –10) repeated measures analysis of variance. The analysis of mean perceived MH revealed significant effects for Response Mode (Fð4; 64Þ ¼ 36:07, P , 0:001), and Position (Fð1; 16Þ ¼ 7:22, P , 0:05), as well as a Response Mode by Position interaction (Fð4; 64Þ ¼ 24:11, P , 0:001). The post-hoc analysis of the interaction (Tukey HSD, P , 0:05) revealed a significant difference in the perceptual judgements between the 15 deg and the 75 deg orientation for all response modes except for the Arm. As evident in Fig. 2, participants perceived their MH more footward in the 15 deg orientation than in the 75 deg orientation with the Verbal, Switch, and Wrist response modes. When participants aimed without the laser, their perception of the MH was more footward in the 75 deg orientation than in the 15 deg orientation. In the Arm response mode, there was no difference in the perceived MH between the two orientations. When the impact of response mode was examined within the 75 deg orientation, participants exhibited greater footward bias in the Arm w/ o Laser response mode than in the other response modes. For the 15 deg orientation, the MH was evaluated more footward with the Verbal and Arm w/o Laser response modes than with the Switch, Wrist, and Arm response modes. The primary purpose of this study was to determine the influence of motor involvement on perceptual biases in selforientation. Participants judged the MH to be more footward in the 15 deg conditions than in the 75 deg conditions, but only with a verbal response mode or when the actual movement demands of the decision were minimal (i.e. use of the switch and wrist rotation to move the laser). There was no difference between the two orientations when a laser
101
Fig. 2. Task by Position interaction for the average straight-ahead perception bias (deg) with standard error bars.
beam provided visual information about the position of the limb during a whole arm movement. Thus, it appears that the response mode influenced perceptual judgements of the MH. Interestingly, when aiming without visual information, participants exhibited greater footward bias than with any other response mode and they perceived the MH more footward in the 75 deg orientation than in the 15 deg situation. These latter results highlight the contribution of visual information processing for the observed effects of action on perceptual judgements. Our findings are similar to other work examining the influence of response mode on perceptual decision-making [1,6,10,11,15]. That is, participants were less influenced by an illusion when the perceptual judgements involved the use of whole limb movements than when only a cognitive response was required (cf. Ref. [8]). As mentioned above, this was only true when visual information was available. Unlike work involving purely visual illusions, it is difficult to attribute the dissociation between perception and action in this study to the use of a context-dependent vs. a contextfree frame of reference [1,11]. Specifically, participants exhibited an illusory bias while making cognitive, but not motor, judgements about self-orientation when no visual context was present. It would appear that the ventral and dorsal visual streams are not the only neural systems that contribute to a dissociation between perception and action. Certainly, studies involving normal participants and patients with cerebellar pathologies have made it clear that any normal reaching gesture requires extensive involvement of the cerebellum and the basal ganglia (see Ref. [18]). The cerebellum contains more than half of the neurons of the human brain, and it maps multiple sources of afferent input, including the types of vestibular input important for the perceptual judgements made in this study. Descending efferent commands are necessarily routed through the cerebellum. In addition, afferent information from the muscle spindles and Golgi tendon organs is encoded in the cerebellum in conjunction with the
102
L. Tremblay, D. Elliott / Neuroscience Letters 349 (2003) 99–102
afferent visual and vestibular information. Perhaps these sensorimotor interactions in the central vermis of the cerebellum modulate perceptual judgements.3 As Roy and Cullen [16] have demonstrated in their single cell work with vestibular neurons, it is not the processing of afferent information alone that determines veridical perception but rather the integration of this information with the intended movement consequences (i.e. efference). In the work reported here, it is possible that motor related cerebellar involvement in the perceptual process makes decisionmaking less susceptible to visual illusions. Regardless of the underlying mechanism, the availability of afferent and efferent sources of information appears to mediate dissociation of perception and action. More importantly, the effect of action on perceptual judgement is not limited to visual processing [15] but extends to visuoocular [9], vestibulo-kinaesthetic [16], and here, to visualvestibular interactions.4 From an evolutionary perspective, it is clear that the ability to move and acquire objects in the environment developed long before neural systems associated with more cognitive decision-making. In this context, it is rather comforting to realize that the perceptual biases associated with many cognitive judgements do not interfere with more basic survival behaviours.
