Running Head: ARM SELECTION AND DYNAMIC COORDINATION 1 2 3

4 Hemispheric Differences in the Control of Limb Dynamics: 5

A Link Between Arm Performance Asymmetries and Arm Selection Patterns 6 7 8 9

Chase J. Coelho1,2, Andrzej Przybyla1,3, Vivek Yadav1, Robert L. Sainburg1,3,4 10 11 12 13 Departments of Kinesiology1, Psychology2, and Neurology3,Pennsylvania State University 14 Huck Life Sciences Institute4, Pennsylvania State University 15 16 17 18 19 20 21 Corresponding Author: 22 Robert L. Sainburg 23 29 Recreation Bldg 24 Pennsylvania State University 25 University Park, PA 16802 26 rls45@psu.edu 27 814-865-7938 28 This research was supported by the National Institute of Health grants R01HD39311 and 29 R01HD059783. 30 31

Articles in PresS. J Neurophysiol (November 14, 2012). doi:10.1152/jn.00885.2012

Copyright © 2012 by the American Physiological Society.

2

Abstract 32 Human handedness has been described and measured from two perspectives: Handedness inventories 33 rate hand preferences, while other tests examine motor performance asymmetries. These two 34 measurement approaches reflect a major controversy in a literature that defines handedness as either 35 a preference or as an asymmetry in sensorimotor processing. Over the past decade, our laboratory has 36 developed a model of handedness based on lateralization of neural processes. This model attributes 37 distinct control processes to each hemisphere, which in turn, lead to observable interlimb 38 sensorimotor performance asymmetries. We now hypothesize that arm preference, or choice, may 39 depend on the interaction between sensorimotor performance asymmetries and the given task. The 40 purpose of this study is to examine whether arm selection is linked to interlimb performance 41 asymmetries during reaching. Right-handers made choice and non-choice reaches to each of 8 targets 42 (d=3.5 cm) arranged radially (r = 13 cm) around a midline starting position. We displaced each cursor 43 (one associated with each hand) 30 cm to the midline start circle to ensure that there were no 44 hemispace-related geometric, mechanical, or perceptual biases to use either arm for the two midline 45 targets. The 3 targets on each side of the midline received mostly reaches from the ipsilateral arm, a 46 tendency previously described as a “hemispace bias.” However, the midline targets, which were 47 equidistant from each hand, received more dominant arm reaches. Dominant arm hand-paths to these 48 targets were straighter and more accurately directed. Inverse dynamics analyses revealed a more 49 proficient dominant arm strategy that exploited intersegmental dynamics to a greater extent than did 50 the non-dominant arm. These findings suggest that sensorimotor asymmetries in dynamic coordination 51 might explain limb choices. We discuss the implications of these results for theories of action selection, 52 models of handedness, and models of neural lateralization. 53 Key words: Neural lateralization, handedness, hand selection, motor control 54 55

3

Introduction 56 Some of the earliest research on human brain lateralization emphasized a left-hemisphere dominance 57 for motor functions in most humans. For example, Broca (1865) described a left-hemisphere 58 specialization for processes that subserve speech and language, including speech motor control. 59 Liepmann (1905) showed that left-hemisphere damage tends to produce greater movement 60 impairment than does right-hemisphere damage, defining Apraxia as a key example of such 61 impairment; see Allen (1983) and Geschwind (1975) for reviews. However, Sperry and Gazzaniga’s 62 seminal research on split-brain patients provided definitive evidence that each hemisphere can be 63 dominant for different neurobehavioral processes; see Gazzaniga (2000a) for a review. While this 64 evidence inspired a departure from Liepmann’s view of a comprehensively dominant left hemisphere, 65 this change was only recently incorporated into theories of handedness (Carson 1993; Sainburg 2002). 66 The view that each hemisphere may be specialized for different aspects of motor control has led to the 67 understanding that unilateral hemisphere damage produces unique motor deficits that depend on the 68 side of the lesion (Mutha Sainburg and Haaland 2011; Haaland and Flaherty 1984; Mani et al. in press). 69 Perhaps more importantly, hemisphere-specific deficits also occur in the ipsilesional arm of stroke 70 patients, demonstrating that each hemisphere contributes different processes to control of each arm 71 (Schaefer Haaland and Sainburg 2007, 2009a). 72 The evidence that hemispheric asymmetries correspond to performance asymmetries in the arms 73 suggests that these asymmetries might give rise to handedness. However, handedness is most often 74 described and measured as a “preference” for performing select motor tasks with the dominant arm. 75 This view of handedness is supported by the fact that most tasks can be accomplished with either limb, 76 regardless of asymmetries in performance measures. In addition, one clearly can alter one’s arm 77 preference for performing a task under different environmental constraints. For example, one might 78 reach for a coffee cup with the left non-dominant arm if the table supporting the cup is situated on the 79 left side of one’s chair. It is therefore easy to understand why handedness has traditionally been 80 characterized as a preference (Bryden 1977; Oldfield 1971), rather than as an asymmetry in 81 performance (Carson 1993; Sainburg 2002). However, it remains unclear whether one’s choice to use a 82 particular hand in a given task results from performance asymmetries in motor control. 83 There has been surprisingly little previous research that has examined both arm selection and arm 84 performance together in the same study. By combining the results of studies on limb selection with 85

4

results from studies on limb performance, one might conclude that the two factors are, in fact, related. 86 For example, it has been shown repeatedly that right-handers tend to reach the nondominant limb to 87 targets on the left side of the workspace and the dominant limb to targets located on the right side 88 (Bryden Pryde and Roy 2000; Bryden and Roy 2006; Gabbard and Helbig 2004; Gonzalez and Goodale 89 2009; Mamolo et al. 2004, 2006; Peters 1995). In addition, studies of interlimb motor performance 90 have shown that reaches to the ipsilateral workspace often show advantages in reaction time, peak 91 velocity, duration, final-position accuracy, and movement trajectory relative to contralateral reaches 92 (Carey Hargreaves and Goodale, 1996; Carson et al. 1992, 1993; Chua Carson Goodman and Elliott 93 1992; van Der Staak 1975; Elliott et al. 1993; Fisk and Goodale 1985; Ingum and Bjorklund 1994; 94 Prablanc et al. 1979). Thus, the tendency to avoid reaching across midline with each arm might depend 95 on sensorimotor performance advantages for ipsilateral reaches. It should, however, be noted that the 96 mechanisms that drive these hemispace biases in both preference and performance are controversial. 97 Some researchers have argued that hemispace biases reflect cognitive effects of attention and 98 stimulus-response compatibility (Gabbard and Helbig, 2004; Gabbard Rabb and Gentry 1998; Hommel 99 1993; Verfaellie and Heilman 1990) -- a suggestion further supported by the finding that the dominant 100 hand receives more attention during bimanual reaches (Buckingham and Carey, 2009; Peters, 1981). 101 However, others have suggested that hemispace effects result from an intrahemispheric information 102 processing advantage, an argument based on the fact that visual stimuli in each hemispace are initially 103 processed in the hemisphere that controls the ipsilateral limb (Bradshaw Bradshaw and Nettleton 104 1990; Fisk and Goodale 1985). Thus, while different underlying processes might give rise to the 105 relations between limb selection patterns and limb performance patterns, we hypothesize that limb 106 selection asymmetries associated with handedness can predict limb performance asymmetries. 107 The goal of the current study is to directly test this hypothesis. We first assessed which arm people 108 chose to reach toward different targets and then examined whether those choices were associated 109 with performance asymmetries observed in the same participants. We controlled for hemispace biases 110 in the present study by providing right-handers the choice of limb to each of 8 targets in a center-out 111 reaching task that was symmetrically centered in the workspace. This design allowed us to control for 112 potential visual-spatial biases by assuring symmetry in midline start positions and in the 8 radially 113 arranged targets. There were three targets left of body-midline, two targets on the midline, and three 114 targets right of the midline. However, if both hands started in the same central location within such a 115 center-out design, each arm would be an obstacle for contralateral reaches with the other arm. We 116

