BRAIN
AND
COGNITION
u),
227-235 (1992)
Asymmetries in the Spatial Localization of Transformed Targets ROMEO CHUA, RICHARD Motor Behaviour
Laboratory,
G. CARSON, AND DAVID GOODMAN
School of Kinesiology,
Simon Fraser University
AND DIGBY ELLIOTT Department of Physical Education,
McMaster
University
This study was designed to examine the contribution of the right cerebral hemisphere in the spatial localization of visual targets for manual aiming. Visual targets were briefly presented to the right and left fields and subjects were required to point either to the target location, or a “mirror” image of the target location with their right or left index finger. Whereas reaction times were faster for lefthand pointing than for right-hand pointing, there was no differential effect of the mirror image transformation. This suggeststhat left-hand reaction time advantages are more related to right hemisphere involvement in the spatial parameterization of the movement than spatial localization of the target. D 1992 Academic PESS, IX.
INTRODUCTION We have recently witnessed debate regarding the relative contributions of the cerebral hemispheres to the preparation for, and execution of, voluntary movements (e.g., Carson, 1989a, b; Peters, 1989). It has been predicted that right hemisphere contributions will be reflected primarily in measures sensitive to movement preparation (Carson, 1989a; Guiard, Diaz, & Beaubaton, 1983). Presumably, this is because the formation of some extant geometric representation of external space is thought to form the final step for the visual system and the initial step for the motor control system (Morass0 & Tagliasco, 1986). Evidence derived from clinThis research was supported by the Natural Sciences and Engineering Research Council of Canada. Reprint requests and correspondence concerning the article should be sent to Richard Carson, Motor Behaviour Laboratory, School of Kinesiology, Simon Fraser University, B.C., Canada V5A lS6. 227 0278-2626192 $5.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
228
CHUA
ET AL
ical populations documents the case for an enhanced role for the right hemisphere for spatial processing in general (DeRenzi, 1982) and for the preparatory stages of movement in particular (Fisk & Goodale, 1988; Haaland & Harrington, 1989). A number of investigators have elicited left-hand advantages for reaction time in manual aiming tasks in normals (Bradshaw, Bradshaw, & Nettleton, 1990; Carson, Chua, Elliott, & Goodman, 1990; Haaland & Harrington, 1989). These asymmetries are suggestive of an increased contribution of the right hemisphere during movement preparation. It is not clear, however, whether the expression of this effect indicates a right hemisphere advantage for the spatial localization of targets or for the spatial parameterization of the movements themselves. We may investigate the former possibility by manipulating the complexity of the spatial processing required to establish the position of the target. If the left-hand advantage for reaction time reduces to an advantage in localizing the target, we might anticipate that the extent of this asymmetry would covary with the complexity of the spatial processing which subserves localization. Carson, Goodman, and Elliott (1992) utilized a reaching task for which targets were not explicitly defined, but were obtained by extrapolating from briefly displayed geometric patterns. Spatial complexity was manipulated by altering the complexity of these patterns. However, Carson et al. (1992) failed to obtain an interaction of response hand with pattern complexity. This would appear to indicate that the degree of right hemisphere involvement does not correlate well with the complexity of the spatial processing necessary to establish the position of the target. However, their manipulation of spatial complexity included a number of possible confounds. The pattern presentation sequence had a strong temporal component. It is therefore possible that the task favored the processing propensities of both hemispheres. Furthermore, subjects were required to distinguish among four possible spatial configurations. Therefore, the task involved a categorization component in addition to the extraction of metric spatial information. Kosslyn (1987) has suggested that these mechanisms are independently lateralized. The left hemisphere, it is proposed, demonstrates greater potential for categorization, while the right hemisphere is predicted to deal predominantly with metric spatial information. The primary objective of the present study was to design a task which would overcome these limitations and permit us to determine the relative contribution of the right hemisphere to the spatial localization of targets prior to movement. To this end we designed an experiment in which subjects were required to make a response movement in a direction other than straight to the stimulus (cf. Georgopoulous & Massey, 1987). There is evidence to suggest that tasks of this nature may be “solved” by “a mental rotation of an imagined movement vector from the direction of the stimulus to the direction of the actual movement” (Georgopoulous & Massey, 1987, p.234).
