Neuropsychologia 43 (2005) 1547–1558

Don’t think twice, it’s all right—contralesional dependency
for bimanual prehension movements
T.D. Punt a, b, ∗ , M.J. Riddoch a , G.W. Humphreys a
a Behavioural Brain Sciences Centre, School of Psychology, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK
b School of Health Sciences, University of Birmingham, Birmingham B15 2TT, UK

Received 3 September 2003; received in revised form 7 October 2004; accepted 3 February 2005
Available online 7 March 2005

Abstract
In bimanual coordination when moving the hands to two separate objects, subjects tend to initiate and terminate the movements together,
even when the targets are at different distances or are of a different size. Additionally, each hand tends to scale its grasp independently to the
object to be grasped. Here, we report the performance of a patient, who had previously shown signs of motor neglect, on two experiments
investigating coupling and independence in bimanual coordination. The patient showed relatively normal bimanual behaviour for the transport
phase of prehension when objects were placed at different distances (Experiment 1), but abnormal behaviour for the grasp component when
objects were of different sizes (Experiment 2). Moreover, the contralesional limb demonstrated a dependency of grasp that was related to the
object grasped by the ipsilesional limb. We discuss the possible underlying mechanisms of this dependency in relation to competitive motor
programming and attentional bias. The results also reinforce the view that the transport and grasp components of prehension are distinct
processes.
© 2005 Elsevier Ltd. All rights reserved.
Keywords: Reach-to-grasp; Bimanual coordination; Motor neglect; Extinction; Attentional bias

1. Introduction
The ability to coordinate movements of the two hands during functional activities is a remarkable skill that humans and
other primates have developed. Many varied skills (e.g. using
a tin-opener, playing a piano) demonstrate both the independence of the two limbs and also the cooperation between
them. In recent years, we have witnessed a rapid expansion
in our understanding of the relationship between the hands
at both behavioural and neural levels. For example, when
the hands move to targets of variable distance and direction,
there appears to be strong coupling of the limbs with almost
synchronous movement onsets, times to peak velocity and
movement end, despite the hands moving at different speeds
(Kelso, Southard, & Goodman, 1979).
The performance of bimanual reach-to-grasp movements
presents potential dilemmas for the organising system, es∗ Corresponding author. Tel.: +44 121 414 4949; fax: +44 121 414 4897.

E-mail address: t.d.punt@bham.ac.uk (T.D. Punt).
0028-3932/$ – see front matter © 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.neuropsychologia.2005.02.001

pecially when the targets are of differing distances or different sizes. For instance, can subjects independently scale
velocity to two targets of differing distance, or independently
shape each hand to targets of differing size while maintaining synchrony between the two limbs in terms of movement
onset and end? The evidence appears to suggest that they
can. Castiello et al.’s initial work on bimanual reach-to-grasp
movements demonstrated that subjects tend to synchronise
movement onset and movement end points when moving
to targets of differing sizes, while maintaining independent
shaping of each hand based on the relative size of the target (Castiello, Bennett, & Stelmach, 1993). However, not
all temporal parameters were coupled. It was found that the
hand moving towards the smaller target, reached the point
of peak velocity earlier than the hand moving towards the
larger target and had a longer deceleration period. It is notable
that in this study, subjects made a ‘whole hand’ grip to the
large target while making a ‘precision’ grip to the small target. More recent work investigating bimanual reach-to-grasp
movements to different sized targets has shown close cou-

1548

T.D. Punt et al. / Neuropsychologia 43 (2005) 1547–1558

pling of all the major temporal parameters (movement onset,
time to peak velocity, time to peak grip aperture, movement
end) (Jackson, Jackson, & Kritikos, 1999). However, in line
with the above study of Castiello et al. (1993), shaping of
the hand was found to be independent for each hand relative
to the target to-be-grasped. This particular aspect of bimanual movements appears to be robust, being further replicated
in a study by Dohle, Ostermann, Hefter, and Freund (2000).
However, the finding may be task-dependent. Further work
by Jackson, German, and Peacock (2002) has shown that
when two different sized targets are ‘unified’ so that they appear to be part of the same object, the hands show a reduced
level of independence. That is, the hand moving towards the
larger target shapes smaller and the hand moving towards the
smaller target shapes larger, relative to when the targets are
not ‘unified’. It should be noted however, that the change,
while significant, was relatively small with the hand moving towards the smaller target shaping much smaller than the
hand moving towards the large target, independent of ‘unification’.
When the two hands move to two targets differing in distance, there is again synchronisation of onset and endpoints
(Kelso, Putnam, & Goodman, 1983). To achieve this synchronisation, it is necessary for the subject to scale the velocity
profile of each hand appropriately (slower to the near target,
faster to the far target). This very synchronisation, however,
leads to the violation of Fitts’ Law, which predicts that movement duration is affected by the distance moved and the precision required by the size of the target, these components
leading to a particular index of difficulty (Fitts, 1954). As the
target becomes smaller or moves further away, the index of
difficulty rises; as the target gets closer or becomes larger, so
the index of difficulty reduces. Interestingly, when moving
to targets of mixed difficulty, one might assume the movements to synchronise to the movement with the lowest index
of difficulty. However, this does not appear to be the case.
Kelso et al. showed that the movement with the higher index
of difficulty moved faster and the hand with the lower index
of difficulty moved slower in order to achieve the synchronisation (Kelso et al., 1979). This finding has been shown to be
true for bimanual reach-to-grasp movements also (Jackson et
al., 1999).
Nevertheless, though aspects of bimanual movements
are usually coordinated, there are also typically costs to
performance. For example, relative to unimanual movements, bimanual reach-to-grasp movements are slower (peak
velocity), take longer (movement duration) and shape
larger (peak grip aperture) (Jackson et al., 1999). These
changes may represent the level of certainty involved in
computing and controlling the movements as the degrees
of freedom for the unimanual movements are doubled
for bimanual actions. This cost for bimanual movements
is typically borne by both hands. Thus, the peak grip
aperture of both hands increases under bimanual conditions compared to unimanual movements (Jackson et al.,
1999).