5

eliminated this potential confound by presenting a virtual reality display in which cursors representing 117 left- and right-hand positions were displaced 30 centimeters to the center of the workspace. Because 118 these cursor displacements allowed each hand to remain in a symmetric configuration in its own 119 motor-hemispace, we assured biomechanical symmetry for both the start positions, and for the two 120 midline targets. 121

Method 122 Participants 123 Ten right-handed Penn State undergraduate students (7 males) participated for course credit. We used 124 a 13-item version of the Edinburgh Handedness Inventory (EHI; Oldfield, 1971) to confirm that all 125 participants were right-handed. The mean number of EHI items that participants endorsed as right-126 limb tasks was 12.70 (SD=.48). The mean height of the participants was 176.28 cm (SD=10.73 cm). 127 Their mean weight was 76.61 kg (SD=11.84 kg), and their mean age was 20.50 yrs (SD=1.41 yrs). All 128 participants signed an informed consent form approved by the Penn State Institutional Review Board. 129 Apparatus 130 Figure 1 (panel A) shows the experimental setup. Participants sat in a dentist-type chair and faced a 131 horizontal workspace. An air sled attached to each arm minimized the effects of friction and fatigue as 132 participants moved their arms across the workspace. A splint on each arm immobilized joints distal to 133 the elbow. A mirror, positioned above the workspace, reflected stimuli projected by an overhead 55” 134 high-definition television (Sony Electronics, Inc). The mirror prevented participants from seeing their 135 arms. The stimuli were displayed with custom software written in REAL BASIC (REAL Software, USA). 136 A six-degree-of-freedom (6-DOF) Flock of Birds (Ascension-Technology, Inc.) magnetic tracking system 137 sampled limb positions and orientations at a rate of 130 Hz. For motion tracking, we digitized the 138 following bony structures in each limb: (1) index fingertip; (2) metacarpal-phalangeal joint; (3) lateral 139 and medial epicondyle of the humerus; and (4) acromion process. Each arm had two 6-DOF sensors 140 securely affixed to it (see Figure 1). 141 Panel B of Figure 1 shows that a 1.5 cm diameter (d) circular cursor with crosshairs represented the 2-D 142 position of the index fingertip of each hand. The software used the two-dimensional position of the 143 finger to project and update the cursor position at a rate of 60 Hz refresh. Each cursor (d=1.5 cm) had a 144 medial displacement of 30 cm. The mirror reflected the following stimuli: (1) two cursors; (2) a 145

6

thermometer-style velocity feedback indicator located at the top of the workspace; (3) a green start 146 circle (d=2.5 cm) located at the body-midline; and (4) a bullseye-style target (d=3.5 cm) located 13 cm 147 away from the start circle. In each trial, the target appeared at one of 8 possible positions. A 45° 148 interval separated each possible target location. 149 Procedure 150 We tested participants in individual sessions. After each participant arrived at the laboratory, s/he 151 signed an informed consent form and filled out demographic and handedness forms. The experimenter 152 then set up the participant in the apparatus. The setup process lasted about ten minutes. The 153 experimenter gave the participant instructions and asked whether the participant had any questions. 154 The session began when the experimenter was confident that the participant understood the 155 procedure. The entire experiment took approximately one hour. 156 Experimental Design 157 We used a fully within-subjects design. In two choice conditions, participants could reach either hand 158 to the target. In four non-choice conditions, participants reached a specified hand to the target. In both 159 types of trials, participants were asked to reach the target as quickly and accurately as possible and to 160 minimize their corrective movements. Table 1 describes the progression of all the conditions. 161 Participants first received practice on 32 choice trials (4 choices to each of 8 targets) for familiarization 162 with the dissociation between the visual feedback of the cursors and the proprioceptive feedback of 163 the hands. To this end, we also encouraged participants to choose both the left and the right hands 164 during different practice trials. In the second choice-condition, they performed 80 more choice trials. In 165 each of the last four conditions, they made 32 non-choice reaches (4 reaches to each of 8 targets). We 166 blocked the non-choice trials on the basis of limb and fixed the order of all conditions; participants first 167 performed 32 non-choice trials with the nondominant limb and then 32 non-choice trials with 168 dominant limb. They repeated this sequence once. We randomized the order of target presentation 169 within each condition so that the target location was unpredictable and fixed that order for all 170 participants. 171 Experimental Task 172 In order to provide visual stimuli that were symmetric relative to the hand locations and the body 173 midline, we centered the 8-direction center-out task at the body-midline (see Figure 1, Panel B). By 174

7

displacing the cursor from the hand 30 cm toward the midline, the arms remained geometrically 175 symmetric at the starting position. This design also separated the hands in space so that they could not 176 interfere with each other during the task. In the choice conditions, we asked participants to reach only 177 one cursor to the displayed target in each trial. Each cursor remained visible throughout the trial. In 178 the non-choice conditions, the set-up was the same except that subjects were instructed to reach with 179 either the left or the right hand for a full block of trials. 180 In the choice conditions, each trial began after participants moved both cursors into the midline start 181 circle and kept them there for .7 seconds (s). Because the target for that trial appeared immediately 182 after the preceding trial, the target was visible while the participants positioned the cursors in the start 183 circle. After the participants held the cursors in the start circle for 700 milliseconds (ms), they received 184 an audiovisual go-cue consisting of a tone and color change of the start circle. 185 Participants had 1 s to complete each 13 cm-long reach. They completed all reaches well before this 1 s 186 time limitation expired. To help ensure maximum tangential hand velocities were similar across targets 187 and across participants, we instructed participants to move as quickly and accurately as possible. We 188 required them to monitor the velocity feedback indicator, which provided feedback of the maximum 189 velocity following each reach. To motivate participants to be both quick and accurate, we awarded 190 points for accuracy only when the maximum tangential velocity exceeded .8 m/s. We showed at the 191 top of the screen a running total of earned points. 192 Participants always had full visual feedback of the cursors. In addition, participants saw a brief 193 feedback display (2 second interval) of the hand-path following each reach. The hand-path display then 194 disappeared and participants began the next trial at their own pace. 195 The only difference between the choice trials and the non-choice trials was that each non-choice trial 196 began after participants moved only a single cursor into the midline start circle. 197 Data Processing and Analysis 198 We processed the data with custom programs written in IgorPro 6.0 (WaveMetrics, Inc). We low-pass 199 filtered the displacement data at 8 Hz with a 3rd-order dual-pass Butterworth filter prior to 200 differentiation to obtain velocity and acceleration profiles. Because there were minor oscillations of 201 the cursors in the start circle, we defined the start of each reach as the first minimum in tangential 202

8

velocity which was under 8% of the maximum velocity for that trial. Likewise, we defined the end of 203 each reach as the first minimum following peak velocity, which was below 8% of maximum velocity. 204 Limb Choice Analysis 205 To assess whether limb choices varied by workspace, we first arcsine-transformed the proportion of 206 right-hand reaches to each target. We did so because the variance of proportions derived from 207 binomial data depends on the values of those proportions. This dependence violates the homogeneity 208 of variance assumptions of traditional inferential statistics. By contrast, the variance of arcsine-209 transformed proportions is independent of the proportion values (Hogg and Craig, 1995). It should be 210 noted that the arcsine transformation yields the inverse sine of the square root of each proportion, 211 which is expressed in radians. Thus, some of the transformed values were greater than 1. 212 We then grouped targets according to their visual workspace location relative to the body midline and 213 pooled the transformed values accordingly (see Figure 2). Hereafter, the three targets left of the body-214 midline are simply called “left” targets, the two targets at the midline are called “midline” targets, and 215 the three targets right of the midline are called “right” targets. 216 Movement Kinematics Analysis 217 We report three kinematic measures to characterize performance: hand-path deviation from linearity, 218 final-position error, and maximum tangential hand velocity (Vmax). Previous work has shown that for 219 reaches with similar maximum velocities, linearity deviation and final-position error systematically vary 220 across the arms (Sainburg 2002; Sainburg and Kalakanis 2002; Wang and Sainburg 2007). 221 We defined hand-path deviation from linearity as the minor axis of the path divided by the major axis 222 of the path. The major axis was the longest distance between any two points on the hand-path, while 223 the minor axis was the longest distance between any 2 points in the hand-path, measured 224 perpendicular to the major axis. We computed final-position error as the Euclidean distance between 225 the cursor center (index of fingertip) at the end of the reach and the target center. 226 Movement Dynamics Analysis 227 Appendix A details the method we used to quantify joint torques. We modeled each limb as a set of 228 three rigid, planar links with 8 degrees of freedom (DOF; see Figure 1 of Appendix A). We then 229 partitioned the equations of motion into three components: (1) net torque, which equaled the sum of 230 muscle and interaction torque components; (2) interaction torque, which reflected the effects of the 231