Acknowledgements This research was supported by a research grant from the Natural Sciences and Engineering Research Council of Canada (NSERC) and a Canada Research Chair (CRC) awarded to Digby Elliott, and a scholarship from les Fonds pour la Formation des Chercheurs et l’Aide a` la Recherche du Que´bec (FCAR) awarded to Luc Tremblay. This work was conducted at McMaster University and was part of the first author’s PhD dissertation. The authors also wish to thank Le Groupe Experts Conseils PM Inc. for providing the KBoum safety harness. Many thanks to John Moroz for his technical assistance as well as Lindsay Finch and Steve Passmore for their help during data collection.
References [1] S. Aglioti, M.A. Goodale, J.F.X. DeSouza, Size-contrast illusions deceive the eye but not the hand, Curr. Biol. 5 (1995) 679–685.
3
It has to be acknowledged that the cerebellum and basal ganglia are not ‘purely motor structures’ and are also associated with cognitive functions [14]. 4 The term visual-vestibular is used for brevity. Although the vestibular system is known to influence visual perception, other sensory systems (e.g. kinaesthetic system) are stimulated during manipulation of whole body orientation. These systems could also be associated with the observed results. The overall observation remains that dissociation between perception and action is not constrained to visual illusions.
[2] U. Brandt, E. Fluur, Postural perceptions and eye displacements produced by a resultant vector acting in the median plane of the head. I. Response along three axes by stepwise increasing F with the subject heading centripetally in an erect and tilted position, Acta Otolaryngol. 63 (1967) 489–502. [3] U. Brandt, E. Fluur, Postural perceptions and eye displacements produced by a resultant vector acting in the median plane of the head. II. Continuous responses along the Y axis with the subject in a vertical position, heading centripetally and centrifugally, Acta Otolaryngol. 63 (1967) 546 –578. [4] M.M. Cohen, C.A. Larson, Human spatial orientation in the pitch dimension, Percept. Psychophys. 16 (1974) 508–512. [5] M.M. Cohen, R.B. Welch, Visual-motor control in altered gravity, in: L. Proteau, D. Elliott (Eds.), Vision and Motor Control, Elsevier, Amsterdam, 1992, pp. 153– 175. [6] E. Daprati, M. Gentilucci, Grasping an illusion, Neuropsychologia 35 (1997) 1577–1582. [7] S.M. Ebenholtz, W. Shebilske, The doll reflex: ocular counterrolling with head-body tilt in the median plane, Vis. Res. 15 (1975) 713 –717. [8] V.H. Franz, Action does not resist visual illusions, Trends Cogn. Sci. 5 (2001) 457 –459. [9] S. Gielen, Helmoltz: founder of the action-perception theory, in: M.L. Latash, V.M. Zatsiorsky (Eds.), Classics in Movement Science, Human Kinetics, Champaign, IL, 2001, pp. 221 –242. [10] S. Glover, Visual illusions affect planning but not control, Trends Cogn. Sci. 6 (2002) 288 –292. [11] S.R. Glover, P. Dixon, Dynamic illusion effects in a reaching task: evidence for separate visual representations in the planning and the control of reaching, J. Exp. Psychol. Hum. Percept. Perform. 27 (2001) 560 –572. [12] A. Graybiel, Oculogravic illusion, Arch. Ophthalmol. 48 (1952) 605 –615. [13] A. Graybiel, R.H. Brown, The delay in visual reorientation following exposure to a change in direction of resultant force on a human centrifuge, J. Gen. Psychol. 45 (1951) 143–150. [14] F.A. Middleton, P.L. Strick, Anatomical evidence for cerebellar and basal ganglia involvement in higher cognitive function, Science 21 (1994) 458. [15] M.A. Milner, A.D. Goodale, The Visual Brain in Action, Oxford University Press, Oxford, 1995. [16] J.E. Roy, K.E. Cullen, Selective processing of vestibular reafference during self-generated head motion, J. Neurosci. 21 (2001) 2131–2142. [17] H. Scho¨ne, On the role of gravity in human spatial orientation, Aerosp. Med. 35 (1964) 764 –772. [18] J.F. Stein, M. Glickstein, Role of the cerebellum in visual guidance of movement, Physiol. Rev. 72 (1992) 967 –1017. [19] L. Tremblay, Dissociation of perception and action or contribution of action to perception? Young scientist award presentation at the meeting of the Canadian Society for Psychomotor Learning and Sport Psychology, Vancouver, BC, Canada.