The performance of bimanual movements is also affected
by attention. For example, bimanual movements are slower
and less stable when carried out under dual task conditions (Monno, Temprado, Zanone, & Laurent, 2002). Also,
performance is affected by directing attention to one of
the hands involved. During bimanual circle-drawing, synchrony is optimised when attention is directed to the dominant hand, while non-dominant hand movements are more
accurate when attention is directed towards it (Swinnen,
Jardin, & Meulenbroek, 1996; Wuyts, Summers, Carson,
Byblow, & Semjen, 1996). However, a recent study of
bimanual reach-to-grasp movements showed no effect of
directing visual attention towards one target (Jackson et
al., 2002). Movements remained tightly synchronised while
each limb expressed independence for the size of target to
be grasped.
Based on the above findings, one might expect that patients
with attentional deficits would have difficulties with bimanual
movements. Recently, evidence from bimanual hand circling
in patients with parietal lesions has suggested that the specific
role of the parietal lobe in directing attention may account
for this abnormal performance (Serrien, Nirkko, Lovblad, &
Wiesendanger, 2001). Three patients with left parietal lesions
showed a lag of the contralesional limb that worsened with
task difficulty. Additionally, Jackson et al. (2000) studied a
patient with a right parietal lesion (DB) on a bimanual prehension task and showed inappropriate coupling of peak velocity
of the contralesional limb when bimanual movements were
made to targets at different distances. The authors attribute
this finding to severe hemianaesthesia of the left side, which
reduced the proprioceptive cues available for DB to control
the movement. It is also possible, though; that some abnormal interactions between the hands for DB may reflect an
attentional bias that affects motor performance. This was examined here in the study of a patient with (recovered) motor
neglect.
Patients with motor neglect are able to move their affected limbs relatively normally when prompted to do so,
but show an under-use under normal conditions when the
ipsilesional limb is favoured. The problems of the contralesional limb may also be exacerbated under bimanual conditions, perhaps due to competition between the hands in
effecting a motor programme, generating a pattern of ‘motor extinction’ (Valenstein & Heilman, 1981; Meador, Watson, Bowers, & Heilman, 1986). Reports of motor extinction are few but case studies of patients who showed motor extinction for movement initiation (hypokinesia) do exist
(Valenstein & Heilman, 1981; Meador et al., 1986), together
with more recent accounts for impersistence (Mattingley &
Driver, 1997; Mattingley, 2002). Motor extinction may be
thought of as an analogue of perceptual extinction, a relatively
well recognised deficit within the neglect syndrome (Driver &
Vuilleumier, 2001). Two problems emerge when attempting
to identify patients with motor neglect or extinction. Firstly,
the diagnosis is a clinical one, based on observation of the
patient’s behaviour and to date there are no specific, recog-

T.D. Punt et al. / Neuropsychologia 43 (2005) 1547–1558

nised tests as there are for visual neglect. Secondly, many
patients with spatial neglect have concurrent hemiparesis,
rendering the ability to test for a neglect-related deficit impossible. Nevertheless, there is a growing recognition that primary motor deficits (e.g. hemiparesis) may be masking motor
neglect in many patients (Robertson, McMillan, MacLeod,
Edgeworth, & Brock, 2002; Vallar, Bottini, & Sterzi,
2003).
Here, we report data on bimanual reach-to grasp movements in a patient who was clinically identified post-stroke
as having motor neglect and extinction. The two experiments were based on those previously performed on normal subjects (Jackson et al., 1999). In Experiment 1, the
distance of the two targets was manipulated to investigate
independence of the two limbs for the transport phase of
reaching-to grasp. In Experiment 2, the size of the targets
was manipulated in order to investigate independence of the
limbs in the grasp phase. Jeannerod (1981) first proposed
that reach-to-grasp movements are “organised on the basis
of separate motor components corresponding to dissociable
sensory channels and perceptual descriptions” (page 155).
These separate motor components are involved in the control
of ‘transport’ and ‘grasp’ components of movement. Thus,
changing the location of an object alters the transport but
not the grasp component of an action (Chieffi, Gentilucci,
Allport, Sasso, & Rizzolatti, 1993), while object size affects
grasp but not necessarily the transport component (Bootsma,
Marteniuk, MacKenzie, & Zaal, 1994). Though the functional separation between the reach and grasp components
of movement should be supported by an underlying neurological separation, there are few neuropsychological data
supporting this. Jeannerod, Decety, and Michel (1994) did
report a selective impairment for the grasp component with
normal transport kinematics in a patient with bilateral parietal damage. However, to our knowledge a double dissociation has not been observed, which sheds doubt on whether
the two components are represented by distinct neural substrates. This was tested here by comparing bimanual coupling between, respectively, the reach and grasp components
of the task. We report a dissociation between relatively preserved bimanual actions for the transport component (Experiment 1), and impaired bimanual actions (contralesional
dependency) for the grasp component (Experiment 2). The
data emphasise the distinction between these components
of action.

2. Case report
MM, a 75-year-old right-handed woman was admitted
to hospital having suffered a cerebrovascular accident in
December 2001. A CT scan revealed infarction in the region of the right middle cerebral artery affecting the inferior
frontal gyrus, the middle frontal gyrus and the superior temporal gyrus (see Fig. 1). She had a mild hemiparesis and
demonstrated motor neglect, an underuse of the contrale-