9

motion of the other limb segments; and (3) muscle torque, which estimated the rotational forces from 232 muscle contraction. This segmentation of the equations of motion was inspired by Schneider and 233 Zernicke (1990), and has previously been reported by our group (Sainburg et al 1999). Here we focus 234 on the torques observed during the time interval between reach onset and Vmax because this interval 235 largely reflects the effects of movement planning processes (Latash 2007; Sainburg et al. 1999). 236 Statistical Analysis 237 We analyzed the transformed choice data using a one-way, repeated measures ANOVA with visual 238 workspace (left, midline, and right) as the factor. We performed planned comparisons between 239 nondominant and dominant limb distributions (for choice, kinematic, and dynamic analyses) with 240 paired-samples t-tests and Bonferonni-corrected the significance threshold for multiple comparisons. 241 Bonferonni-corrected alpha was defined as, αBon =α/n, where α was .05 and n was the number of 242 comparisons (Abdi 2007). 243 Non-choice data came from conditions in which participants made an equal number of reaches to each 244 target. We analyzed each dependent measure using a 2 X 2 X 2 repeated-measures ANOVA with the 245 following factors: limb (nondominant vs. dominant), target (90° vs. 270°), and non-choice condition 246 order (first vs. second). 247

Results 248 Limb Choice 249 Previous research has shown that right-handers tend to reach the nondominant limb to left targets and 250 the dominant limb to right targets (see Peters 1995 for review). We therefore expected our 251 participants to reach the dominant limb to left targets less often than to right targets. Indeed, a 252 repeated-measures ANOVA on the arcsine-transformed values revealed a significant effect of target 253 location on dominant reaches, F(2,18)=42.935, p<.001 (see Figure 2). The size of this effect, estimated by 254 partial eta-squared (ηp

2), was .827. Post-hoc tests confirmed that left targets received fewer dominant 255 reaches (M=.319, SE=.153) than right targets (M=1.549, SE=.014), t(9)=7.622, p<.0001. 256 Previous work has also shown that the tendency for people to reach each limb to its own hemispace is 257 asymmetric, such that people reach the dominant limb to targets in the nondominant (left) space more 258 often than they reach the nondominant limb to targets in the dominant (right) space (Bryden and Roy 259 2006; Bryden et al. 2000; Mamolo et al. 2004, 2006; Gabbard and Helbig, 2004; Peters, 1995). Figure 2 260

10

shows that our participants displayed a similar pattern. There was a trend for the rate of nondominant 261 reaches to left targets (M=1.252, SE=.153) to be lower than the rate of dominant reaches to right 262 targets, although this trend did not reach statistical significance (M=1.549, SE=.014), t(9)=2.044, p=.071. 263 The main aim of this study was to determine whether limb performance asymmetries can predict limb 264 choices in the absence of hemispace and proximity effects. Thus, we focused on the midline targets 265 because they were equidistant from each hand. Post-hoc tests showed that the midline targets 266 (M=1.239, SE=.075) received a larger proportion of dominant reaches than left targets, t(9)=5.74, 267 p<.001, but a lower proportion of dominant reaches than right targets, t(9)=4.186, p=.002. 268 If limb choices depended entirely on hemispace effects on performance and preference, then one 269 would expect equal rates of nondominant and dominant reaches to midline targets. We tested these 270 predictions by assessing whether the midline targets received more dominant reaches than would 271 have been expected by chance alone. A one-sample t-test showed that midline targets received an 272 above-chance number of dominant reaches, t(9)=6.061, p<.001. We observed this effect at both the 90° 273 target (M=1.374, SE=.074), t(9)=7.938, p<.001, and at the 270° target (M=1.104, SE=.095), t(9)=3.347, 274 p=.009 (αBon=.05/3=.017). 275 Finally, to ensure that these effects were not due to differences across targets in the same visual 276 workspace, we conducted seven pair-wise t-tests -- three for the left targets, one for the midline 277 targets, and three for the right targets. These tests confirmed that there were no significant differences 278 in dominant reaches across left (ts(9)<1.539, ps>.158) or right targets (ts(9)<1.500, ps>.168). There was a 279 trend suggesting that the 90° midline target received more dominant reaches (M=1.374, SE=.074) than 280 the 270° midline target (M=1.104, SE=.095), but this effect did not survive the Bonferonni-corrected 281 significance threshold (αBon=.05/7=.007), t(9)=3.347, p=.009. It should be noted that despite this trend, 282 both midline targets received an above-chance number of dominant reaches. 283 Kinematics 284 Having confirmed that left targets received mostly nondominant reaches and that right targets 285 received mostly dominant reaches, we asked whether limb performance asymmetries can explain the 286 preponderance of dominant reaches to midline targets. In order to address this issue, we compared 287 movements made under non-choice conditions with the left and right arms. The non-choice conditions 288 were necessary for drawing statistical comparisons, as they allowed us to have equal numbers of 289 observations in each condition of interest. 290

11

Below we first detail kinematic differences between nondominant and dominant reaches to each of 291 the two midline targets. We then relate these kinematic results to interlimb differences in dynamic 292 coordination. As seen in Figure 3 (panel A), there was a main effect of limb on linearity deviation, 293 F(1,9)=20.397, p=.001, ηp

2=.694; nondominant reaches (M=.145 au, SE=.022 au) displayed greater 294 deviation than dominant reaches (M=.061 au, SE=.007 au).1 There was also a main effect of order, 295 F(1,9)=12.035, p=.007, ηp

2=.572, such that linearity deviation was greater in the first non-choice 296 condition (M=.108 au, SE=.014 au) than in the second (M=.099 au, SE=.013 au). This order effect was 297 the same for each limb and for each target as there were no interactions (Fs<2.639, ps>.130), so we 298 collapsed across order and performed planned comparisons. These tests confirmed that nondominant 299 linearity deviation was greater (M=.142 au, SE=.015 au) than dominant linearity deviation (M=.070 au, 300 SE=.006 au) to the 90° target, t(9)=3.819, p=.004. The same pattern between nondominant linearity 301 deviation (M=.147 au, SE=.017 au) and dominant linearity deviation (M=.054 au, SE=.007 au) was seen 302 at the 270° target, t(9)=4.380, p=.002. 303 Figure 3 (panel B) shows a main effect of limb on final-position error, F(1,9)=8.640, p=.017, ηp

2=.490, 304 such that the nondominant limb (M=2.50 cm, SE=.20 cm) showed larger final-position errors than the 305 dominant limb (M=2.00 cm, SE=.20 cm). This is consistent with previous findings for reaching under 306 visual feedback conditions (Carson et al. 1990). There was a main effect of target, F(1,9)=5.515, p=.043, 307 ηp