Contribution of action to perception of self-orientation in humans Luc Tremblaya,*, Digby Elliottb a
Department of Health and Human Performance, University of Houston, 104 Garrison Gym, 3855 Holman, Houston, TX 77204, USA b Department of Kinesiology, McMaster University, 1280 Main St. W., Hamilton, ON L8S 4K1, Canada Received 4 December 2002; received in revised form 11 June 2003; accepted 30 June 2003
Abstract In this study, we evaluated the effect of action on the perception of an egocentric illusion. Eighteen participants were asked to indicate the perceived morphological horizon under two backward body tilts from upright in the median plane (i.e. pitch) using five different response modes. The response modes varied in the degree of motor and cognitive involvement. Differences in perception of the morphological horizon between the two body tilts were significant only when proximal limb control was not involved. These results suggest that motor involvement and frame of reference may both be important in visual-vestibular illusions. q 2003 Published by Elsevier Ireland Ltd. Keywords: Human perception; Morphological horizon; Perception-action coupling; Dorsal-ventral streams; Visual-vestibular illusion
The typical procedure used to investigate sensorimotor interactions has been to manipulate stimulus variables and to observe their impact on perceptual-motor behaviours. In the last decade, however, researchers have accumulated a large body of evidence to indicate that the type of motor or cognitive response a person makes also influences an individual’s perceptual judgements. In this context, a number of recent studies have shown that susceptibility to certain visual illusions depends on the response requirements associated with the perceptual decision. For example, Aglioti et al. [1] used grip size to evaluate the perceived diameter of poker chip disks surrounded by TitchenerEbbinghaus circles. When simply estimating the size of the poker chip with their grip size, participants’ judgements were affected by the size of the surrounding circles. However, during a reach and grasp movement, the same grip component was found to be unaffected by the illusion. Similar dissociations between cognitive and motor judgements have been reported for the Mu¨ller-Lyer illusion [6]. Milner and Goodale [15] have used this type of finding, as well as clinical and animal evidence, to forward a neurophysiological model of visual processing. They suggested that there are two distinct visual streams. These streams are associated with different types of perceptual judgements and response requirements. *
Corresponding author. Tel.: þ 1-419-946-0200; fax: þ1-416-971-2118. E-mail addresses: [email protected] (L. Tremblay), [email protected] (D. Elliott) . 0304-3940/03/$ - see front matter q 2003 Published by Elsevier Ireland Ltd. doi:10.1016/S0304-3940(03)00797-3
The ventral stream, which projects from the visual cortex to the inferotemporal cortex, is specialized for form perception and object recognition. The response mode associated with the ventral stream is usually cognitive and language-based. Moreover, judgements of form usually require an ‘object-centred’ view of the stimulus (i.e. frame of reference centred on the object) as well as the encoding of the object’s intrinsic properties (e.g. its use). In ‘objectcentred’ perceptual decision-making individuals adopt an allocentric frame of reference. In the case of visual illusions (e.g. Ref. [1]), an allocentric frame of reference entails consideration of the target stimulus in the context of the objects around it (e.g. adjacent circles in the TitchenerEbbinghaus illusion). The dorsal stream extends from the visual cortex to the superior parietal areas. This stream is associated with more action-based perceptual judgements, which require the control of limb and body movements (e.g. moving to a location in space, intercepting a moving object). The dorsal stream provides a ‘viewer-centred’ perspective of the stimulus. This is sometimes referred to as ‘egocentric coding’. Thus, when participants actually reach for an object, they adopt a perspective in which the encoding of the object is independent of the visual surround. Presumably, this makes visual-motor behaviours associated with the dorsal stream uninfluenced by context-induced illusions. In the majority of the illusion studies designed to examine the dissociation between perception and action, the response requirements and the frame of reference required for a decision covary (see Ref. [8] for a review).