SPATIAL LOCALIZATION
229
It has been suggested that the spatial components of mental rotation tasks are lateralized to a greater degree to the right hemisphere (e.g., Corballis & McLaren, 1984; Ratcliffe, 1979). These suspicions have been supported by “direct” physiological measures (Deutsch, Bourbon, Papanicolau, & Eisenberg, 1988; Osaka, 1984; Papanicolau, Deutsch, Bourbon, Will, Loring, & Eisenberg, 1987). If the left-hand advantage for reaction time reflects an enhanced ability of the right hemisphere to establish the position of a target prior to the initiation of a movement, we might anticipate that this advantage would be greatest in conditions in which the difficulty of establishing the target position was increased. Specifically, we might expect to see greater right hemisphere advantages when subjects are required to perform a mental rotation to establish the position of a target relative to a control condition in which the target is presented directly. Therefore, in this study, subjects made aiming movements in two conditions. In the first (mirror) condition, a stimulus was presented in either the left or the right visual field. Subjects were required to perform a transformation, or mental rotation, of the stimulus position about a virtual line dividing the left and right visual fields. The spatial position which resulted from this transformation constituted the target. The target was thus the mirror image of the stimulus position. Subjects were required to move to this target position. In a second (control) condition, subjects were required to move directly to the stimulus position. METHODS Subjects. Subjects were eight experiment-naive male volunteers. All subjects were right-
handed (Oldfield, 1971) and had normal or corrected to normal vision. Apparatus for data collection. Subjects were seated facing a 50-by-50-cm display panel, constructed such that a 21-by-21 array of LEDs (centers spaced by 2 cm in the horizontal and vertical directions) was not normally visible behind a translucent hardened glass screen. Activation of single LEDs (target lights) resulted in a projection which was viewed by the subject as a point source of light. The central light of the array was the fixation point, located 50 cm directly ahead of the subject at eye level. The display panel was composed of an Interaction Systems Series 4000 touch-sensitive screen interfaced with a microcomputer such that points of contact of a subject’s finger upon the surface were registered as twodimensional coordinates. In the configuration used, the screen had a horizontal resolution of < 1.20 mm and a vertical resolution of < 1.65 mm. Prior to all trials the subject’s finger was placed upon a microswitch located on a virtual line between the center of the display panel and the subject’s midline, 40 cm from the surface of the panel and 42 cm below the fixation point. The microswitch and LED array were interfaced with the controlling microcomputer. Ambient illumination was provided by a custom-modified single fluorescent lamp.’ A control device allowing almost instantaneous offset of the lamp (~25 ms) was triggered under microcomputer control, thereby allowing for the lamp to be extinguished upon movement initiation. Twelve target positions were employed, six to each side of the vertical midline of the ’ Goodman, D., Keogh, P., Carson, R. G., & Elliott, D. (In preparation), a simple modification for instantaneous offset of fluorescent lamps,
230
CHUA ET AL. 0
1
2
3
4
5
6
7
8
9
1011121314151617181920
0 1
......
2 3 4 5 6 7 a 9
...... ...... ...... ......
0 i
0 il
.a..
0.0 ......
......
4
5
B..
10 11 12 13
. . .
.O’O
14 15
10, ! 1: j2:
16 17
:
:
.
18 19
20
_ .
. .
.
,
,
I
FIG. 1. Schematic of the target display panel. In the mirror aiming condition, for movements made into the left field, upon illumination of positions 4, 5, 6, 10, 11, or 12, subjects moved to positions 3, 2, 1, 9, 8, or 7, respectively. The reverse applied for movements into the right field. In the control condition, subjects moved directly to the illuminated target positions.
panel (Fig. 1). Each target could be specified in terms of field, eccentricity, and midpoint. These factors designated side (left, right), horizontal distance from fixation (inner, middle, outer), and vertical distance from fixation (above, below), respectively. Inner targets were 60 mm eccentric from the vertical midline, representing 6.87” of visual angle; middle targets were 100 mm eccentric (11.46”), while outer targets were 140 mm eccentric (16.04”). With respect to the relation to the midpoint, targets were 9.17” of visual angle (80 mm) above or below the fixation point. Procedure. A repeated measures (2 x 2 x 2 x 2 x 3) factorial design was utilized. The five independent variables were Hand (left or right), Aiming Condition (mirror or control), Movement Field (left or right), position of the target in Relation to the Midpoint (lower or upper), and Target Eccentricity (inner, middle, or outer). Each subject underwent two experimental sessionswhich were generally held on successive days. Each session consisted of one aiming condition. The order in which each aiming condition occurred was counterbalanced across subjects. Test sessions consisted of four blocks (in random order) of 60 trials, two blocks with each hand. Each target position (also in random order) appeared on five occasions within a block. Each session was preceded by two blocks of 12 practice trials, one block being made with each hand.’ ’ A software random-number generator (seeded prior to each trial block) was used to generate randomization of trial conditions and of foreperiod durations. The sequences were truly random in that it was possible to have several consecutive trials to the same target location. The order of hand blocks was randomized for each subject by drawing an identifier for each of the four blocks from a “hat.” Practice blocks consisted of a single trial at each target position. The order of hand use in practice blocks was counterbalanced across subjects.
231
SPATIAL LOCALIZATION TABLE 1
REACTION TIME, MOVEMENTTIME, AND RADIAL ERRORAS A FUNCTIONOF HAND AND
AIMING CONDITION Mirror
Left Right
Control
RT
MT
RE
RT
MT
RE
314.4 321.7
483.6 479.6
26.2 23.8
283.8 294.7
464.4 440.8
18.5 17.0
Note. Reaction time and movement time measured in msec; radial error measured in mm. Subjects were informed of the aiming condition prior to each session and of the hand prior to each block. They were instructed to react as quickly as possible when the “target” was presented and to use the index finger to move to the defined target position as accurately as possible. All trials began with a microcomputer-generated tone. Upon subsequent closure of the microswitch, a warning tone was followed by illumination of the central fixation light. Subjects were required to fixate upon this position until, after a variable interval (of between 500 and 3000 msec) the fixation was extinguished. Simultaneously the “target” LED was illuminated for 200 msec. Upon inititation of a response movement, ambient illumination was removed, thus eliminating external visual cues that subjects could use during the movement. A trial was terminated when the subject completed the movement to the display panel, at which point ambient illumination was again provided. There was a break of approximately 5 set between trials. For each trial Radial Error (movement accuracy) was calculated as the absolute distance between the calibrated target position and the point of initial contact with the display panel. Reaction Time, assessed as being the time from the onset of the target until movement initiation, and Movement Time, the time between movement initiation and contact with the surface of the screen, were also calculated. Precise millisecond timing was achieved through the use of the 1-kHz clock on a LabMaster data acquisition board (Scientific Solutions, Inc.). Custom code for the control of display onsets and the recording of critical events was written in Turbo C using LabPac (Scientific Solutions Inc.) library routines which permit direct control over the LabMaster board. The status of the touchscreen was sampled “continuously.” In the configuration used, a “minimum contact time” of 1 msec was used, such that it was necessary for the subject’s finger to remain in contact with the screen for at least 1 msec in order that a contact was registered and the touch coordinates determined. The test session lasted approximately 45 min. Subjects were permitted rest periods between blocks, as required.