2=.380; reaches to the 270° target (M=2.00 cm, SE=.20 cm) showed more final-position error than 308 reaches to the 90° target (M=2.5 cm, SE=.20 cm). There was no main effect of order, and there were no 309 interactions (Fs<3.278, ps>.100). 310 Planned comparisons showed no difference between nondominant final-position error (M=2.673, 311 SE=.273) and dominant final-position error (M=2.359, SE=.264) at the 270° target, t(9)=1.378, p=.201. At 312 the 90° target, however, the nondominant limb showed greater final-position error (M=2.269 cm, 313 SE=.185 cm) than the dominant limb (M=1.680 cm, SE=.164 cm), t(9)=5.163, p=.001. 314 Figure 3 (panel C) confirms that none of these interlimb differences could be attributed to differences 315 in maximum velocity. There was only a main effect of target on maximum velocity, F(1,9)=7.494, p=.023; 316 reaches were slower to the 90° target (M=.844 m/s, SE=.026 m/s) than to the 270° target (M=.875 m/s, 317 SE=.020 m/s). There were no other main effects, and there were no interactions (Fs<3.519, ps>.050). 318

12

Inverse Dynamics 319 The previous sections established that limb choice varied by workspace and that midline workspace 320 received mostly dominant reaches. They also showed that the dominant limb produced straighter 321 hand-paths to midline targets relative to the nondominant limb. By contrast, it was not clear that 322 interlimb differences in final-position accuracy could explain reaches to both midline targets. 323 These results indicate that asymmetries in rates of nondominant and dominant reaches to midline 324 targets might be attributable to limb performance asymmetries. More specifically, the preponderance 325 of dominant midline target reaches might be attributable to a dominant-limb coordination advantage. 326 This notion is consistent with previous work showing that the dominant limb/hemisphere system 327 appears specialized for the coordination of intersegmental dynamics (Sainburg and Kalakanis 2000; 328 Sainburg 2002). Thus, we now examine how the observed asymmetries in linearity deviation related to 329 interlimb differences in dynamic control. 330 Panels C and D of Figure 4 show representative shoulder and elbow joint torques, respectively. Two 331 torque components contributed to the shoulder (or elbow) net torque: muscle torque and interaction 332 torque (see Method). The torque profiles in Figure 4 show that each limb produced similar peak net 333 torque magnitudes at both the elbow and shoulder. For both limbs, these magnitudes were higher at 334 the shoulders than at the elbows (see y-axis scales) because the shoulders carry greater inertial loads. 335 Overall, there were no significant interlimb differences between absolute peak net torques at the 336 shoulders (Fs<4.504, ps>.050). There was only a statistical trend for a main effect of limb, F(1,9)=4.503, 337 p=.063. The dominant limb tended to display slightly greater peak net torques (M=13.229, SE=1.102) 338 compared to the nondominant limb (M=11.594, SE=1.167). However, this effect did not reach 339 traditional levels of statistical significance. There were also no significant interlimb differences 340 between absolute net torques at the elbows (Fs<2.508, ps>.147).2 341 Although initial peak net torques were similar across the limbs, closer inspection of peak muscle and 342 peak interaction torques revealed substantial interlimb differences in dynamic coordination. Panel C of 343 Figure 4 shows that the nondominant arm generated smaller shoulder peak muscle torques (M=6.635 344 N.m, SE=.634 N.m.) than the dominant arm (M=8.838 N.m, SE=.850 N.m.). Across subjects, this 345 difference was borne out as a main effect of limb, F(1,9)=35.425, p<.001, ηp

2=.797. There were no other 346 main effects and or interactions (Fs<4.462, ps>.050). 347

13

Figure 4 (panels C and D) shows that shoulder flexion in each limb was accompanied by elbow 348 extension as the hand moved toward the target. However, interaction and muscle torque made 349 different contributions to net torque at the shoulder and at the elbow in each limb. At the 350 nondominant shoulder, interaction and muscle torque made similar contributions to net torque. By 351 contrast, interaction torque at the dominant shoulder contributed less to net torque and muscle 352 torque contributed more. 353 These asymmetries in shoulder torque components were associated with clear differences in elbow 354 torque components. As shown in Figure 4 (panel D), nondominant elbow muscle torque made a 355 substantial contribution to net torque. By contrast, dominant elbow muscle torque made very little 356 contribution to net torque, as interaction torque arising from motion of the upper arm accounted for 357 most of the net torque and remained near zero throughout the trial. As a result, the dominant elbow 358 displayed very little extensor muscle torque while driving the hand toward the target. Indeed, there 359 was a main effect of limb on peak elbow muscle torque, F(1,9)=10.082, p=.011, ηp

2=.582, such that 360 nondominant reaches showed higher peak elbow muscle torques (M=1.618 N.m., SE=.228 N.m.) than 361 dominant reaches (M=1.002 N.m., SE=.091 N.m.). There was also a main effect of target, F(1,9)=40.194, 362 p<.001, ηp

2=.817; participants generated less elbow muscle torque to reach to the 90° target (M=1.105 363 N.m., SE=.136 N.m) than to the 270° target (M=1.515 N.m., SE=.158 N.m.). There were no other main 364 effects, and there were no interactions (Fs<3.580, ps>.050). 365 These results indicate that the coordination patterns displayed by the nondominant and dominant 366 limbs were different. Relative to the dominant limb, coordination of the nondominant limb was 367 characterized by a strategy that generated smaller peak muscle torques at the shoulder and greater 368 peak muscle torques at the elbow. By contrast, coordination of the dominant limb was characterized 369 by greater peak muscle torques at the shoulder, which produced elbow interaction torques that were 370 efficiently coordinated with small elbow muscle torques. 371 These results are consistent with our previous findings that the movement of the shoulder joint creates 372 large interaction torques at the elbow that are proficiently used to reduce muscle torque requirements 373 at the elbow (Kalakanis and Sainburg 2000). However, nondominant arm movements in the current 374 study were characterized by greater elbow and lower shoulder muscle torques. This strategy resulted 375 in lower interaction torques from the shoulder acting at the elbow, which resulted in greater elbow 376 muscle torque requirements. Furthermore, increases in muscle torques at the nondominant elbow 377

14

resulted in greater interaction torques at the shoulder, an effect that required active compensation at 378 the shoulder. That is, muscle torques and interaction torques acted “antagonistically,” suggesting that 379 the nondominant arm uses a less efficient coordination strategy than does the dominant arm. 380 Figure 5 shows that these coordination patterns also characterized reaches to the 270° target, even 381 though those reaches were made in the opposite direction from the 90° target. Both limbs tended to 382 move the hand medial to the 270° target. However, linearity deviation was again greater for the left 383 limb, which generated greater flexor elbow muscle torque. 384 Figure 6 confirms that the patterns of shoulder and elbow muscle torques were relatively consistent 385 across targets and participants. More specifically, panel A of Figure 6 shows that initial peak muscle 386 torques were smaller at the shoulder of the nondominant arm compared to the shoulder of the 387 dominant arm. Panel B shows that this pattern was reversed at the elbow where peak muscle torques 388 were greater for the nondominant arm compared to the dominant arm. 389 To further support the idea that interlimb differences in dynamic coordination may have led to 390 interlimb differences in linearity deviation, we analyzed the relation, across participants, between 391 initial peak elbow muscle torque magnitude and linearity for nondominant and dominant reaches. If 392 higher peak elbow muscle torques led to increases in linearity deviation, then one would expect them 393 to be positively correlated with linearity deviation for the nondominant limb but not for the dominant 394 limb. Indeed, magnitudes of nondominant peak elbow muscle torques positively correlated with 395 nondominant linearity deviation at the 90° target, r(77)=.39, R2=.15, p=.022, and at the 270° target, 396 r(77)=.37, R2=.14, p=.027 (see Figure 7, panels A and C). By contrast, dominant peak elbow muscle 397 torques did not correlate with dominant linearity deviation at the 90° target, r(78)=.03, R2<.01, p=.832, 398 or at the 270° target, r(76)=.04, R2<.01, p=.737 (see Figure 7, panels B and D). 399 The non-zero regression slopes in panels A and C of Figure 7 indicate that nondominant peak elbow 400 muscle torque explained a moderate but significant amount of variance in nondominant linearity 401 deviation at each target. The moderate nature of these relations may be explained by the fact that for 402 each limb, elbow muscle torque contributed less to elbow net torque than did elbow interaction 403 torque. Variations in the shape of the torque profiles also contributes significantly to the relative 404 displacement profiles of the joints, and thus ultimately to hand-path linearity. These points 405 notwithstanding, the non-zero slopes for the nondominant limb contrast the near-zero slopes for the 406 dominant limb (see Figure 7, panels B and D). 407