100
L. Tremblay, D. Elliott / Neuroscience Letters 349 (2003) 99–102
The purpose of this study was to examine the influence of response mode on the manifestation of an egocentric illusion induced by body tilt. Specifically, we measured the morphological horizon (MH) [4]. This illusion is egocentric in nature as it involves a single visual stimulus in an otherwise dark environment. The MH is associated with vestibulo-ocular interactions, which can be altered when the magnitude and/or direction of inertial forces are manipulated [5,17] (see also Refs. [12,13]). These vestibulo-ocular biases are associated with, but are not completely explained by, a different resting eye orientation [2,3]. For example, a backward body pitch (i.e. supine position) leads to a footward bias in the perceived MH [7]. This illusion is unconscious and usually examined with verbal judgements.1 Eighteen naive members of the McMaster University community (nine females and nine males) took part in the experiment (mean age 25.3 years, SD 3.9 years) in exchange for a small financial compensation (i.e. $10). Each participant provided informed consent according to the guidelines of the ethics board of McMaster University. The apparatus included a tilting device to incline the participant and a device to record the perceived MH. The tilting device was installed in a dark room and consisted of a bed that could rotate around its transverse axis. The participant was secured to the bed by means of a waist harness (KBoum). Spine board straps (Ferno) and Velcro straps were used to stabilize the shoulders and the feet, respectively. Head movements were restricted by an adjustable cervical immobilization collar (Wizloc). The bed was equipped with a 8 cm thick foam mattress for comfort. The measurement device consisted of an arc with a radius of 53.5 cm, 4.5 cm wide and painted black. In the centre of the arc, a smooth round manipulandum was equipped with a potentiometer and a laser pointer (see Fig. 1). The potentiometer was linked through an Analog-to-Digital converter (Dataq Instruments – Model 220) to a computer (PC compatible, AMD450MHz processor), which acquired data with the Windaq program (Dataq Instruments) at 200 Hz (accuracy of 0.08 deg). This device was used to obtain 1 The perceived morphological horizon should not be confused with visually perceived eye-level as the latter is related to the perceived horizon independent of body orientation. In order to examine response mode, we employed a task that involved a visual stimulus manipulated by the participant to provide her/his response instead of using a visual target per se. 2 The specific sequence order adopted in this study was designed to reduce the opportunity for protocol to protocol carry-over effects. In past work, we have demonstrated that there is no reduction in judgement bias across trials, or blocks of trials for either the switch response mode or the wrist response mode when no feedback is provided about performance [19]. We were concerned, however, that the work against gravity associated with a whole arm movement (e.g. fourth and fifth response modes) might provide participants with enough information to modify their performance. The absence of any influence of block, as well as the reversal in the direction of bias between the fourth and the fifth response mode, however, indicates that this was not the case.