RESULTS AND DISCUSSION
Median Reaction Time, Movement Time, and Radial Error measures were obtained for cells formed from the combination of all factors, each median value being derived from 10 trials. The three dependent variables were analyzed separately using a 2 Hand by 2 Aiming Condition by 2 Movement Field by 3 Target Eccentricity by 2 Relation to Midpoint repeated measures analysis of variance. A summary of these data is presented in Table 1. The mean values shown are those obtained by collapsing the cell structure over the factors of Field, Eccentricity, and Relation to Midpoint. Although all factors were entered into the analyses of variance,
232
CHUA ET AL.
the hypotheses of interest make their most specific predictions in terms of Aiming Condition and Hand. However, statistically significant main effects and interactions involving Field, Eccentricity, and Relation to Midpoint are noted below. In this study, we were concerned with movement preparation and, in particular, with processes associated with the localization of target positions prior to movement initiation. Therefore, reaction time was the the dependent measure of primary interest. Analysis of reaction time data indicated a main effect for aiming condition, F(1, 7) = 14.43, p < 0.01. Movements were initiated more quickly in the target condition (3 = 289.3 msec) compared to the mirror condition (X = 318.0 msec). It appears that when a transformation of the visual stimulus was required to establish the position of the target, greater “processing” time was required. This is consistent with data obtained from studies dealing explicitly with processesof mental rotation (e.g., Cooper, 1976). Of interest was whether this increment was asymmetrical in nature. There was a main effect for Hand, F(1, 7) = 7.08, p < 0.05. Left hand movements (X = 299.1 msec) were initiated, on the average, 9 msec quicker than right hand movements (a = 308.2 msec). Th e advantage for the left hand is in line with data reported previously (Bradshaw et al., 1989; Carson et al., 1990; Carson, Goodman, Chua, & Elliott, (in press); Elliott, Roy, Goodman, Carson, Chua, & Maraj, (in press); Haaland & Harrington, 1989). Significantly, however, the magnitude of the left-hand advantage for reaction time did not increase in the mirror condition (see Table 1). This suggests that although there may be greater right hemisphere involvement prior to the initiation of aiming movements, the shorter reaction times exhibited by the left hand do not reflect an advantage of the right hemisphere in establishing the spatial position of a target. There was also a main effect for Relation to Midpoint, F(1, 7) = 8.17, p < 0.05. Movements to lower targets (X = 299.9 msec) were initiated more quickly than movements to upper targets (X = 307.4 msec). Analysis of movement time data revealed a tendency for movements in the mirror condition (X = 481.6 msec) to take longer than those in the control condition (Z = 452.6 msec), F(1, 7) = 5.41, p = 0.053. Although the hands were not differentiated on the basis of movement time, there was an interaction of hand with field, F(1, 7) = 96.75, p < 0.0001. Movements made by the left hand into the left field, and by the right hand into the right field, were of shorter duration than movements made contralaterally. This effect was mediated by target eccentricity, as revealed by a Hand by Field by Eccentricity interactions, F(1,7) = 20.26, p < 0.0001. Movements made into ipsilateral fields were of shorter duration when targets were more peripheral, whereas the reverse was true when movements were made into contralateral fields (cf. Carson et al., 1990). These data suggest the intrusion of biomechanical factors. There
SPATIAL
LOCALIZATION
233
was also a main effect for Midpoint, F(1,7) = 35.38, p < 0.001. Consistent with expectations, movements made to upper targets (Z = 477.4 msec) took longer to complete than movements to lower targets (X = 456.8 msec). Analysis of radial error data revealed a main effect for Aiming Condition, F(1, 7) = 9.05, p < 0.05, with movements being more accurate in the control condition (X = 17.8 mm) than in the mirror condition (X = 25.0 mm). There was also a trend toward a right-hand advantage which failed to reach conventional levels of statistical significance, F(1, 7) = 5.46, p = 0.06. However, there was an interaction of hand with Midpoint, F(1, 7) = 8.04, p < 0.05. Movements made by the left hand were less accurate when moving to upper targets (upper x = 23.4 mm, lower K = 21.2 mm), whereas movements made by the right hand (upper x = 20.4 mm, lower x = 20.4 mm) were not differentiated on the basis of the relation of the target to the midpoint (cf. Carson et al., 1990). Although movements made to targets in extrapersonal space may have an inherent spatial complexity such that greater right hemisphere involvement ensues, the results of the present study suggest that the complexity is not that of localizing the targets. Increasing the complexity of the spatial transformation required to localize the targets did not increase the extent of the left hand advantage for reaction time. It may be the case that the import of the right hemisphere for movement preparation pertains to other components of the task. These components may include those concerned with establishing the spatial parameters of the movement itself. Another possibility exists, based upon an assumption that the righthand movement system is more “precise” than the left-hand system and that there exists a computational variant of the speed-accuracy trade-off phenomena.3 If it requires longer to complete the computations necessary for a more precise movement, the observation that reaction times for the left hand are shorter than those for the right may actually provide no information concerning the processing capabilities of the respective handhemisphere systems. In order to assessthis proposal, an individual trials analysis was performed to determine, for each subject, the correlation between Reaction Time and Radial Error. Each correlation was based upon 10 pairs of values. These correlation values were collapsed over the factors of Movement Field, Relation to Midpoint, and Target Eccentricity. The corresponding z values were analyzed using a repeated measures (2 (Hand) x 2 (Aiming Condition)) factorial design analysis of variance. There were were no indications of main effects for Hand or Aiming Condition or of an interaction of these factors. In addition, the grand mean correlation assumed a value of r = -0.009. Thus, rather than there being a trade3 We are grateful to an anonymous reviewer
for bringing this possibility
to our attention.