15

Finally, we analyzed the relation between initial peak shoulder muscle torque magnitude and linearity 408 for dominant and nondominant reaches. If interlimb differences in linearity deviation were based on 409 interlimb differences in dynamic coordination across the elbow and shoulder, then linearity deviations 410 should also be more strongly negatively correlated with peak shoulder muscle torques in the dominant 411 arm. We found some support for this prediction, as dominant peak shoulder muscle torques negatively 412 correlated with dominant arm linearity deviation at both the 90° target, r(77)=-.23, R2=.05, p=.011, and 413 at the 270° target, r(77)=-.24, R2=.06, p=.042. By contrast, nondominant peak shoulder muscle torques 414 only correlated with nondominant linearity deviation at the 270° target, r(76)=-.30, R2=.09, p=.010, and 415 not at the 90° target, r(78)=.09, R2<.01, p=.477. 416

Discussion 417 Early research on brain lateralization proposed that the left-hemisphere plays a special role in motor 418 processes (Broca 1865; Geschwind 1975; Leipmann 1905). Later work focusing mainly on higher 419 cognitive and perceptual asymmetries established that both hemispheres are specialized for different 420 functions (Corballis 1998; Gazzaniga 2000a; Sperry 1982). Recent work from our lab has extended 421 these ideas to the motor system. We have proposed that the two hemisphere/limb systems are 422 specialized for distinct, but complementary, motor control processes: the dominant system for 423 coordination of limb and task dynamics and the nondominant system for stabilizing position through 424 impedance control mechanisms (Bagesteiro and Sainburg 2002, 2003; Przybyla et al. 2011; Sainburg 425 2002; Sainburg and Kalakanis 2000; Wang and Sainburg 2004, 2007). The idea of distinct and 426 complementary control systems was previously proposed by Guiard (1987), who described a similar 427 division of labor between the arms during bimanual actions. Our research emphasizes the 428 complementary division of labor between the two hemispheres during unimanual actions (Schaefer et 429 al, 2007). This characterization of handedness provides a unique opportunity to study the link between 430 functional lateralization and the decisions that people make about how to move. In fact, handedness 431 has traditionally been characterized by the arm choices that people make when they approach a given 432 task (Bryden 1977; Oldfield 1971). We now ask whether and how limb choice is related to, and may 433 therefore depend upon, sensorimotor asymmetries in dynamic coordination. 434 To address these questions, we had right-handed participants reach either limb to a single target in an 435 eight-direction center-out task. We centered the task in the workspace so there were three targets left 436 of body-midline, three targets right of midline, and two targets on the midline. We displaced both 437

16

cursors to a midline start circle so that each limb remained in its own hemispace (see Figure 1). This 438 design assured geometric symmetry between the arms for the midline targets and eliminated potential 439 mechanical conflicts between the limbs. 440 Our main question was whether asymmetries in limb choice for symmetric targets correspond to 441 asymmetries in performance. To address this question, we assessed whether the two midline targets 442 would receive more dominant reaches than would be expected by chance. These targets were 443 equidistant from the actual start locations of each hand and were neither in the left nor right 444 hemispaces. Therefore, if limb selection were dictated solely by a hemispace bias, one would expect an 445 equal distribution of dominant and nondominant reaches. However, data from the choice conditions 446 showed that the midline targets received more dominant than nondominant reaches (see Figure 2). 447 Data from our non-choice conditions clarified the relations between this pattern of limb choice and 448 performance asymmetries. In these conditions, the dominant hand-paths were substantially straighter 449 (see Figure 3, panel A); they reflected more proficient utilization of limb dynamics (shoulder interaction 450 torques), as compared to the nondominant arm (Bagesteiro and Sainburg 2002, 2003; Sainburg 2002; 451 Sainburg and Kalakanis 2000). While both arms had similar initial peak elbow and shoulder net torques, 452 the dominant arm strategy was characterized by higher shoulder muscle torque that drove motion at 453 both the shoulder and elbow joints. Critically, nondominant initial peak elbow muscle torques 454 positively correlated with linearity deviations, whereas no such correlations occurred for the dominant 455 arm (see Figure 7). Together, these findings suggest that in the absence of hemispace effects, 456 sensorimotor asymmetries in dynamic coordination may contribute to limb choice. 457 As predicted by hemispace biases, left targets received mostly nondominant reaches and right targets 458 received mostly dominant reaches (see Figure 2). Our data confirmed a trend for the dominant (right) 459 arm to reach to left targets more than the nondominant (left) limb reached to right targets. Previous 460 research has reported that the distribution of reaches is asymmetric across the workspace, such that 461 right-handers prefer dominant reaches to targets located in the middle of the workspace and to targets 462 just left of the body-midline (Bryden and Roy 2006; Bryden et al. 2000; Mamolo et al. 2004; Gabbard 463 and Helbig 2004; Peters 1995). To the best of our knowledge, however, the current study is the first to 464 show that the dominant limb reaches into contralateral space more than the nondominant arm even 465 when only the visual representation of the hand (i.e., the cursor) moves across the body-midline while 466 the hand does not. This finding suggests a strong influence of vision on the dominant-arm preference. 467

17

Consistent with this idea, Przybyla et al. (in press) reported that people choose significantly fewer 468 contralateral reaches with the dominant limb, when visual feedback is not available. Moreover, this 469 decrease appears to be driven by a corresponding change in sensorimotor performance that is 470 consistent with the dynamic dominance hypothesis. This hypothesis states that dominant reaches rely 471 largely on predictive mechanisms that specify mechanically efficient movement patterns. Such 472 predictive control has been shown to depend on visual feedback mechanisms to update predictive 473 mechanisms on a trial-to-trial basis (Sainburg 2002; Yadav and Sainburg 2011). Thus, if removing vision 474 decreases the relative proficiency of dominant reaches, and if limb choice depends on sensorimotor 475 asymmetries, then the dominant limb should reach less. The link between limb choice and the dynamic 476 dominance model that has emerged in this study and in the study of Przybyla et al. has implications for 477 theories of action selection, for our understanding of handedness, and possibly for models of neural 478 lateralization. We address these issues next. 479 A popular approach to the problem of action selection has been to identify linkages within the 480 neuromuscular system and then to model behavior as optimizations performed under those 481 constraints (Bernstein 1967; Todorov 2004). Using this approach, a number of factors have been 482 suggested to govern action selection; see Rosenbaum et al. (1990) for reviews. Although some 90% of 483 all humans prefer the right limb for most unimanual actions (Annett 1972; Corballis 1997, 2003), 484 researchers have been reluctant to characterize handedness as an important constraint on action 485 selection. This reluctance has likely been due to the fact that it has proven difficult to establish 486 whether sensorimotor performance asymmetries underlie handedness; see Carson (1993) and 487 Sainburg (2002) for discussions. Indeed, it would be difficult to argue that choosing the dominant arm 488 is optimal without knowing if there is a performance advantage for the dominant arm in that particular 489 instance. A related problem is the absence of studies that have examined limb choice and 490 sensorimotor asymmetries together in the same experiments. 491 The current study addresses these issues by assessing both limb choice and limb performance in the 492 same participants and by linking limb choice to a reliable sensorimotor asymmetry in dynamic 493 coordination. However, we do not suggest that asymmetries in dynamic coordination fully explain limb 494 choice. By contrast, we acknowledge that limb choice likely depends upon multiple factors. In the 495 current study, for example, it is clear that limb choice to targets that were visually located to the left 496 and right strongly depended on hemispace biases. But where hemispace biases could not predict limb 497