Fig. 1. Drawing of the arc-manipulandum part of the straight-ahead measurement device installed on the inverting bed.
perceptual judgements with five response modes. The participants were able to adjust the position of the laser beam (or just their limb depending on the response mode) at the ‘perceived straight-ahead’ (i.e. perceived MH) in a selfpaced manner without gravitational influences other than their own limb weight. This could be done by using the manipulandum alone (grip diameter 7.3 cm) or when the device was equipped with a counterbalanced aiming rod (42 cm in length) installed parallel to the orientation of the laser. In addition, it was possible to engage a switch-controlled DC motor against the manipulandum, which moved the laser at 4.55 deg/s and allowed either the experimenter or the participant to adjust the position of the laser at the perceived MH. The link between the motor and the manipulandum was manipulated according to the experimental task. This arc-manipulandum was installed on a 7 cm by 107 cm piece of plywood, which could be moved to align the shaft of the potentiometer with the participant’s eyes in the frontal and the saggital planes. This device was on the right-hand side of the participant. First, the participant was explained the experimental task and then secured to the bed. She/he was asked to make perceptual judgements of her/his ‘straight-ahead at eyelevel’. This was further described as ‘a point 90 deg relative to head orientation at eye-level’, and the explanation was visually supported with a stick figure on a chalkboard. The participant was instructed to keep her/his eyes still during each trial. The experimental conditions included perceptual decisions from five response modes performed in two body orientations. The DC motor linked to the laser was controlled by the experimenter in the first response mode (Verbal) and by the participant in the second response mode (Switch). The third response mode involved the use of the smooth manipulandum alone (Wrist). For the last response modes, the aiming rod and its counterweight were installed on the manipulandum. In the fourth response mode, the laser was shut off and the participant was required to aim her/his index finger at the MH by manipulating the rod (Arm w/o Laser). The fifth response mode was similar to the fourth except that the laser was turned on (Arm). The presentation of the response modes was sequential (i.e. from Verbal to Arm)2 and half of the participants performed in the 15 deg orientation first while the other half started with the 75 deg
L. Tremblay, D. Elliott / Neuroscience Letters 349 (2003) 99–102
orientation. Whole body pitch, backward from the anatomical upright, was manipulated prior to each block of ten trials. As vestibulo-ocular biases take time to stabilize (see Ref. [13]), there was a waiting delay of 90 s between the time of body pitch and the first perceptual response. Each trial was performed as follows. The direction of the laser was oriented footward and outside the participant’s field of view before each trial even if it was shut off (i.e. Arm w/o Laser). Then, the experimenter gave a signal that allowed the participants to perform the trial. The participants were instructed to keep their eyes still during the trial and to position the laser and/or finger at ‘the perceived straight-ahead’, giving a verbal signal to the experimenter when the final position was achieved. This signal was used by the experimenter to insert a marker in the data collection program and determine the perceived MH (i.e. signed angle value). Afterwards, the manipulandum/aiming rod was moved footward, outside the field of view before the next trial. The perceived MH was calculated for each block of five trials, for each orientation (i.e. 15 deg and 75 deg) and in each response mode. This led to a 5 Response Mode (Verbal, Switch, Wrist, Arm w/o Laser, Arm) £ 2 Position (15 deg, 75 deg) £ 2 Block (trials 1 – 5, trials 6 –10) repeated measures analysis of variance. The analysis of mean perceived MH revealed significant effects for Response Mode (Fð4; 64Þ ¼ 36:07, P , 0:001), and Position (Fð1; 16Þ ¼ 7:22, P , 0:05), as well as a Response Mode by Position interaction (Fð4; 64Þ ¼ 24:11, P , 0:001). The post-hoc analysis of the interaction (Tukey HSD, P , 0:05) revealed a significant difference in the perceptual judgements between the 15 deg and the 75 deg orientation for all response modes except for the Arm. As evident in Fig. 2, participants perceived their MH more footward in the 15 deg orientation than in the 75 deg orientation with the Verbal, Switch, and Wrist response modes. When participants aimed without the laser, their perception of the MH was more footward in the 75 deg orientation than in the 15 deg orientation. In the Arm response mode, there was no difference in the perceived MH between the two orientations. When the impact of response mode was examined within the 75 deg orientation, participants exhibited greater footward bias in the Arm w/ o Laser response mode than in the other response modes. For the 15 deg orientation, the MH was evaluated more footward with the Verbal and Arm w/o Laser response modes than with the Switch, Wrist, and Arm response modes. The primary purpose of this study was to determine the influence of motor involvement on perceptual biases in selforientation. Participants judged the MH to be more footward in the 15 deg conditions than in the 75 deg conditions, but only with a verbal response mode or when the actual movement demands of the decision were minimal (i.e. use of the switch and wrist rotation to move the laser). There was no difference between the two orientations when a laser
101
Fig. 2. Task by Position interaction for the average straight-ahead perception bias (deg) with standard error bars.