234
CHUA ET AL.
off, reaction times are, on the average, simply faster for the left hand than for right hand. It also appears that the magnitude of the left-hand advantage for RT increases with the degree of prior uncertainty concerning the spatial position of a movement target (Carson, Chua, Goodman, & Elliott, submitted). When the number of target position alternatives is increased, thus diminishing the effectuality of advance preparation, the magnitude of the left-hand advantage for RT increases. When subjects are precued as to the exact spatial position in advance of stimulus presentation, and can thus complete some portion of movement preparation in advance, the left-hand advantage for reaction time is eliminated. In addition, the left-hand advantage does not appear to be simply due to differences in the speed with which cued parameters per se are integrated into a movement plan. Carson, Chua, Byblow, and Goodman (1991) required that subjects make their aiming responses in two further cue conditions. In one case, response hand was cued simultaneously with the presentation of a target stimulus. In the other case, hand was cued in advance of the stimulus. In the latter condition (analogous to the target condition in the present study) a left-hand advantage for RT was observed. In the former condition a right-hand advantage for RT was present. As in these studies, the various cue conditions were not distinguished by measures of response execution (movement time and radial error), it is suggested that the lefthand advantage for reaction time is specific to the organization of movement with respect to spatial parameters and is independent of relative or absolute movement precision. REFERENCES Bradshaw, J. L., Bradshaw, J. A., & Nettleton, N. C. 1990. Abduction, adduction and hand differences in simple and serial movements. Neuropsychologia, 28, 917-931. Carson, R. G. 1989a. Manual asymmetries: Feedback processing, output variability and spatial complexity: Resolving some inconsistencies. Journal of Motor Behavior, 21,3847. Carson, R. G. 1989b. Manual asymmetries: In defense of a multifactorial account. Journal of Motor Behavior, 21, 157-162. Carson, R. G., Chua, R., Byblow, W. D., & Goodman, D. 1991, October. Are left hand Paper preadvantages for movement preparation specific to spatial parameterization? sented at the Canadian Society for Psychomotor Learning and Sport Psychology, London, Ontario. Carson, R. G., Chua, R., Goodman, D., & Elliott, D. Submitted. Asymmetries in the preprogramming of aiming movements. Carson, R. G., Chua, R., Elliott, D., & Goodman, D. 1990. The contribution of vision to asymmetries in manual aiming. Neuropsychologia, 28, 1215-1220. Carson, R. G., Goodman, D. Chua, R., & Elliott, D. In press. Asymmetries in the regulation of visually guided aiming. Journal of Motor Behavior. Carson, R. G., & Goodman, D., & Elliott, D. 1992. Asymmetries in the discrete and pseudocontinuous regulation of visually guided reaching. Brain and Cognition, l&169191.
SPATIAL LOCALIZATION
235
Cooper, L. A. 1976. Demonstration of a mental analog of an external rotation. Perception and Psychophysics,
19, 296-302.
Corballis, M. C., & McLaren, R. 1984. Winding one’s Ps and Qs: Mental rotation and mirror image discrimination. Journal of Experimental Psychology: Human Perception and Performance,
10, 318-327.
DeRenzi, E. 1982. Disorders of space exploration and cognition. Chichester: Wiley. Deutsch, G., Bourbon, W. T., Papanicolaou, A. C., & Eisenberg, H. M. 1988. Visuospatial tasks compared via activation of regional cerebral blood flow. Neuropsychologia, 26, 445-452. Elliott, D., Roy, E. A., Goodman, D., Carson, R. G., Chua, R., & Maraj, B. K. V. In press. Asymmetries in the preparation and control of manual aiming movements. Canadian Journal of Psychology.
Fisk, J. D., & Goodale, M. A. 1988. The effects of unilateral brain damage on visually guided reaching: Hemispheric differences in the nature of the deficit. Experimental Brain Research, 72, 425-435.
Georgopoulous, A. P., & Massey, J. T. 1987. Cognitive spatial-motor processes. 1. The making of movements at various angles from a stimulus direction. Experimental Brain Research, 65, 361-370. Guiard, Y., Diaz, D., & Beaubaton, D. 1983. Left hand advantage for right handers for spatial constant error: Preliminary evidence in a unimanual ballistic aimed movement. Neuropyschologia, 21, 11l-l 15. Haaland, K. Y., & Harrington, D. 1989. The role of the hemispheres in closed loop movements. Brain and Cognition, 9, 158-180. Kosslyn, S. M. (1987). Seeing and imaging in the cerebral hemispheres: A computational approach. Psychological Review, 94, 148-175. Morasso, P., & Tagliasco, V. 1986. Human movement understanding: From computational geometry to artificial intelligence. Amsterdam: North Holland. Oldfield, R. C. 1971. The assessmentand analysis of handedness: The Edinburgh inventory. Neuropsychologia, 9, 97- 113. Osaka, M. 1984. Peak alpha frequency of EEG during a mental task: Task difficulty and hemispheric differences. Psychophysiology, 21, 101-105. Papanicolau, A. C., Deutsch, G., Bourbon, W. T., Will, K. W., Loring, D. W., & Eisenberg, H. M. 1987. Convergent evoked potential and cerbral blood flow evidence of taskspecific hemispheric differences. Electroencephalography and Clinical Neuropsychology, 66, 515-520. Peters, M. 1989. Do feedback processing, output variability, and spatial complexity account for manual asymmetries? Journal of Motor Behavior, 21, 151-155. Ratcliffe, G. 1979. Spatial thought, mental rotation and the right cerebral hemisphere. Neuropsychologia,
17, 49-54.