18

choice we found evidence for a link between limb choice and dynamic coordination differences. This 498 case was one in which the mechanical requirements for the two possible actions – a dominant or 499 nondominant reach – were symmetric and equivalent. Thus, the current findings suggest that the 500 hemispace bias may act as the most useful constraint on limb choice to lateralized targets, but also 501 because dynamic coordination asymmetries may facilitate action selection in the face of geometric and 502 mechanical symmetry. This perspective is consistent with the idea that action selection in general likely 503 depends on a hierarchy of multiple constraints whose relative priority may depend on task demands 504 (Rosenbaum et al. 2001). 505 In conclusion, our current results establish a clear relation between sensorimotor asymmetries in 506 dynamic coordination and the decisions people make about how to move. However, we do not argue 507 on the basis of this association that differences in dynamic coordination caused particular patterns of 508 arm choices. Our aim in this study is simply to demonstrate a link between arm selection and reliable 509 interlimb performance asymmetries. We can speculate, however, that although the interlimb 510 differences in energetics we observed may be small in the current task, the choices we observed could 511 reflect a well-established strategy based on dynamics. Moreover, that difference could be substantial 512 when stressed by either the need for continuous action (fatigue) or strenuous action (approximating 513 maximal output). The choices we observed in this sub-maximal task may reflect such rules that may 514 have been developed to deal with more taxing conditions under the stress of evolutionary adaptation 515 processes. We suggest that limb selection may in fact be driven by a lifetime (ontogenetic) and by 516 many generations (phylogenetic) of movement choices that involve more energetically costly 517 movements, such as those required during tool making, tool use, hunting, and fighting. Indeed, we 518 have recently verified that interlimb differences in dynamic coordination persist during more stressful 519 vertical reaching movements performed in a gravitational field (Tomlinson & Sainburg, 2012). It will be 520 important for future work to test whether hand choice corresponds to interlimb coordination 521 asymmetries in this and other energetically demanding tasks. 522 This link between arm selection and arm performance may also have a more general implication for 523 models of neural lateralization. It has been suggested that lateralization of neural function reflects an 524 adaptive “no-cost extension” whereby more cortical tissue can be devoted to expansion of function 525 while restricting control to local circuits -- a factor that is beneficial in terms of both speed and 526 energetics of neural processing (Corballis 2009; Gazzaniga 2000a; Tommasi 2009). Consistent with this 527

19

view, Sainburg and Eckhardt (2005) have argued that handedness first reflected adaptation at the 528 individual level that resulted from the demand for skilled behavior and later became conserved 529 genetically through the process of natural selection. Our current findings emphasize the link between 530 hemispheric specializations in dynamic coordination and action selection, thereby supporting this skill-531 based evolutionary perspective. Moreover, Gazzaniga (2000a) has pointed out that specialization of a 532 given hemisphere for a particular function implies that the other hemisphere is functionally limited for 533 that function. Thus, a primary benefit of lateralization is that new abilities can emerge over the course 534 of evolution through the expansion of intrahemispheric circuitry, while evolutionarily older abilities are 535 sustained. It follows, then, that lateralization should also underlie the ability to make adaptive 536 behavioral decisions that are based on lateralized skills. Future work can test this idea further by asking 537 whether people use the nondominant limb more in tasks that match that arm’s specialization for 538 position stabilization via impedance-based control mechanisms (e.g., Wang and Sainburg 2007). 539 540

20

References 541 Abdi H. The Bonferonni and Šidák corrections for multiple comparisons. In Neil Salkind (Ed.), 542

Encyclopedia of Measurement and Statistics. Thousand Oaks, CA: Sage, 2007. 543 Allen M. Models of hemispheric specialization. Psychological Bulletin 93: 73-104, 1983. 544 Annett M. The distribution of manual asymmetry. British Journal of Psychology 63: 343-358, 1972. 545 Bagesteiro LB, Sainburg RL. Handedness: Dominant arm advantages in control of limb dynamics. J 546

Neurophysiol 88: 2408-2421, 2002. 547 Bagesteiro LB, Sainburg RL. Nondominant arm advantages in load compensation during rapid elbow 548

joint movements. J Neurophysiol 90: 1503-1513, 2003. 549 Bernstein N. The coordination and regulation of movements. London: Pergamon, 1967. 550 Bradshaw JL, Bradshaw JA, Nettleton NC. Abduction, adduction and hand differences in simple serial 551

movements. Neuropsychologia 28: 917-931, 1990. 552 Broca P. Sur la facultfe du langage articulfe. Bulletin de la Societt d’Anthropologie de Paris 6: 493-494, 553

1865. 554 Bryden MP. Measuring handedness with questionnaires. Neuropsychologia 15: 617-624, 1977. 555 Bryden PJ, Roy EA. Preferential reaching across regions of hemispace in adults and children. 556

Developmental Psychobiology 48: 121-132, 2006. 557 Bryden PJ, Pryde KM, Roy EA. A performance measure of the degree of hand preference. Brain and Cog 558

44: 402-414, 2000. 559 Buckingham, G, Carey DP. Rightward biases during bimanual reaching. Exp Brain Res 194: 197-206, 560

2009. 561 Carey DP, Hargreaves EL, Goodale MA. Reaching to ipsilateral or contralateral targets: within-562

hemisphere visuomotor processing cannot explain hemispatial differences in motor control. Exp 563 Brain Res 112: 496-504, 1996. 564

Carson RG. Manual asymmetries: Old problems and new directions. Human Movement Science 12: 565 479-506, 1993. 566

Carson RG, Chua R, Goodman D, Byblow WD, Elliott D. The preparation of aiming movements. Brain 567 and Cog 28: 133-154, 1995. 568

Carson RG, Elliott D, Chua R, Goodman D. Dickinson J. Manual asymmetries in the reproduction of a 3-569 dimensional spatial location. Neuropsychologia 28: 99-103, 1990. 570

21

Carson RG, Goodman D, Elliott D. Asymmetries in the discrete and pseudocontinuous regulation of 571 visually guided reaching. Brain and Cog 18: 169-191, 1992. 572

Chua R, Carson, RG, Goodman D, Elliott D. Asymmetries in the spatial localization of transformed 573 targets. Brain and Cog 20: 227-235, 1992. 574

Corballis MC. The genetics and evolution of handedness. Psychological Review 104: 714-727, 1997. 575 Corballis MC. Cerebral asymmetry: Motoring on. Trends in Cognitive Sciences 2: 152-157, 1998. 576 Corballis MC. From mouth to hand: gesture, speech, and the evolution of right-handedness. Behavioral 577

Brain Sciences 26: 199-208, 2003. 578 Corballis MC. The evolution and genetics of cerebral asymmetry. Phil. Trans. R. Soc. B 364: 867-879, 579

2009. 580 Elliott D, Roy EA, Goodman D, Chua R, Carson RG, Maraj BKV. Asymmetries in the preparation and 581

control of manual aiming movements. Can Journal of Experimental Psychology 47: 570-589, 1993. 582 Fisk JD, Goodale MA. The organization of eye and limb movements during unrestricted reaching to 583

targets in contralateral and ipsilateral visual space. Exp Brain Res 60: 159-178, 1985. 584 Gabbard C, Helbig CR. What drives children’s limb selection for reaching in hemispace? Exp Brain Res 585