beam provided visual information about the position of the limb during a whole arm movement. Thus, it appears that the response mode influenced perceptual judgements of the MH. Interestingly, when aiming without visual information, participants exhibited greater footward bias than with any other response mode and they perceived the MH more footward in the 75 deg orientation than in the 15 deg situation. These latter results highlight the contribution of visual information processing for the observed effects of action on perceptual judgements. Our findings are similar to other work examining the influence of response mode on perceptual decision-making [1,6,10,11,15]. That is, participants were less influenced by an illusion when the perceptual judgements involved the use of whole limb movements than when only a cognitive response was required (cf. Ref. [8]). As mentioned above, this was only true when visual information was available. Unlike work involving purely visual illusions, it is difficult to attribute the dissociation between perception and action in this study to the use of a context-dependent vs. a contextfree frame of reference [1,11]. Specifically, participants exhibited an illusory bias while making cognitive, but not motor, judgements about self-orientation when no visual context was present. It would appear that the ventral and dorsal visual streams are not the only neural systems that contribute to a dissociation between perception and action. Certainly, studies involving normal participants and patients with cerebellar pathologies have made it clear that any normal reaching gesture requires extensive involvement of the cerebellum and the basal ganglia (see Ref. [18]). The cerebellum contains more than half of the neurons of the human brain, and it maps multiple sources of afferent input, including the types of vestibular input important for the perceptual judgements made in this study. Descending efferent commands are necessarily routed through the cerebellum. In addition, afferent information from the muscle spindles and Golgi tendon organs is encoded in the cerebellum in conjunction with the
102
L. Tremblay, D. Elliott / Neuroscience Letters 349 (2003) 99–102
afferent visual and vestibular information. Perhaps these sensorimotor interactions in the central vermis of the cerebellum modulate perceptual judgements.3 As Roy and Cullen [16] have demonstrated in their single cell work with vestibular neurons, it is not the processing of afferent information alone that determines veridical perception but rather the integration of this information with the intended movement consequences (i.e. efference). In the work reported here, it is possible that motor related cerebellar involvement in the perceptual process makes decisionmaking less susceptible to visual illusions. Regardless of the underlying mechanism, the availability of afferent and efferent sources of information appears to mediate dissociation of perception and action. More importantly, the effect of action on perceptual judgement is not limited to visual processing [15] but extends to visuoocular [9], vestibulo-kinaesthetic [16], and here, to visualvestibular interactions.4 From an evolutionary perspective, it is clear that the ability to move and acquire objects in the environment developed long before neural systems associated with more cognitive decision-making. In this context, it is rather comforting to realize that the perceptual biases associated with many cognitive judgements do not interfere with more basic survival behaviours.
Acknowledgements This research was supported by a research grant from the Natural Sciences and Engineering Research Council of Canada (NSERC) and a Canada Research Chair (CRC) awarded to Digby Elliott, and a scholarship from les Fonds pour la Formation des Chercheurs et l’Aide a` la Recherche du Que´bec (FCAR) awarded to Luc Tremblay. This work was conducted at McMaster University and was part of the first author’s PhD dissertation. The authors also wish to thank Le Groupe Experts Conseils PM Inc. for providing the KBoum safety harness. Many thanks to John Moroz for his technical assistance as well as Lindsay Finch and Steve Passmore for their help during data collection.