1549

sional limbs. This was most obvious in the initial weeks after stroke when MM appeared to have total paralysis of the
contralesional (left) arm and yet used the arm relatively normally when specifically asked to use it. Motor neglect was
also apparent during gait, when MM would tend to ‘drag’
the left leg but walked relatively normally when reminded
to step the left foot. At this time, neurological examination
revealed the following information. Muscle power was 4/5
on the left and 5/5 on the right (MRC scale). Assessment
of tone showed no abnormalities with equal tendon reflexes
left and right. Plantar responses were downward bilaterally
and there was no clonus. She made a good recovery and
was largely independent in tasks of everyday living. This
study was undertaken 6 months following MM’s stroke in
June 2002.
At the time of the current tests of motor performance, MM
scored at ceiling on the star cancellation, line cancellation
and letter cancellation subtests of the Behavioural Inattention Test (Wilson, Cockburn, & Halligan, 1987). She also
performed at ceiling when examined using confrontation for
visual extinction (10 left, 10 right and 10 bilateral stimuli
presented). In order to ‘push’ this further, MM was tested
on the Posner Cueing Paradigm (Posner, Inhoff, & Friedrich,
1987). Here, MM performed well (see Fig. 2) and, while
responding marginally slower to contralesional targets, was
not abnormally affected by ipsilesional cues occurring prior
to contralesional targets1 . MM did show a degree of tactile
extinction showing contralesional extinction on 4 out of 10
presentations of bilateral light touch stimuli to the hands. Proprioception was intact for both hands when tested for movement detection and discrimination but MM had difficulties
in limb matching (examiner moves contralesional limb and
subject moves ipsilesional limb to the same ‘mirrored’ position). However, this particular test, while clinically popular,
appears to confound the relative contributions of proprioception and attention and as such should be interpreted with
care. MM was also tested on a novel test of motor extinction
both shortly following stroke and at the time of this study. She
was asked to tap as quickly as possible using WPS Electronic
Tapping Tests (WPS ETT). Shortly following stroke, when
tapping with the right hand unimanually, MM registered 46
taps in a 10 s period, compared with 44 left hand taps. However, when tapping bimanually she registered 44 taps with
the right hand but only 19 with the left. When repeated on
the task at the time of this study, MM made equivalent numbers of taps with the left and right hands both unimanually
and bimanually. Thus, motor extinction was reduced at the
time of testing, though it could still contribute to bimanual
performance under more taxing motor conditions (reaching
and grasping objects, as here).

1 MM responded slower to contralesional targets (mean = 582 ms) than
ipsilesional targets (mean = 495 ms) but crucially did not show a significant
Side × Validity interaction (F(1, 286) < 1.0, p = 0.7) indicative of visual extinction.

1550

T.D. Punt et al. / Neuropsychologia 43 (2005) 1547–1558

Fig. 1. Lesion reconstruction for MM, from CT scan. The lesion has been drawn onto standard slices from Gado, Hanaway, and Frank (1979). The bottom
figure shows the 10 slices used. Only slices 3–8 are depicted here. The left of each slice represents the right hemisphere.

3. Experiment 1
In Experiment 1, we examined the ability of MM to scale
the transport component of reach-to-grasp movements by
comparing unimanual with bimanual movements. In addition, for bimanual movements, we examined both congruent (target of equal distance away) and incongruent (targets
placed at different distances) movements. We hypothesised

Fig. 2. MM’s results on the valid vs. invalid cue paradigm (Posner et al.,
1987). While reaction times were slightly slower to contralesional targets,
the typical severe slowing to contralesional targets following an ipsilesional
cue was not observed.

that MM may inappropriately couple the velocity profiles of
her two hands to incongruent targets as described above for
DB (Jackson et al., 2000).
3.1. Method
MM was asked to make either unimanual or bimanual
movements to targets placed at positions on the table in front
of her (see Fig. 3). Each trial began with both of MM’s hands
resting on the table with the thumbs and index fingers gently resting around small cylinders fixed to the table. Targets
(cylinders 6 cm in diameter and 2 cm deep) were placed either
15 cm or 30 cm directly in front of each hand. On unimanual

Fig. 3. Set-up for Experiment 1. MM made unimanual reaches to either near
or far targets, bimanual congruent reaches (far/far; near/near) or bimanual
incongruent reaches (near/far; far/near).

T.D. Punt et al. / Neuropsychologia 43 (2005) 1547–1558

1551

(near, far) and condition (unimanual, bimanual congruent,
bimanual incongruent).

trials, a single target was presented; on bimanual trials, two
targets were presented. When cued to move by a ‘beep’, MM
reached ‘quickly but accurately’ to the targets. Each trial was
presented in full vision of the subject and it was introduced
by the experimenter (“this time there is a near target on the
left and a far target on the right”, or, “this time there is just a
far target on the right”, etc.). Conditions were presented randomly over trials and MM performed 10 repetitions for each
of the 8 conditions (unimanual: left near, left far, right near,
right far; bimanual: congruent near, congruent far, incongruent left near right far and incongruent left far right near).
The movements were measured using a 3D motion tracking
system (ProReflex, Qualysis Ltd.), sampling at 100 Hz. Reflective markers were placed on the nail of the thumb and
index fingers, the radial head at the wrist and on each of the
targets (the target markers were not visible to the subject).

3.2.1. Kinematics for the transport phase
3.2.1.1. Movement duration. Movements to ‘far’ targets
were longer than those to ‘near’ targets (near = 717 ms,
far = 859 ms, F(1, 78) = 19.0, p < 0.0001). There was also
a significant main effect of condition (F(2, 78) = 10.2,
p < 0.0005). Unimanual movements (682 ms) were significantly shorter than both bimanual congruent movements
(876 ms, F(1, 78) = 18.4, p < 0.0001) and bimanual incongruent movements (817 ms, F(1, 78) = 11.6, p < 0.005). However, the two bimanual conditions were equivalent (F(1,
79) = 1.0, p = 0.3). Left hand movements took slightly longer
than right hand movements (left = 821 ms, right = 765 ms,
F(1, 78) = 4.3, p = 0.04). None of the two-way or three-way
interactions approached statistical significance.

3.2. Results
3.2.1.2. Peak velocity. The left and right hands had
equivalent values for peak velocity (left = 554 mm/s,
right = 572 mm/s, F(1, 93) < 1.0, p = 0.4). Peak velocity was
higher for targets that were further away than for targets that were closer (near = 457 mm/s, far = 660 mm/s, F(1,
93) = 114.3, p < 0.0001). There was a significant main effect for condition (F(2, 93) = 16.5, p < 0.0001). Peak velocity for unimanual movements (630 mm/s) was significantly
higher than both bimanual congruent movements (540 mm/s,
F(1, 93) = 17.8, p < 0.0001) and bimanual incongruent movements (513 mm/s, F(1, 93) = 29.3, p < 0.0001). However, the
two bimanual conditions were equivalent (F(1, 93) < 1.0,
p = 0.3). There was a significant Hand × Distance interaction (F(1, 93) = 8.0, p < 0.01) with the left hand on average tending to have a higher peak velocity to near targets
and a lower peak velocity to far targets than the right hand
(Fig. 4).