AND
COGNITION
u),
227-235 (1992)
Asymmetries in the Spatial Localization of Transformed Targets ROMEO CHUA, RICHARD Motor Behaviour
Laboratory,
G. CARSON, AND DAVID GOODMAN
School of Kinesiology,
Simon Fraser University
AND DIGBY ELLIOTT Department of Physical Education,
McMaster
University
This study was designed to examine the contribution of the right cerebral hemisphere in the spatial localization of visual targets for manual aiming. Visual targets were briefly presented to the right and left fields and subjects were required to point either to the target location, or a “mirror” image of the target location with their right or left index finger. Whereas reaction times were faster for lefthand pointing than for right-hand pointing, there was no differential effect of the mirror image transformation. This suggeststhat left-hand reaction time advantages are more related to right hemisphere involvement in the spatial parameterization of the movement than spatial localization of the target. D 1992 Academic PESS, IX.
INTRODUCTION We have recently witnessed debate regarding the relative contributions of the cerebral hemispheres to the preparation for, and execution of, voluntary movements (e.g., Carson, 1989a, b; Peters, 1989). It has been predicted that right hemisphere contributions will be reflected primarily in measures sensitive to movement preparation (Carson, 1989a; Guiard, Diaz, & Beaubaton, 1983). Presumably, this is because the formation of some extant geometric representation of external space is thought to form the final step for the visual system and the initial step for the motor control system (Morass0 & Tagliasco, 1986). Evidence derived from clinThis research was supported by the Natural Sciences and Engineering Research Council of Canada. Reprint requests and correspondence concerning the article should be sent to Richard Carson, Motor Behaviour Laboratory, School of Kinesiology, Simon Fraser University, B.C., Canada V5A lS6. 227 0278-2626192 $5.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
228
CHUA
ET AL
ical populations documents the case for an enhanced role for the right hemisphere for spatial processing in general (DeRenzi, 1982) and for the preparatory stages of movement in particular (Fisk & Goodale, 1988; Haaland & Harrington, 1989). A number of investigators have elicited left-hand advantages for reaction time in manual aiming tasks in normals (Bradshaw, Bradshaw, & Nettleton, 1990; Carson, Chua, Elliott, & Goodman, 1990; Haaland & Harrington, 1989). These asymmetries are suggestive of an increased contribution of the right hemisphere during movement preparation. It is not clear, however, whether the expression of this effect indicates a right hemisphere advantage for the spatial localization of targets or for the spatial parameterization of the movements themselves. We may investigate the former possibility by manipulating the complexity of the spatial processing required to establish the position of the target. If the left-hand advantage for reaction time reduces to an advantage in localizing the target, we might anticipate that the extent of this asymmetry would covary with the complexity of the spatial processing which subserves localization. Carson, Goodman, and Elliott (1992) utilized a reaching task for which targets were not explicitly defined, but were obtained by extrapolating from briefly displayed geometric patterns. Spatial complexity was manipulated by altering the complexity of these patterns. However, Carson et al. (1992) failed to obtain an interaction of response hand with pattern complexity. This would appear to indicate that the degree of right hemisphere involvement does not correlate well with the complexity of the spatial processing necessary to establish the position of the target. However, their manipulation of spatial complexity included a number of possible confounds. The pattern presentation sequence had a strong temporal component. It is therefore possible that the task favored the processing propensities of both hemispheres. Furthermore, subjects were required to distinguish among four possible spatial configurations. Therefore, the task involved a categorization component in addition to the extraction of metric spatial information. Kosslyn (1987) has suggested that these mechanisms are independently lateralized. The left hemisphere, it is proposed, demonstrates greater potential for categorization, while the right hemisphere is predicted to deal predominantly with metric spatial information. The primary objective of the present study was to design a task which would overcome these limitations and permit us to determine the relative contribution of the right hemisphere to the spatial localization of targets prior to movement. To this end we designed an experiment in which subjects were required to make a response movement in a direction other than straight to the stimulus (cf. Georgopoulous & Massey, 1987). There is evidence to suggest that tasks of this nature may be “solved” by “a mental rotation of an imagined movement vector from the direction of the stimulus to the direction of the actual movement” (Georgopoulous & Massey, 1987, p.234).
SPATIAL LOCALIZATION
229
It has been suggested that the spatial components of mental rotation tasks are lateralized to a greater degree to the right hemisphere (e.g., Corballis & McLaren, 1984; Ratcliffe, 1979). These suspicions have been supported by “direct” physiological measures (Deutsch, Bourbon, Papanicolau, & Eisenberg, 1988; Osaka, 1984; Papanicolau, Deutsch, Bourbon, Will, Loring, & Eisenberg, 1987). If the left-hand advantage for reaction time reflects an enhanced ability of the right hemisphere to establish the position of a target prior to the initiation of a movement, we might anticipate that this advantage would be greatest in conditions in which the difficulty of establishing the target position was increased. Specifically, we might expect to see greater right hemisphere advantages when subjects are required to perform a mental rotation to establish the position of a target relative to a control condition in which the target is presented directly. Therefore, in this study, subjects made aiming movements in two conditions. In the first (mirror) condition, a stimulus was presented in either the left or the right visual field. Subjects were required to perform a transformation, or mental rotation, of the stimulus position about a virtual line dividing the left and right visual fields. The spatial position which resulted from this transformation constituted the target. The target was thus the mirror image of the stimulus position. Subjects were required to move to this target position. In a second (control) condition, subjects were required to move directly to the stimulus position. METHODS Subjects. Subjects were eight experiment-naive male volunteers. All subjects were right-
handed (Oldfield, 1971) and had normal or corrected to normal vision. Apparatus for data collection. Subjects were seated facing a 50-by-50-cm display panel, constructed such that a 21-by-21 array of LEDs (centers spaced by 2 cm in the horizontal and vertical directions) was not normally visible behind a translucent hardened glass screen. Activation of single LEDs (target lights) resulted in a projection which was viewed by the subject as a point source of light. The central light of the array was the fixation point, located 50 cm directly ahead of the subject at eye level. The display panel was composed of an Interaction Systems Series 4000 touch-sensitive screen interfaced with a microcomputer such that points of contact of a subject’s finger upon the surface were registered as twodimensional coordinates. In the configuration used, the screen had a horizontal resolution of < 1.20 mm and a vertical resolution of < 1.65 mm. Prior to all trials the subject’s finger was placed upon a microswitch located on a virtual line between the center of the display panel and the subject’s midline, 40 cm from the surface of the panel and 42 cm below the fixation point. The microswitch and LED array were interfaced with the controlling microcomputer. Ambient illumination was provided by a custom-modified single fluorescent lamp.’ A control device allowing almost instantaneous offset of the lamp (~25 ms) was triggered under microcomputer control, thereby allowing for the lamp to be extinguished upon movement initiation. Twelve target positions were employed, six to each side of the vertical midline of the ’ Goodman, D., Keogh, P., Carson, R. G., & Elliott, D. (In preparation), a simple modification for instantaneous offset of fluorescent lamps,
230
CHUA ET AL. 0
1
2
3
4
5
6
7
8
9
1011121314151617181920
0 1
......