References [1] S. Aglioti, M.A. Goodale, J.F.X. DeSouza, Size-contrast illusions deceive the eye but not the hand, Curr. Biol. 5 (1995) 679–685.
3
It has to be acknowledged that the cerebellum and basal ganglia are not ‘purely motor structures’ and are also associated with cognitive functions [14]. 4 The term visual-vestibular is used for brevity. Although the vestibular system is known to influence visual perception, other sensory systems (e.g. kinaesthetic system) are stimulated during manipulation of whole body orientation. These systems could also be associated with the observed results. The overall observation remains that dissociation between perception and action is not constrained to visual illusions.
[2] U. Brandt, E. Fluur, Postural perceptions and eye displacements produced by a resultant vector acting in the median plane of the head. I. Response along three axes by stepwise increasing F with the subject heading centripetally in an erect and tilted position, Acta Otolaryngol. 63 (1967) 489–502. [3] U. Brandt, E. Fluur, Postural perceptions and eye displacements produced by a resultant vector acting in the median plane of the head. II. Continuous responses along the Y axis with the subject in a vertical position, heading centripetally and centrifugally, Acta Otolaryngol. 63 (1967) 546 –578. [4] M.M. Cohen, C.A. Larson, Human spatial orientation in the pitch dimension, Percept. Psychophys. 16 (1974) 508–512. [5] M.M. Cohen, R.B. Welch, Visual-motor control in altered gravity, in: L. Proteau, D. Elliott (Eds.), Vision and Motor Control, Elsevier, Amsterdam, 1992, pp. 153– 175. [6] E. Daprati, M. Gentilucci, Grasping an illusion, Neuropsychologia 35 (1997) 1577–1582. [7] S.M. Ebenholtz, W. Shebilske, The doll reflex: ocular counterrolling with head-body tilt in the median plane, Vis. Res. 15 (1975) 713 –717. [8] V.H. Franz, Action does not resist visual illusions, Trends Cogn. Sci. 5 (2001) 457 –459. [9] S. Gielen, Helmoltz: founder of the action-perception theory, in: M.L. Latash, V.M. Zatsiorsky (Eds.), Classics in Movement Science, Human Kinetics, Champaign, IL, 2001, pp. 221 –242. [10] S. Glover, Visual illusions affect planning but not control, Trends Cogn. Sci. 6 (2002) 288 –292. [11] S.R. Glover, P. Dixon, Dynamic illusion effects in a reaching task: evidence for separate visual representations in the planning and the control of reaching, J. Exp. Psychol. Hum. Percept. Perform. 27 (2001) 560 –572. [12] A. Graybiel, Oculogravic illusion, Arch. Ophthalmol. 48 (1952) 605 –615. [13] A. Graybiel, R.H. Brown, The delay in visual reorientation following exposure to a change in direction of resultant force on a human centrifuge, J. Gen. Psychol. 45 (1951) 143–150. [14] F.A. Middleton, P.L. Strick, Anatomical evidence for cerebellar and basal ganglia involvement in higher cognitive function, Science 21 (1994) 458. [15] M.A. Milner, A.D. Goodale, The Visual Brain in Action, Oxford University Press, Oxford, 1995. [16] J.E. Roy, K.E. Cullen, Selective processing of vestibular reafference during self-generated head motion, J. Neurosci. 21 (2001) 2131–2142. [17] H. Scho¨ne, On the role of gravity in human spatial orientation, Aerosp. Med. 35 (1964) 764 –772. [18] J.F. Stein, M. Glickstein, Role of the cerebellum in visual guidance of movement, Physiol. Rev. 72 (1992) 967 –1017. [19] L. Tremblay, Dissociation of perception and action or contribution of action to perception? Young scientist award presentation at the meeting of the Canadian Society for Psychomotor Learning and Sport Psychology, Vancouver, BC, Canada.