Movement parameters were chosen and calculated using
customised software (QTools, Qualysis Ltd.), defined as:
Movement onset—this was taken as the first frame in which
wrist marker velocity consistently exceeded 50 mm/s.
Movement end—first frame in which target marker position
moved 1 mm.
Movement duration (MD)—movement end time minus
movement onset time.
Peak velocity (PV)—the highest velocity value reached during the prehension movement.
Time to peak velocity (TTPV)—the frame in which PV was
reached.
Peak grip aperture (PGA)—the maximum distance found
between the index finger and thumb markers during the
movement.
Time to peak grip aperture (TTPGA)—the frame in which
PGA was reached.
Deceleration time—movement end time minus TTPV.

3.2.1.3. Time to peak velocity. Across the experiment, the
left hand took longer to reach peak velocity than the right hand
(left = 407 ms, right = 311 ms, F(1, 79) = 27.1, p < 0.0001).
A Hand × Condition interaction was marginally significant
(F(2, 79) = 3.0, p = 0.06) with unimanual movements largely
accounting for the difference in the hands. As can be seen
from Fig. 5, TTPV for bimanual incongruent movements

The individual parameter values calculated in all the analyses were treated as individual subjects. Table 1 shows the
mean kinematic parameters for each hand in all conditions.
An independent subjects 2 × 2 × 3 ANOVA was conducted
containing the following factors: hand (left, right); distance
Table 1
Mean kinematic values for Experiment 1
Unimanual
Near

Duration (ms)
Peak velocity (mm/s)
Time to peak velocity (ms)
Deceleration time (ms)
Peak grip aperture (mm)
Time to peak grip aperture (ms)

Far

Bimanual—congruent

Bimanual—incongruent

Near

Near

Far

Far

Left

Right

Left

Right

Left

Right

Left

Right

Left

Right

Left

Right

676
503
396
294
145
334

575
511
260
292
123
400

834
680
419
446
141
407

636
833
268
374
125
465

769
477
427
494
141
436

734
429
264
623
122
488

949
614
426
523
143
521

974
609
333
641
129
658

753
439
355
414
146
436

769
371
350
419
129
531

931
578
425
469
147
583

814
646
354
450
128
581

1552

T.D. Punt et al. / Neuropsychologia 43 (2005) 1547–1558

Fig. 4. Mean peak velocity values across the three conditions in Experiment 1.

tended to be well coupled. No other main effects or interactions approached significance.
3.2.1.4. Deceleration time. There was a significant main
effect for condition (F(2, 95) = 9.1, p < 0.0005). Contrasts
showed that unimanual movements (353 ms) had significantly shorter deceleration times than bimanual congruent
movements (572 ms, F(1, 95) = 18.3, p < 0.0001) but not
bimanual incongruent movements (439 ms, F(1, 95) = 3.0,
p = 0.09). Bimanual incongruent movements had significantly shorter deceleration times than bimanual congruent
movements (F(1, 95) = 6.5, p < 0.05). None of the other main
effects or interactions reached statistical significance.
3.2.1.5. Temporal coupling of the transport phase. In addition to the above, we also examined the coupling of the
hands on bimanual trials for movement onset and movement
endpoint both in terms of the ‘lag’ (left hand–right hand)

and the correlation between the limbs. For Experiment 1,
the mean movement onset lag was −74 ms demonstrating
that the left hand tended to initiate the movement before the
right hand. However, the time lag was relatively small and the
movement onsets were ostensibly well coupled. Furthermore,
movement onset lag did not differ significantly between the
two bimanual conditions (means: congruent = −67 (15) ms;
incongruent = −82 (14) ms; F(1, 26) < 1.0, p = 0.5). In addition, a significant correlation was found between the left and
right hands for movement onset (r = 0.9, p < 0.0001).
The mean movement endpoint lag was −38 ms again suggesting close coupling between the limbs. Lags did not vary
significantly between conditions (means: congruent = −49
(33) ms; incongruent = −27 (30) ms; F(1, 35) < 1.0, p = 0.6)
and were significantly correlated (r = 0.8, p < 0.0001).
For the transport phase, MM was clearly able to couple
her hands appropriately and she also scaled peak velocity
appropriately for the bimanual incongruent condition.

Fig. 5. Representative velocity profiles for the right and left hands for the incongruent conditions in Experiment 1. These profiles show the normal scaling of
peak velocity and the tight coupling of movement onset, time to peak velocity and movement end. Peak velocity for the left and right hands varies according
to which object is more distant.

T.D. Punt et al. / Neuropsychologia 43 (2005) 1547–1558

1553

3.2.2. Kinematics for the grasp phase
3.2.2.1. Peak grip aperture. There was a significant main
effect for hand (F(1, 105) = 142.4, p < 0.0001), the left
hand (mean = 144 mm) shaping larger than the right hand
(mean = 126 mm). No other main or interaction effects
reached significance.
3.2.2.2. Time to peak grip aperture. The left hand reached
peak grip aperture significantly earlier than the right hand
leading to a main effect for hand (left hand = 454 ms,
right hand = 530 ms; F(1, 78) = 7.8, p < 0.01). Movements to
near targets reached peak grip aperture earlier than those
to far targets (near = 438 ms, far = 540 ms; F(1, 78) = 16.5,
p < 0.0005). There was also a main effect for condition (F(2,
78) = 12.3, p < 0.0001). Contrasts revealed that in the unimanual condition (mean = 401 ms) the peak grip aperture was
reached significantly faster than in both the bimanual conditions (congruent = 537 ms; F(1, 78) = 17.1, p < 0.0001; incongruent = 531 ms; F(1, 78) = 20.4, p < 0.0001). There was no
significant difference between the two bimanual conditions
(F(1, 78) < 1.0, p = 0.8). There were no significant interactions for this parameter.
3.3. Discussion
Despite her initial difficulties in performing bimanual
movements after her stroke, at the time of testing MM was
able to couple bimanual movements in terms of movement
onset and movement duration. Furthermore, she showed appropriate uncoupling or independence of peak velocity for
incongruent bimanual movements when the targets were positioned at different distances. These results are in line with
previous findings of bimanual coupling in normal subjects
(Jackson et al., 1999; Kelso et al., 1979). However, MM is
quite different from patient DB, a patient with contralesional
sensory impairment as a result of a right anterior parietal
lesion who tended to couple peak velocity inappropriately
for incongruent bimanual movements (Jackson et al., 2000).
Jackson et al. (2000) argue that proprioceptive cues were
not available for DB’s left hand, and that the inappropriate
movement coupling reflects dominance for the ipsilesional
limb under those conditions. In contrast to DB, MM was able
to perform incongruent bimanual movements (in terms of
distance) relatively normally.