2 3 4 5 6 7 a 9
...... ...... ...... ......
0 i
0 il
.a..
0.0 ......
......
4
5
B..
10 11 12 13
. . .
.O’O
14 15
10, ! 1: j2:
16 17
:
:
.
18 19
20
_ .
. .
.
,
,
I
FIG. 1. Schematic of the target display panel. In the mirror aiming condition, for movements made into the left field, upon illumination of positions 4, 5, 6, 10, 11, or 12, subjects moved to positions 3, 2, 1, 9, 8, or 7, respectively. The reverse applied for movements into the right field. In the control condition, subjects moved directly to the illuminated target positions.
panel (Fig. 1). Each target could be specified in terms of field, eccentricity, and midpoint. These factors designated side (left, right), horizontal distance from fixation (inner, middle, outer), and vertical distance from fixation (above, below), respectively. Inner targets were 60 mm eccentric from the vertical midline, representing 6.87” of visual angle; middle targets were 100 mm eccentric (11.46”), while outer targets were 140 mm eccentric (16.04”). With respect to the relation to the midpoint, targets were 9.17” of visual angle (80 mm) above or below the fixation point. Procedure. A repeated measures (2 x 2 x 2 x 2 x 3) factorial design was utilized. The five independent variables were Hand (left or right), Aiming Condition (mirror or control), Movement Field (left or right), position of the target in Relation to the Midpoint (lower or upper), and Target Eccentricity (inner, middle, or outer). Each subject underwent two experimental sessionswhich were generally held on successive days. Each session consisted of one aiming condition. The order in which each aiming condition occurred was counterbalanced across subjects. Test sessions consisted of four blocks (in random order) of 60 trials, two blocks with each hand. Each target position (also in random order) appeared on five occasions within a block. Each session was preceded by two blocks of 12 practice trials, one block being made with each hand.’ ’ A software random-number generator (seeded prior to each trial block) was used to generate randomization of trial conditions and of foreperiod durations. The sequences were truly random in that it was possible to have several consecutive trials to the same target location. The order of hand blocks was randomized for each subject by drawing an identifier for each of the four blocks from a “hat.” Practice blocks consisted of a single trial at each target position. The order of hand use in practice blocks was counterbalanced across subjects.
231
SPATIAL LOCALIZATION TABLE 1
REACTION TIME, MOVEMENTTIME, AND RADIAL ERRORAS A FUNCTIONOF HAND AND
AIMING CONDITION Mirror
Left Right
Control
RT
MT
RE
RT
MT
RE
314.4 321.7
483.6 479.6
26.2 23.8
283.8 294.7
464.4 440.8
18.5 17.0
Note. Reaction time and movement time measured in msec; radial error measured in mm. Subjects were informed of the aiming condition prior to each session and of the hand prior to each block. They were instructed to react as quickly as possible when the “target” was presented and to use the index finger to move to the defined target position as accurately as possible. All trials began with a microcomputer-generated tone. Upon subsequent closure of the microswitch, a warning tone was followed by illumination of the central fixation light. Subjects were required to fixate upon this position until, after a variable interval (of between 500 and 3000 msec) the fixation was extinguished. Simultaneously the “target” LED was illuminated for 200 msec. Upon inititation of a response movement, ambient illumination was removed, thus eliminating external visual cues that subjects could use during the movement. A trial was terminated when the subject completed the movement to the display panel, at which point ambient illumination was again provided. There was a break of approximately 5 set between trials. For each trial Radial Error (movement accuracy) was calculated as the absolute distance between the calibrated target position and the point of initial contact with the display panel. Reaction Time, assessed as being the time from the onset of the target until movement initiation, and Movement Time, the time between movement initiation and contact with the surface of the screen, were also calculated. Precise millisecond timing was achieved through the use of the 1-kHz clock on a LabMaster data acquisition board (Scientific Solutions, Inc.). Custom code for the control of display onsets and the recording of critical events was written in Turbo C using LabPac (Scientific Solutions Inc.) library routines which permit direct control over the LabMaster board. The status of the touchscreen was sampled “continuously.” In the configuration used, a “minimum contact time” of 1 msec was used, such that it was necessary for the subject’s finger to remain in contact with the screen for at least 1 msec in order that a contact was registered and the touch coordinates determined. The test session lasted approximately 45 min. Subjects were permitted rest periods between blocks, as required.