156: 325-332, 2004. 586 Gabbard C, Rabb C, Gentry V. Attentional stimuli and programming hand selection: A developmental 587

perspective. International Journal of Neuroscience 96: 205-215, 1998. 588 Gazzaniga MS. Cerebral specialization and interhemispheric communication: Does the corpus callosum 589

enable the human condition? Brain 123: 1293-1326, 2000a. 590 Geschwind N. The apraxias: Neural mechanisms of disorders of learned movement. American Scientist 591

63:188-195, 1975. 592 Gonzalez CLR, Goodale MA. Hand preference for precision grasping predicts language lateralization. 593

Neuropsychologia 47: 3182-3189, 2009. 594 Guiard, Y. Asymmetric division of labor in human skilled bimanual action: The kinematic chain as a 595

model. Journal of Motor Behavior 19: 486-517, 1987. 596 Haaland KY, Flaherty D. The different types of limb apraxia errors made by patients with left vs. right 597

hemisphere damage. Brain Cognition 3: 370–384, 1984. 598 Hogg RV, Craig AT. Introduction to Mathematical Statistics (5th ed.). Englewood Cliffs, NJ: Prentice Hall. 599 Hommel B. Inverting the Simon effect by intention: Determinants of direction and extent of effects of 600

irrelevant spatial information, Psychological Research 55: 270-279, 1993. 601

22

Ingum J, Bjorklund R. Effects of flunitrazepam on responses to lateralized visual stimuli: evidence for 602 cerebral asymmetry of execution of manual movements to targets in contralateral and ipsilateral 603 visual space. Psychopharmacology 114: 551-558, 1994. 604

Latash ML. Neurophysiological basis of movement (2nd ed.). Champaign, IL: Human Kinetics, 2007. 605 Liepmann H. Die linke Hemisphäre und das Handeln. Münchener Medizinische Wochenschrift 49: 2375-606

2378, 1905. 607 Mamolo CM, Roy EA, Bryden PJ, Rohr LE. The effects of skill demands and object position on the 608

distribution of preferred hand reaches. Brain and Cog 55: 349-351, 2004. 609 Mamolo CM, Roy EA, Rohr LE, Bryden PJ. Reaching patterns across working space: The effects of 610

handedness, task demands, and comfort levels. Laterality 11: 465-492, 2006. 611 Mani S, Mutha PK, Przyblya A, Haaland K, Good D, Sainburg RL. Contralesional motor deficits after 612

unilateral stroke reflect hemisphere-specific control mechanisms. Brain, in press. 613 Mutha PK, Sainburg RL, Haaland KY. Left parietal regions are critical for adaptive visuomotor control. 614

Journal of Neuroscience 31: 6972-6981, 2011. 615 Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 9: 616

97-113, 1971. 617 Peters M. Attentional asymmetries during concurrent bimanual performance. Quarterly Journal of Exp 618

Psychology 33: 95-103, 1981. 619 Peters M. Handedness and its relation to other indices of cerebral lateralization. In R.J. Davidson, & K. 620

Hugdahl (Eds.), Brain asymmetry (pp. 183-214). Cambridge, MA: MIT Press, 1995. 621 Prablanc C, Echallier JF, Komilis E, Jeannerod M. Optimal response of eye and hand motor systems in 622

pointing at a visual target 1: Spatio-temporal characteristics of eye and hand movements and their 623 relationships when varying the amount of visual information. Biological Cybernetics 35: 113-124, 624 1979. 625

Przybyla A, Good D, Sainburg RL. Dynamic dominance varies with handedness: reduced interlimb 626 asymmetries in left-handers. Exp Brain Res 216: 419-431, 2012. 627

Przybyla A, Coelho C, Akpinar S, Kirazci S, Sainburg RL. Sensorimotor performance asymmetries predict 628 hand selection. Neuroscience: in press. 629

23

Rosenbaum DA, Marchak F, Barnes H J, Vaughan J, Slotta J, Jorgensen M. Constraints for action 630 selection: Overhand versus underhand grips. In M. Jeannerod (Ed.), Attention and Performance XIII: 631 Motor representation and control (pp. 321-342). Hillsdale, NJ: Lawrence Erlbaum Associates, 1990. 632

Rosenbaum DA, Meulenbroek RG, Vaughan J, Jansen C. Posture-based motion planning: Applications to 633 grasping. Psychological Review 108: 709-734, 2001. 634

Sainburg RL. Evidence for a dynamic-dominance hypothesis of handedness. Exp Brain Res 142: 241-635 258, 2002. 636

Sainburg RL, Eckhardt RB. Optimization through lateralization: The evolution of handedness. Behavioral 637 and Brain Sciences 28: 575-633, 2005. 638

Sainburg RL, Ghez C, Kalakanis D. Intersegmental dynamics are controlled by sequential anticipatory, 639 error correction, and postural mechanisms. J Neurophysiol 81: 1045-1056, 1999. 640

Sainburg RL, Kalakanis D. Differences in control of limb dynamics during dominant and nondominant 641 arm reaching. J Neurophysiol 83: 2661-2675, 2000. 642

Sainburg RL, Wang J. Interlimb transfer of visuomotor rotations: independence of direction and final 643 position information. Exp Brain Res 145: 437-447, 2002. 644

Schaefer SY, Haaland KY, Sainburg RL. Ipsilesional motor deficits following stroke reflect hemispheric 645 specializations for movement control. Brain 130: 2146–2158, 2007. 646

Schaefer SY, Haaland KY, Sainburg RL. Hemispheric specialization and functional impact of ipsilesional 647 deficits in movement coordination and accuracy. Neuropsychologia 47: 2953-2966, 2009a. 648

Schneider K, Zernicke RF. A Fortran package for the planar analysis of limb intersegmental dynamics 649 from spatial coordinate-time data. Adv Eng Software. 12: 123–128, 1990. 650

Sperry R. Some effects of disconnecting the cerebral hemispheres. Science 217: 1223-1226, 1982. 651 Todorov E. Optimality principles in sensorimotor control. Nature Neuroscience 7: 907-915, 2004. 652 Tomassi L. Mechanisms and functions of brain and behavioral asymmetries. Phil. Trans. R. Soc. 364: 653

855-859, 2009. 654 Tomlinson T, Sainburg RL. Dynamic Dominance Persists During Unsupported Reaching. Journal of 655

Motor Behavior: 13-25, 2012. 656 Van Der Staak C. Intra- and interhemispheric visual-motor control of human arm movements. 657

Neuropsychologia 13: 439-448, 1975. 658 Verfaellie M, Heilman KM. Hemispheric Asymmetries in Attentional Control: Implications for Hand 659

Preference in Sensorimotor Tasks. Brain and Cog 14: 70-80, 1990. 660

24

Wang J, Sainburg RL. Mechanisms underlying interlimb transfer of visuomotor rotations. Exp Brain Res 661 149: 520-526, 2003. 662

Wang J, Sainburg RL. The dominant and nondominant arms are specialized for stabilizing different 663 features of task performance. Exp Brain Res 178: 565-570, 2007. 664

Yadav V, Sainburg RL. Motor Lateralization is characterized by a serial hybrid control scheme. 665 Neuroscience 196: 153-167, 2011. 666

667 668

25

Notes 669 1. Three trials (.94%) were excluded from the linearity analysis because they yielded values that were 670

more than three standard deviations above the mean. Nine trials (2.8%) were excluded from the 671 final-position error analysis, and four trials (1.3%) were excluded from the elbow muscle torque 672 analysis for the same reason. 673

2. The ANOVAs that yielded these results also included target as a factor. Thus, there were no 674 differences in shoulder or elbow net torque across the two midline targets. 675