4. Experiment 2
Here, we investigated MM’s ability to show inter-limb
independence for grasp scaling by manupulating object size.

Fig. 6. Set-up for Experiment 2. MM made unimanual reaches to either
small or large targets, bimanual congruent reaches (small/small; large/large)
or bimanual incongruent reaches (small/large; large/small).

ter) or large targets (6 cm diameter) (see Fig. 6). The distance
of the targets was fixed at 30 cm from the hand starting position and MM again completed 10 repetitions for each of the
eight conditions (left small, left large, right small, right large,
bimanual congruent small, bimanual congruent large, bimanual incongruent left small right large, bimanual incongruent
left large right small). The conditions were randomised over
trials.
4.2. Results
Again, the individual parameter values calculated in all the
analyses were treated as individual subjects. Table 2 shows
the mean kinematic parameters for each hand in all conditions. An independent subjects 2 × 2 × 3 ANOVA was conducted containing the following factors: hand (left, right);
size (small, large) and condition (unimanual, bimanual congruent, bimanual incongruent).
4.2.1. Kinematics for the transport phase
4.2.1.1. Movement duration. Movement duration did not
differ significantly for hand (left = 1089 ms, right = 1028 ms;
F(1, 98) = 1.9, p = 0.2) or for size (small = 1073 ms,
large = 1048 ms; F(1, 98) < 1.0, p = 0.4). There was a significant effect of condition (F(2, 98) = 36.2, p < 0.0001).
Contrasts showed that the unimanual movement durations
(804 ms) were significantly shorter than both bimanual conditions (congruent = 1184 ms, F(1, 98) = 51.6, p < 0.0001; incongruent = 1198 ms, F(1, 98) = 56.3, p < 0.0001) but the two
bimanual conditions had comparable movement durations
(F(1, 98) < 1.0, p = 0.8). None of the interactions reached significance.

4.1. Method
MM was asked to make unimanual or bimanual movements in much the same way as in Experiment 1. MM had to
reach unimanually or bimanually towards small (3 cm diame-

4.2.1.2. Peak velocity. Across the experiment, peak velocity of the right hand was higher than that for the left
hand (left = 604 mm/s, right = 690 mm/s; F(1, 101) = 17.6,
p < 0.0001). There was also a significant main effect

1554

T.D. Punt et al. / Neuropsychologia 43 (2005) 1547–1558

Table 2
Mean kinematic values for Experiment 2
Unimanual
Small

Duration (ms)
Peak velocity (mm/s)
Time to peak velocity (ms)
Deceleration time (ms)
Peak grip aperture (mm)
Time to peak grip aperture (ms)

Bimanual—congruent
Large

Small

Bimanual—incongruent
Large

Small

Large

Left

Right

Left

Right

Left

Right

Left

Right

Left

Right

Left

Right

879
658
322
557
113
468

691
828
223
468
107
353

894
659
317
577
138
414

729
874
247
481
122
516

1297
568
319
972
126
585

1220
589
330
888
106
644

1143
596
327
816
138
652

1086
624
321
812
128
733

1268
606
350
979
133
573

1143
591
345
794
109
721

1076
536
346
730
132
663

1326
639
386
921
124
695

of condition (F(2, 101) = 26.9, p < 0.0001). Contrasts revealed that unimanual movements had a significantly higher
peak velocity than each of the two bimanual conditions
(congruent, F(1, 101) = 40.1, p < 0.0001; incongruent, F(1,
101) = 40.4, p < 0.0001) but there was no difference between the two bimanual conditions (F(1, 101) < 1.0, p = 0.9).
There was a significant interaction of Hand × Condition
(F(1, 101) = 6.5, p < 0.005). The right hand had a significantly higher peak velocity than the left hand, but only in
the unimanual condition (left = 660 mm/s, right = 848 mm/s;
F(1, 34) = 35.4, p < 0.0001). Peak velocity was comparable
across the hands for the two bimanual conditions (congruent,
left = 582 mm/s, right = 606 mm/s, F(1, 34) < 1.0, p = 0.4; incongruent, left = 569 mm/s, right = 615 mm/s, F(1, 33) < 1.0,
p = 0.4). There were no other significant main effects or interactions.
4.2.1.3. Time to peak velocity. TTPV was equivalent for
the left and right hands (left = 330 ms, right = 305 ms;
F(1, 96) = 1.3, p = 0.3) and also for small and large targets (small = 313 ms, right = 325 ms; F(1, 96) < 1.0, p = 0.6).
There was a significant main effect of condition (F(2,
96) = 5.7, p < 0.005). Contrasts showed TTPV to be shorter
for unimanual movements (280 ms) than for movements in
the bimanual congruent condition (324 ms; F(1, 96) = 4.2,
p < 0.05) and in the bimanual incongruent condition (355 ms;
F(1, 96) = 10.9, p < 0.005); however, the two bimanual conditions were comparable (F(1, 96) = 1.8, p = 0.2). No other
main effects or interactions proved significant.
4.2.1.4. Deceleration time. The left and right hands did not
differ in the time taken between TTPV and movement end
(left = 768 ms, right = 729 ms, F(1, 99) < 1.0, p = 0.3), nor did
deceleration times for small and large targets (small = 769 ms,
large = 731 ms, F(1, 99) = 1.3, p = 0.3). There was a significant main effect of condition (F(2, 99) = 24.1, p < 0.0001).
Contrasts showed that deceleration time for the unimanual condition (524 ms) was significantly shorter than for
the bimanual congruent condition (870 ms, F(1, 99) = 37.9,
p < 0.0001) and the bimanual incongruent condition (856 ms,
F(1, 99) = 34.5, p < 0.0001). However, the two bimanual conditions were equivalent (F(1, 99) < 1.0, p = 0.8). There were
no other significant main effects or interactions.