RESULTS AND DISCUSSION
Median Reaction Time, Movement Time, and Radial Error measures were obtained for cells formed from the combination of all factors, each median value being derived from 10 trials. The three dependent variables were analyzed separately using a 2 Hand by 2 Aiming Condition by 2 Movement Field by 3 Target Eccentricity by 2 Relation to Midpoint repeated measures analysis of variance. A summary of these data is presented in Table 1. The mean values shown are those obtained by collapsing the cell structure over the factors of Field, Eccentricity, and Relation to Midpoint. Although all factors were entered into the analyses of variance,
232
CHUA ET AL.
the hypotheses of interest make their most specific predictions in terms of Aiming Condition and Hand. However, statistically significant main effects and interactions involving Field, Eccentricity, and Relation to Midpoint are noted below. In this study, we were concerned with movement preparation and, in particular, with processes associated with the localization of target positions prior to movement initiation. Therefore, reaction time was the the dependent measure of primary interest. Analysis of reaction time data indicated a main effect for aiming condition, F(1, 7) = 14.43, p < 0.01. Movements were initiated more quickly in the target condition (3 = 289.3 msec) compared to the mirror condition (X = 318.0 msec). It appears that when a transformation of the visual stimulus was required to establish the position of the target, greater “processing” time was required. This is consistent with data obtained from studies dealing explicitly with processesof mental rotation (e.g., Cooper, 1976). Of interest was whether this increment was asymmetrical in nature. There was a main effect for Hand, F(1, 7) = 7.08, p < 0.05. Left hand movements (X = 299.1 msec) were initiated, on the average, 9 msec quicker than right hand movements (a = 308.2 msec). Th e advantage for the left hand is in line with data reported previously (Bradshaw et al., 1989; Carson et al., 1990; Carson, Goodman, Chua, & Elliott, (in press); Elliott, Roy, Goodman, Carson, Chua, & Maraj, (in press); Haaland & Harrington, 1989). Significantly, however, the magnitude of the left-hand advantage for reaction time did not increase in the mirror condition (see Table 1). This suggests that although there may be greater right hemisphere involvement prior to the initiation of aiming movements, the shorter reaction times exhibited by the left hand do not reflect an advantage of the right hemisphere in establishing the spatial position of a target. There was also a main effect for Relation to Midpoint, F(1, 7) = 8.17, p < 0.05. Movements to lower targets (X = 299.9 msec) were initiated more quickly than movements to upper targets (X = 307.4 msec). Analysis of movement time data revealed a tendency for movements in the mirror condition (X = 481.6 msec) to take longer than those in the control condition (Z = 452.6 msec), F(1, 7) = 5.41, p = 0.053. Although the hands were not differentiated on the basis of movement time, there was an interaction of hand with field, F(1, 7) = 96.75, p < 0.0001. Movements made by the left hand into the left field, and by the right hand into the right field, were of shorter duration than movements made contralaterally. This effect was mediated by target eccentricity, as revealed by a Hand by Field by Eccentricity interactions, F(1,7) = 20.26, p < 0.0001. Movements made into ipsilateral fields were of shorter duration when targets were more peripheral, whereas the reverse was true when movements were made into contralateral fields (cf. Carson et al., 1990). These data suggest the intrusion of biomechanical factors. There
SPATIAL
LOCALIZATION
233
was also a main effect for Midpoint, F(1,7) = 35.38, p < 0.001. Consistent with expectations, movements made to upper targets (Z = 477.4 msec) took longer to complete than movements to lower targets (X = 456.8 msec). Analysis of radial error data revealed a main effect for Aiming Condition, F(1, 7) = 9.05, p < 0.05, with movements being more accurate in the control condition (X = 17.8 mm) than in the mirror condition (X = 25.0 mm). There was also a trend toward a right-hand advantage which failed to reach conventional levels of statistical significance, F(1, 7) = 5.46, p = 0.06. However, there was an interaction of hand with Midpoint, F(1, 7) = 8.04, p < 0.05. Movements made by the left hand were less accurate when moving to upper targets (upper x = 23.4 mm, lower K = 21.2 mm), whereas movements made by the right hand (upper x = 20.4 mm, lower x = 20.4 mm) were not differentiated on the basis of the relation of the target to the midpoint (cf. Carson et al., 1990). Although movements made to targets in extrapersonal space may have an inherent spatial complexity such that greater right hemisphere involvement ensues, the results of the present study suggest that the complexity is not that of localizing the targets. Increasing the complexity of the spatial transformation required to localize the targets did not increase the extent of the left hand advantage for reaction time. It may be the case that the import of the right hemisphere for movement preparation pertains to other components of the task. These components may include those concerned with establishing the spatial parameters of the movement itself. Another possibility exists, based upon an assumption that the righthand movement system is more “precise” than the left-hand system and that there exists a computational variant of the speed-accuracy trade-off phenomena.3 If it requires longer to complete the computations necessary for a more precise movement, the observation that reaction times for the left hand are shorter than those for the right may actually provide no information concerning the processing capabilities of the respective handhemisphere systems. In order to assessthis proposal, an individual trials analysis was performed to determine, for each subject, the correlation between Reaction Time and Radial Error. Each correlation was based upon 10 pairs of values. These correlation values were collapsed over the factors of Movement Field, Relation to Midpoint, and Target Eccentricity. The corresponding z values were analyzed using a repeated measures (2 (Hand) x 2 (Aiming Condition)) factorial design analysis of variance. There were were no indications of main effects for Hand or Aiming Condition or of an interaction of these factors. In addition, the grand mean correlation assumed a value of r = -0.009. Thus, rather than there being a trade3 We are grateful to an anonymous reviewer
for bringing this possibility
to our attention.