676 677 678

26

Appendix A 679 We modeled each limb as a serial three-link manipulator with eight degrees of freedom (DOF). The 680 eight DOFs came from three translational joints and five rotational joints. The first three translational 681 (x-,y-, and z-) DOFs at the shoulder described movement of the shoulder that resulted from movement 682 of the trunk. The next three DOFs at the shoulder corresponded to shoulder rotation. The first two 683 DOFs at shoulder defined the orientation of the plane containing the shoulder, elbow, and wrist joints; 684 the third described the orientation of the upper arm in this plane. The last two DOFs were 685 intersegment joint angles at the elbow and wrist. We did not include pronation, wrist flexion, or wrist 686 extension in the model because we immobilized the wrist and allowed movement only in the 687 horizontal plane. These features of our experiment made negligible the dynamic effects at the elbow 688 and at the shoulder arising from pronation, wriest flexion, or wrist extension. 689 In Appendix Figure 1 (panel A), i = 1, 2,3 represent the three links corresponding to upper arm, 690 forearm, and hand, respectively. The terms mi, Ixi, Iyi, and Izi represent masses and moments of inertia 691 of each limb; we computed these variables with the method described in Zatsiorsky (2002). The terms li 692 and ri represent the length and center of mass location for ith link, respectively. 693 As shown in Figure 1 (panel B), we described the configuration of the limb model in generalized 694 coordinates, q, as follows: 695

(1) 696 We used a screw theory-based approach to compute these joint angles and joint torques from the 697 position data of the Flock of Birds motion sensors. Screw theory describes motion of a rigid body as 698 rotation about an axis and translation along that axis (Ball, 1876). When combined with Lie algebra, 699 screw theory provides tools that can be used to derive equations of motions for complex mechanical 700 systems. In turn, these equations can be used to study mechanical properties of the system and to 701 develop control algorithms (Murray, Li, & Sastry, 1994). Below, we present the related inverse 702 kinematics and inverse dynamics method. We computed joint angles and derived equations of motion 703 using screw theory as follows: 704 First, we described a velocity twist. For a purely rotational or a revolute joint, the velocity twist is 705 defined by the orientation of the screw axis and a point on the screw axis by 706

27

(2) 707 where w is the unit vector along axis of rotation, and r is any point on the screw axis -- the axis of 708 rotation. For example, the velocity twist corresponding to the elbow joint is given by 709

(3) 710 For a purely translational or prismatic joint, the velocity twist is defined as the direction of movement: 711 (4) 712 where h is the unit vector along the direction of motion. For example, the velocity twist corresponding 713 to translation degree of freedom along x axis is given by 714 (5) 715 Therefore, the matrix of twists defining the arm model in Panel B of Figure 1 is 716

(6) 717 Second, we computed inverse kinematics. To obtain equations for forward kinematics using screw 718 theory, we needed the velocity twists that defined the mechanical system (Eq. 6) and the forward 719 kinematic map in zero configuration, 720

(7) 721 where I3X3 is the 3 dimensional identity matrix. 722 723 We then computed the forward kinematics as 724

(8) 725 where we computed matrix exponents as described in Murray et al. (1994). Once we had the forward 726 kinematic mapping between joint angles and marker positions, we computed the inverse kinematics 727 using analytical methods, when possible, or by using Newton Raphson root solving methods. 728 Third, we computed inverse dynamics. We calculated the joint torques that generated movement 729 using the joint angles computed in the previous step. We did so by first deriving equations of motion 730

28

for the arm model in Figure 1 (Panel B). The general form of the equations of motion of the arm model 731 in a zero-gravity condition is 732

(9) 733 where M is the mass-inertia matrix of the human arm and C is the matrix corresponding to Coriolis and 734 centrifugal contributions from movement of joints. We therefore computed M and C matrices for our 735 arm model. 736 Computing equations of motions for complex mechanical systems is straightforward with this screw 737 theory-based approach in conjunction with methods from Lie Algebra (Murray et al). To derive 738 equations of motion using screw theory, we needed the velocity twist that defined the pose of the 739 center of mass of each link. For example, in the model of human arm in Figure 1 (panel B), the twists 740 that defined movement of the center of mass and forward dynamics in zero configuration of the fore 741 arm are given by 742

(10) 743 and 744

(11) 745 The twists and forward kinematics in zero-configuration for other links were computed in a similar 746 manner. We next used these formulations to compute M and C in Eq. 9. 747 To compute M, we first computed the mass-inertia matrix. The kinetic energy of the arm model in 748 Figure 1 is 749

(12) 750 where Jb is the body Jacobian of the mechanical system. Note that there are different types of 751 manipulator Jacobian for different applications. We used the body Jacobian as defined in Murray et al. 752 The mass-inertia matrix was then simply M(q) in Eq. 12. 753 To calculate C, we computed the matrix corresponding to Coriolis and centrifugal terms from M using 754 Christoffel Symbols 755

29

756 (13) 757

Fourth, we partitioned joint torques into interaction, muscle and net torques. 758 We defined muscle torque as the torques injected into the mechanical system at the joints 759

. (14) 760 Net torque was defined as the joint torques corresponding to the acceleration of the ith joint 761

(15) 762 where Dj,k = 1, if j=k=i else Dj,k = 0. 763 Once we obtained the net and muscle torques, we computed the interaction torques. Noting that 764

, we rewrote Eq. 8 as 765 (16) 766

Rearranging terms in Eq. 16, we obtained interaction torques. 767 (17) 768

Equations 14, 15, and 17 gave muscle, net, and interaction torques, respectively, for the human arm 769 model described in Figure 1. 770 771 772

30

Table 773

774 775 776

31

Figure Captions 777 Appendix Figure 1. The human arm modeled as a 3-link system with three translational DOFs and five 778 rotational DOFs (Panel A). The mechanical equivalent of the human arm model in zero configuration 779 (Panel B). 780 Figure 1. Schematics of the experimental apparatus (panel A) and task (panel B). The distance between 781 the start circle (d=2.5cm) and each target (d=3.5 cm) was 13 cm. Only one target was visible in each 782 trial. There were eight potential targets that were spaced at 45° intervals. 783 Figure 2. Mean proportions of nondominant (ND) and dominant (D) reaches to each target. Means 784 were computed across participants (N=10). Rectangles show how targets were grouped into the left 785 workspace, midline workspace, and right workspace. 786 Figure 3. Mean hand-path deviation from linearity (minor axis divided by major axis; see Method), final 787 position error (absolute distance between the center of the target and the center of the cursor), and 788 maximum tangential hand velocity as a function of nondominant (ND) and dominant (D) limbs, and 789 midline targets (90° and 270°). Error bars represent ± one standard error of the mean. 790 Figure 4. Representative hand paths for each limb (panel A), tangential hand velocity profiles (panel B), 791 shoulder torque components (panel C), and elbow torque components (panel D) toward the 90° target. 792 Example data were taken from the same participant. Torque components include net torque (thick 793 black line), interaction torque (solid, light-gray line), elbow muscle torque (solid, dark-gray line), and 794 shoulder muscle torque (dashed gray line). Elbow muscle torque is equal in amplitude but opposite in 795 sign at the elbow and shoulder. 796 Figure 5. Representative hand paths for each limb (panel A), tangential hand velocity profiles (panel B), 797 shoulder torque components (panel C), and elbow torque components (panel D) toward the 270° 798 target. Example data were taken from the same participant; these data and previous example data 799 (Figure 4) came from different participants. The signs of the torque components are the opposite those 800 shown in Figure 4 because movements to the 270° target were made in the opposite direction. 801 Figure 6. Mean absolute initial peak shoulder (panel A) and elbow (panel B) muscle torque, averaged 802 across subjects. Data are plotted as a function of nondominant (ND) and dominant (D) limbs and 803 midline targets (90° and 270°). Error bars represent ± one standard error of the mean. 804 Figure 7. Hand-path deviation from linearity in arbitrary units (minor axis divided by major axis; see 805 Method) as a function of absolute initial peak elbow muscle torque on each of 80 non-choice trials. 806 Panels A and B show data from nondominant (circles) and dominant (triangles) reaches to the 90° 807 target. Panels C and D show data from nondominant (circles) and dominant (triangles) reaches to the 808 270° target. 809

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Appendix Figure 1

A B