4.2.1.5. Temporal coupling of the transport phase. Temporal coupling was analysed in the same way as it was in Experiment 1. For Experiment 2, mean movement onset lag
was −25 ms suggesting close coupling between the limbs.
Movement onset lag did not differ significantly between the
two bimanual conditions (means: congruent = −19 (13) ms;
incongruent = −32 (14) ms; F(1, 33) < 1.0, p = 0.5). In addition, a significant correlation was found between the left and
right hands for movement onset (r = 0.9, p < 0.0001).
The mean movement endpoint lag was −11 ms showing
close coupling between the limbs. Lags did not vary significantly between conditions (means: congruent = 43 (48) ms;
incongruent = −66 (47) ms; F(1, 37) = 2.6, p = 0.1) and were
significantly correlated (r = 0.7, p < 0.0001).
In summary, for the transport phase in Experiment 2, MM
showed appropriate coupling of the limbs with bimanual
movements demonstrating the typical ‘costs’ associated with
bimanual movements compared with unimanual movements.
4.2.2. Kinematics for the grasp phase
4.2.2.1. Peak grip aperture. The ANOVA revealed a significant main effect of hand (F(1, 106) = 89.6, p < 0.0001) with
the left hand shaping larger than the right (left = 130 mm,
right = 116 mm). Grip apertures to large targets were significantly greater than to small targets (small = 116 mm,
large = 130 mm, F(1, 106) = 93.5, p < 0.0001). There was
also a significant main effect of condition F(2, 106) = 3.9,
p < 0.05). Contrasts revealed that peak grip aperture was
narrower for the unimanual condition (119 mm) than the
bimanual congruent condition (125 mm, F(1, 106) = 6.1,
p < 0.05) and the bimanual incongruent condition (124 mm,
F(1, 106) = 5.5, p < 0.05). However, the two bimanual conditions were equivalent (F(1, 106) < 1.0, p = 0.9). In addition, there was a significant two-way interaction of
Size × Condition (F(2, 106) = 7.3, p < 0.005), and a significant three-way interaction for Hand × Size × Condition
(F(2, 106) = 6.8, p < 0.005). The difference in behaviour between the hands becomes clear when the interactions are
explored fully. For the right hand, while there is a significant difference in scaling to different sized objects (F(1,
53) = 83.5, p < 0.0001), this scaling is consistent across conditions (F(2, 53) < 1.0, p = 0.5) and there was no significant Size × Condition interaction (F(2, 53) = 1.3, p = 0.3).

T.D. Punt et al. / Neuropsychologia 43 (2005) 1547–1558

1555

Fig. 7. PGA values (and error bars) for the left and right hands. While the right hand remains relatively stable across conditions, the left (contralesional) hand
shows a dependency on the right (ipsilesional) target.

However, for the left hand, while again there was a significant main effect of size (F(1, 53) = 26.1, p < 0.0001), this
scaling changed as a function of condition (F(2, 53) = 3.6,
p < 0.05) and there was a significant Size × Condition interaction (F(2, 53) = 10.6, p < 0.0005). If the Size × Condition
interaction for each hand is explored further, it can be
seen that the right hand demonstrates significantly different scaling to small and large objects for each condition (unimanual, small = 106.9 mm, large = 122.0 mm, F(1,
19) = 18.2, p < 0.005; bimanual congruent, small = 106.3 mm,
large = 127.5 mm, F(1, 18) = 39.9, p < 0.0001; bimanual incongruent, small = 109.0 mm, large = 123.6 mm, F(1,
18) = 28.6, p < 0.0001). The left hand also showed a
significant size effect for unimanual (small = 112.8 mm,
large = 138.4 mm, F(1, 17) = 35.2, p < 0.0001) and bimanual

congruent movements (small = 126.5 mm, large = 138.2 mm,
F(1, 18) = 6.6, p < 0.025) but this was lost for bimanual incongruent movements (small = 133.0 mm, large = 131.9 mm,
F(1, 18) < 1.0, p = 0.7) (see Fig. 7). Grip profiles from representative trials of the bimanual incongruent condition are
shown in Fig. 8.
4.2.2.2. Time to peak grip aperture. The time from movement onset to when peak grip aperture was reached was equivalent for each hand (left = 0.56 s, left = 0.61 s, F(1, 99) = 2.1,
p = 0.2) and for the two sizes of target (small = 0.55 s,
large = 0.62 s, F(1, 99) = 2.4, p = 0.1). There was a significant main effect of condition (F(2, 99) = 16.7, p < 0.0001).
Contrasts revealed that MM was significantly faster to
reach peak grip aperture in the unimanual condition (0.43 s)

Fig. 8. Representative grip aperture for the left and right hands for the incongruent conditions in Experiment 2. While the right hand scales normally, the left
hand shows an abnormal dependence related to the target grasped by the right hand.