234
CHUA ET AL.
off, reaction times are, on the average, simply faster for the left hand than for right hand. It also appears that the magnitude of the left-hand advantage for RT increases with the degree of prior uncertainty concerning the spatial position of a movement target (Carson, Chua, Goodman, & Elliott, submitted). When the number of target position alternatives is increased, thus diminishing the effectuality of advance preparation, the magnitude of the left-hand advantage for RT increases. When subjects are precued as to the exact spatial position in advance of stimulus presentation, and can thus complete some portion of movement preparation in advance, the left-hand advantage for reaction time is eliminated. In addition, the left-hand advantage does not appear to be simply due to differences in the speed with which cued parameters per se are integrated into a movement plan. Carson, Chua, Byblow, and Goodman (1991) required that subjects make their aiming responses in two further cue conditions. In one case, response hand was cued simultaneously with the presentation of a target stimulus. In the other case, hand was cued in advance of the stimulus. In the latter condition (analogous to the target condition in the present study) a left-hand advantage for RT was observed. In the former condition a right-hand advantage for RT was present. As in these studies, the various cue conditions were not distinguished by measures of response execution (movement time and radial error), it is suggested that the lefthand advantage for reaction time is specific to the organization of movement with respect to spatial parameters and is independent of relative or absolute movement precision. REFERENCES Bradshaw, J. L., Bradshaw, J. A., & Nettleton, N. C. 1990. Abduction, adduction and hand differences in simple and serial movements. Neuropsychologia, 28, 917-931. Carson, R. G. 1989a. Manual asymmetries: Feedback processing, output variability and spatial complexity: Resolving some inconsistencies. Journal of Motor Behavior, 21,3847. Carson, R. G. 1989b. Manual asymmetries: In defense of a multifactorial account. Journal of Motor Behavior, 21, 157-162. Carson, R. G., Chua, R., Byblow, W. D., & Goodman, D. 1991, October. Are left hand Paper preadvantages for movement preparation specific to spatial parameterization? sented at the Canadian Society for Psychomotor Learning and Sport Psychology, London, Ontario. Carson, R. G., Chua, R., Goodman, D., & Elliott, D. Submitted. Asymmetries in the preprogramming of aiming movements. Carson, R. G., Chua, R., Elliott, D., & Goodman, D. 1990. The contribution of vision to asymmetries in manual aiming. Neuropsychologia, 28, 1215-1220. Carson, R. G., Goodman, D. Chua, R., & Elliott, D. In press. Asymmetries in the regulation of visually guided aiming. Journal of Motor Behavior. Carson, R. G., & Goodman, D., & Elliott, D. 1992. Asymmetries in the discrete and pseudocontinuous regulation of visually guided reaching. Brain and Cognition, l&169191.
SPATIAL LOCALIZATION
235
Cooper, L. A. 1976. Demonstration of a mental analog of an external rotation. Perception and Psychophysics,
19, 296-302.
Corballis, M. C., & McLaren, R. 1984. Winding one’s Ps and Qs: Mental rotation and mirror image discrimination. Journal of Experimental Psychology: Human Perception and Performance,
10, 318-327.
DeRenzi, E. 1982. Disorders of space exploration and cognition. Chichester: Wiley. Deutsch, G., Bourbon, W. T., Papanicolaou, A. C., & Eisenberg, H. M. 1988. Visuospatial tasks compared via activation of regional cerebral blood flow. Neuropsychologia, 26, 445-452. Elliott, D., Roy, E. A., Goodman, D., Carson, R. G., Chua, R., & Maraj, B. K. V. In press. Asymmetries in the preparation and control of manual aiming movements. Canadian Journal of Psychology.
Fisk, J. D., & Goodale, M. A. 1988. The effects of unilateral brain damage on visually guided reaching: Hemispheric differences in the nature of the deficit. Experimental Brain Research, 72, 425-435.
Georgopoulous, A. P., & Massey, J. T. 1987. Cognitive spatial-motor processes. 1. The making of movements at various angles from a stimulus direction. Experimental Brain Research, 65, 361-370. Guiard, Y., Diaz, D., & Beaubaton, D. 1983. Left hand advantage for right handers for spatial constant error: Preliminary evidence in a unimanual ballistic aimed movement. Neuropyschologia, 21, 11l-l 15. Haaland, K. Y., & Harrington, D. 1989. The role of the hemispheres in closed loop movements. Brain and Cognition, 9, 158-180. Kosslyn, S. M. (1987). Seeing and imaging in the cerebral hemispheres: A computational approach. Psychological Review, 94, 148-175. Morasso, P., & Tagliasco, V. 1986. Human movement understanding: From computational geometry to artificial intelligence. Amsterdam: North Holland. Oldfield, R. C. 1971. The assessmentand analysis of handedness: The Edinburgh inventory. Neuropsychologia, 9, 97- 113. Osaka, M. 1984. Peak alpha frequency of EEG during a mental task: Task difficulty and hemispheric differences. Psychophysiology, 21, 101-105. Papanicolau, A. C., Deutsch, G., Bourbon, W. T., Will, K. W., Loring, D. W., & Eisenberg, H. M. 1987. Convergent evoked potential and cerbral blood flow evidence of taskspecific hemispheric differences. Electroencephalography and Clinical Neuropsychology, 66, 515-520. Peters, M. 1989. Do feedback processing, output variability, and spatial complexity account for manual asymmetries? Journal of Motor Behavior, 21, 151-155. Ratcliffe, G. 1979. Spatial thought, mental rotation and the right cerebral hemisphere. Neuropsychologia,
17, 49-54.