1556

T.D. Punt et al. / Neuropsychologia 43 (2005) 1547–1558

than in both the bimanual congruent condition (0.65 s,
F(1, 99) = 24.1, p < 0.0001) and the bimanual incongruent condition (0.66 s, F(1, 99) = 26.2, p < 0.0001). However, the two bimanual conditions were equivalent (F(1,
99) < 1.0, p = 0.8). None of the interactions approached significance.
In summary, MM showed appropriate scaling for unimanual movements to the small and large object, with both the left
and right hands, though the left hand scaled wider. There was
also appropriate scaling of grasp in the bimanual congruent
condition. However, the interaction between the hands broke
down in the bimanual incongruent condition. The left hand
was modulated by the grip aperture adopted for the right hand
while the right hand grip aperture showed no effect of the left
hand.
4.3. Discussion
In Experiment 2 as in Experiment 1, MM showed relatively normal coupling for the transport phase of bimanual
movements. However, here we were primarily interested in
whether MM was able to scale the contralesional hand appropriately to the size of the target. In this respect, the experiment revealed a number of interesting findings. Firstly,
and most noticeably, the contralesional hand did not demonstrate the same pattern of performance as the ipsilesional
hand across conditions. It is tempting to explain the results
in terms of a contralesional deficit alone. However, closer
inspection shows that the relative scaling of each hand is
more illuminating. On trials involving incongruent bimanual movements, the contralesional left hand demonstrated an
influence of the ipsilesional hand; relative to the unimanual
baseline, the grip aperture was smaller for large objects and
larger for small objects. In contrast, there was no modulation
of grasp aperture for the ipsilesional hand. This represents
a significant shift away from the behaviour seen in normal
subjects where both hands share the cost of bimanual activity (Castiello et al., 1993; Dohle et al., 2000; Jackson et al.,
1999). The bimanual actions of MM here are indicative of
an attentional bias in favour of the ipsilesional target and/or
the ipsilesional hand; although the contralesional hand acts
appropriately under unimanual movement conditions, grasp
programming is unduly influenced by the ipsilesional action
being conducted concurrently. This may be similar to the inappropriate coupling of the transport component of prehension reported for patient DB by Jackson et al. (2000). In DB’s
case, the transport component of movement made by the contralesional limb was coupled to that made by the ipsilesional
limb even when the reached-for objects were at different distances. In MM’s case, there was a complete lack of an effect
of the grip aperture of the contralesional hand on the ipsilesional hand, while the ipsilesional hand strongly influenced
the contralesional one. This suggests a classic ‘winner takes
all’ situation where the ipsilesional hand wins the selection
of the parameters for the grasp action

5. General discussion
We investigated bimanual coordination in a patient, MM,
who had motor neglect and extinction following stroke. MM
demonstrated an interesting pattern of performance when required to make bimanual reach-to-grasp movements to objects. When the objects were at different distances, MM
showed appropriate uncoupling of the ‘reach’ components
of her actions. For example, there was a cross-over of peak
velocities with the conralesional (left) hand showing higher
peak velocity than the ipsilesional (right) hand when the left
object was far and the right near; the opposite occurred when
the distances of the objects reversed. There was also some
‘cost’ to this, since peak velocities for both hands were lower
for bimanual than for unimanual movements. In contrast, a
different pattern arose when the objects were at the same
distance but were different sizes. The grasp components of
the actions were not uncoupled and there was dominance of
the ipsi-(right) over the contralesional (left) hand. The ipsilesional hand was scaled appropriately to the size of the
object, but this did not change as a function of whether the
other object was the same or a different size. However, the
contralesional hand was not scaled appropriately. Grip aperture to a small object was wider than in the unimanual and
congruent bimanual condition; grip aperture to a large object
was smaller than in the unimanual and congruent bimanual
conditions.
Since the impairment in the grasp component emerged in
bimanual but not unimanual conditions, it may be considered
as part of the syndrome of motor extinction. Motor extinction has only a limited history of study but has been shown to
adversely affect contralesional movement initiation (hypokinesia) (Valenstein & Heilman, 1981) and the sustaining of
movement (Mattingley, 2002). It is conceivable that, if an attentional bias in movement can lead to contralesional hypokinesia and hypometria, then other aspects of motor production
might be similarly affected. The present data suggest that aspects of grasp production can be affected, while the reach
components of an action are relatively spared. The deficit in
the grasp component indicates a bias in motor programming,
with greater weight applied to this component of the ipsicompared with the contralesional motor programme, when
both are effected simultaneously. The problem was greater
for the grasp than the reach component of MM’s actions, perhaps representing a selective deficit in the grasp component.
This would fit with Jeannerod’s (1981) original distinction
between reaching and grasping actions.
The selective effect of the deficit on the grasp component of actions fits with a ‘motor competition’ account and it
presents some difficulties to some other accounts, suggesting
that performance is affected by a more general attentional
bias. For example, one alternative account is that the deficit
is induced by a bias in visual attention to the object on the
ipsilesional side. Such an attentional bias could arise if motor programmes feedback to influence visual attention (cf.
Deubel & Schneider, 2004; Humphreys et al., 2004), and

T.D. Punt et al. / Neuropsychologia 43 (2005) 1547–1558

there is stronger activation of a motor response with the ipsilesional limb. Due to this stronger motor activation, MM
might orient to the ipsilesional side, and the object on that
side then dominates selection of the appropriate grip. In simple terms, however, this account would also predict an affect
on the transport component of the action, which was not supported here. The pattern of this deficit on the grasp component
specifically under bimanual movement conditions, provides a
form of double dissociation when considered in conjunction
with the data from DB (Jackson et al., 2000). DB also showed
a deficit in bimanual action, but this affected the transport
component of reaching. However, Jackson et al. did not test
DB on bimanual movements to incongruently sized objects
as in Experiment 2 here. It is, therefore hard to tell if contralesional grip was abnormally affected by the ipsilesional side.
It could be that both were affected in this case. MM shows
a selective grip affect but this may reflect the more difficult
component, holding a higher position in the hierarchy of motor programming. Nevertheless the attentional account can
be maintained if we distinguish between the affects of attention on the grasp and transport components of movement. For
example, if attention influences the relatively late stages of
action, then it might exert a greater effect on grip aperture as
this may be programmed later in the reach and grasp action
(Jeannerod, 1981).
Whichever account of MM’s performance is favoured, the
data demonstrate that there can be differential effects of a
brain lesion on bimanual motor coupling. The coupling may
be relatively intact for reach components of action, but not
for the grasp component. The data emphasise the distinction
between these two components of motor programming.

Acknowledgements
This work was supported by grants from The Stroke Association to all three authors, and by grants from the MRC
and Wellcome Trust to Jane Riddoch and Glyn Humphreys.
We gratefully acknowledge this support.

References
Bootsma, R. J., Marteniuk, R. G., MacKenzie, C. L., & Zaal, F. T. (1994).
The speed-accuracy trade-off in manual prehension: Effects of movement amplitude, object size and object width on kinematic characteristics. Experiments Brain Research, 98, 535–541.
Castiello, U., Bennett, K. M. B., & Stelmach, G. E. (1993). The bilateral
reach to grasp movement. Behavioural Brain Research, 56, 43–57.
Chieffi, S., Gentilucci, M., Allport, A., Sasso, E., & Rizzolatti, G. (1993).
Study of selective reaching and grasping in a patient with unilateral
parietal lesion. Dissociated effects of residual spatial neglect. Brain,
116(Pt 5), 1119–1137.
Deubel, H., & Schneider, W. X. (2004). Attentional selection in sequential
manual movements, movements around an obstacle and in grasping.
In G. W. Humphreys & M. J. Riddoch (Eds.), Attention in action.
Hove: Psychology Press.

1557

Dohle, C., Ostermann, G., Hefter, H., & Freund, H. J. (2000). Different
coupling for the reach and grasp components in bimanual prehension
movements. Neuroreport, 11, 3787–3791.
Driver, J., & Vuilleumier, P. (2001). Perceptual awareness and its loss in
unilateral neglect and extinction. Cognition, 79, 39–88.
Fitts, P. M. (1954). The information capacity of the human motor system in controlling the amplitude of movement. Journal Experimental
Psychology, 47, 381–391.
Gado, M., Hanaway, J., & Frank, R. (1979). Functional anatomy of the
cerebral cortex by computed tomography. Journal of Computer Assisted Tomography, 3, 1–19.
Humphreys, G. W., Riddoch, M. J., Punt, T. D., Linnell, K. J., Edwards,
M. G., & Wing, A. M. (2004). Attending to what you are doing:
Neuropsychological and experimental evidence for interactions between perception and action. In G. W. Humphreys & M. J. Riddoch
(Eds.), Attention in action. Hove: Psychology Press.
Jackson, G. M., German, K., & Peacock, K. (2002). Functional coupling
between the limbs during bimanual reach-to-grasp movements. Human
Movement Science, 21, 317–333.
Jackson, G. M., Jackson, S. R., Husain, M., Harvey, M., Kramer, T., &
Dow, L. (2000). The coordination of bimanual prehension movements
in a centrally deafferented patient. Brain, 123(Pt 2), 380–393.
Jackson, G. M., Jackson, S. R., & Kritikos, A. (1999). Attention for action: Coordinating bimanual reach-to-grasp movements. British Journal of Psychology, 90(Pt 2), 247–270.
Jeannerod, M. (1981). Intersegmental coordination during reaching at natural visual objects. In J. Long & A. Baddeley (Eds.), Attention and
performance IX (pp. 153–169). Hillsdale: Lawrence Erlbaum Associates.
Jeannerod, M., Decety, J., & Michel, F. (1994). Impairment of grasping
movements following a bilateral posterior parietal lesion. Neuropsychologia, 32, 369–380.
Kelso, J. A., Putnam, C. A., & Goodman, D. (1983). On the space-time
structure of human interlimb co-ordination. The Quarterly Journal of
Experimental Psychology A, 35(Pt 2), 347–375.
Kelso, J. A., Southard, D. L., & Goodman, D. (1979). On the coordination of two-handed movements. Journal of Experimental Psychology.
Human Perception and Performance, 5, 229–238.
Mattingley, J. B. (2002). Spatial extinction and its relation to mechanisms of normal attention. In H. O. Karnath, A. D. Milner, & G.
Vallar (Eds.), The cognitive and neural bases of spatial neglect (pp.
289–309). Oxford: Oxford University Press.
Mattingley, J. B., & Driver, J. (1997). Distinguishing motor and sensory
deficits after parietal damage: An evaluation of response selection biases in unilateral neglect. In P. Thier & H.-O. Karnath (Eds.), Parietal
lobe contributions to orientation in 3D space (pp. 309–337). Berlin:
Springer-Verlag.
Meador, K. J., Watson, R. T., Bowers, D., & Heilman, K. M. (1986).
Hypometria with hemispatial and limb motor neglect. Brain, 109,
293–305.
Monno, A., Temprado, J. J., Zanone, P. G., & Laurent, M. (2002). The
interplay of attention and bimanual coordination dynamics. Acta Psychologica (Amsterdam), 110, 187–211.
Posner, M. I., Inhoff, A. W., & Friedrich, F. J. (1987). Isolating attentional systems—A cognitive-anatomical analysis. Psychobiology, 15,
107–121.
Robertson, I. H., McMillan, T. M., MacLeod, E., Edgeworth, J., &
Brock, D. (2002). Rehabilitation by limb activation training reduces
left-sided motor impairment in unilateral neglect patients: A singleblind randomised control trial. Neuropsychological Rehabilitation, 12,
439–454.
Serrien, D. J., Nirkko, A. C., Lovblad, K. O., & Wiesendanger, M. (2001).
Damage to the parietal lobe impairs bimanual coordination. Neuroreport, 12, 2721–2724.
Swinnen, S. P., Jardin, K., & Meulenbroek, R. (1996). Between-limb
asynchronies during bimanual coordination: Effects of manual dominance and attentional cueing. Neuropsychologia, 34, 1203–1213.

1558

T.D. Punt et al. / Neuropsychologia 43 (2005) 1547–1558

Valenstein, E., & Heilman, K. M. (1981). Unilateral hypokinesia and
motor extinction. Neurology, 31, 445–448.
Vallar, G., Bottini, G., & Sterzi, R. (2003). Anosognosia for left-sided
motor and sensory deficits, motor neglect, and sensory hemiinattention: Is there a relationship? Progress in Brain Research, 142, 289–
301.

Wilson, B., Cockburn, J., & Halligan, P. (1987). Development of a behavioral test of visuospatial neglect. Archives of Physical Medicine
Rehabilitation, 68, 98–102.
Wuyts, I. J., Summers, J. J., Carson, R. G., Byblow, W. D., & Semjen, A.
(1996). Attention as a mediating variable in the dynamics of bimanual
coordination. Human Movement Science, 15, 877–897.