Neurop~ycholoyw. Vol 33. No 5. pp 611422, 1995 Copyright q 1995 Elsewer Science Ltd

Prmted III Great Br~tam. All rights reserved 002&3932/95 $9 50+0 00

0028-3932(95)00005-4

RECALL OF SELF-GENERATED ARM MOVEMENTS BY PATIENTS WITH UNILATERAL CORTICAL EXCISIONS

GABRIEL LEONARD* and BRENDA MILNER

Department of Neurology and Neurosurgery, McGill University and the Montreal Neurological Institute. 3801 University Street, Montreal, Quebec, Canada H3A 284

(Received 2 August 1994; accepted 21 November 1994)

Abstract-In previous work, Leonard and Mimer (Neuropsychologia, 29,47758,199l) demonstrated that patients with large excisions from the right frontal lobe have difficulty in reproducing accurately the extent of examiner-defined arm movements, the displacements being made without the aid of vision. The impairment in the right frontal-lobe group was not dependent on recall-condition, being apparent irrespective of the presence of a delay, suggesting that the deficit was primarily one of encoding. We then went on to show that these same patients have a short-term memory deficit when recalling terminal position of examiner-defined arm movements (Neuropsychologia 29, 6299640, 1991). From these investigations we concluded that the right frontal lobe is critically involved in the monitoring of information related to movement. In the present study 58 patients with unilateral temporal- or frontal-lobe excisions were tested on two kinesthetic tasks that required the subjects themselves to select terminal positions, or movement extents, thereby reducing dependence on peripheral feedback. Patients with right frontal-lobe lesions could reproduce these self-generated movements normally, indicating that when demands on feedback are reduced the frontal-lobe contribution is not critical.

Key Words: frontal-lobe; arm movements.

INTRODUCTION

Leonard and Milner [23] have demonstrated that patients with large excisions from the right frontal lobe are deficient in reproducing the extent of examiner-defined arm movements, starting-position being changed on each trial, and vision being excluded. This loss was manifest on both arms. Patients with left frontal-lobe excisions, whether small or large, were unimpaired, and this was also true for patients who had had an anterior temporal-lobe resection from either hemisphere, regardless of the extent of encroachment upon the hippocampus and parahippocampal gyrus. The deficit shown by the right frontal group was apparent in immediate recall and was not disproportionately exacerbated by delay, indicating that the difficulty was primarily one of encoding and/or retrieval, rather than a retention deficit per se. It was concluded from this study that the right frontal lobe is critically involved in the monitoring of information related to limb movement.

That the deficit after right frontal lobectomy was not restricted to the recall of distance traversed was shown by a further study [24], which examined how well patients with frontal- lobe lesions could encode and recall the end-position of simple arm movements under

*Address for correspondence: Gabriel Leonard, Department of Neuropsychology, Montreal Neurological Institute, 3801 University Street, Montreal, Quebec, Canada H3A 2B4.

611

Pergamon Neurop~ychologza. Vol 33, No 5, pp 611~522, 1995

Copyright ~ 1995 Elsevier Science Ltd Printed m Great Britain. All rights reserved

0028-3932/95 $9 50+0 00

0028-3932(95)00005-4

R E C A L L O F S E L F - G E N E R A T E D A R M M O V E M E N T S BY P A T I E N T S W I T H U N I L A T E R A L C O R T I C A L E X C I S I O N S

GABRIEL LEONARD* and BRENDA MILNER

Department of Neurology and Neurosurgery, McGill University and the Montreal Neurological Institute, 3801 University Street, Montreal, Quebec, Canada H3A 2B4

(Received 2 August 1994; accepted 21 November 1994)

Abstract--In previous work, Leonard and Milner (Neuropsychologia, 29, 47--58, 1991 ) demonstrated that patients with large excisions from the right frontal lobe have difficulty in reproducing accurately the extent of examiner-defined arm movements, the displacements being made without the aid of vision. The impairment in the right frontal-lobe group was not dependent on recall-condition, being apparent irrespective of the presence of a delay, suggesting that the deficit was primarily one of encoding. We then went on to show that these same patients have a short-term memory deficit when recalling terminal position of examiner-defined arm movements (Neuropsychologia 29, 629-640, 1991 ). From these investigations we concluded that the right frontal lobe is critically involved in the monitoring of information related to movement. In the present study 58 patients with unilateral temporal- or frontal-lobe excisions were tested on two kinesthetic tasks that required the subjects themselves to select terminal positions, or movement extents, thereby reducing dependence on peripheral feedback. Patients with right frontal-lobe lesions could reproduce these self-generated movements normally, indicating that when demands on feedback are reduced the frontal-lobe contribution is not critical.

Key Words: frontal-lobe; arm movements.

INTRODUCTION

Leonard and Milner 1,23] have demonstrated that patients with large excisions from the right frontal lobe are deficient in reproducing the extent of examiner-defined arm movements, starting-position being changed on each trial, and vision being excluded. This loss was manifest on both arms. Patients with left frontal-lobe excisions, whether small or large, were unimpaired, and this was also true for patients who had had an anterior temporal-lobe resection from either hemisphere, regardless of the extent of encroachment upon the hippocampus and parahippocampal gyrus. The deficit shown by the right frontal group was apparent in immediate recall and was not disproportionately exacerbated by delay, indicating that the difficulty was primarily one of encoding and/or retrieval, rather than a retention deficit per se. It was concluded from this study that the right frontal lobe is critically involved in the monitoring of information related to limb movement.

That the deficit after right frontal lobectomy was not restricted to the recall of distance traversed was shown by a further study 1-24], which examined how well patients with frontal- lobe lesions could encode and recall the end-position of simple arm movements under

*Address for correspondence: Gabriel Leonard, Department of Neuropsychology, Montreal Neurological Institute, 3801 University Street, Montreal, Quebec, Canada H3A 2B4.

611

612 G. LEONARD and B MILNER

conditions in which the distance traversed offered no useful cue. In this case, the group with large removals from the right frontal lobe was again impaired, but now the deficit was only apparent after a delay. Patients in both temporal-lobe groups with extensive medial removal were also impaired, but only after a filled recall interval. This latter finding is in contrast to the frontal-lobe deficit, which was apparent following a 30-see delay, regardless of whether or not a distracting task was performed during the interval.

In the present investigation, the paradigm was changed to permit subjects to form their own internal representation of the extent or end-position of the arm displacement, thereby diminishing dependence on feedback from external sensory receptors. This was achieved by having subjects select in advance either distance or end-position before making a response, and then examining them on their ability to recall these self-generated responses. For the distance task, because reliance on feedback from peripheral receptors was reduced, it was predicted that patients with frontal lobe lesions would perform normally. It remained an open question as to how such patients would perform on the preselected location task.

For these studies, subjects were required to vary their choices from trial to trial. Given the fluency deficits frequently associated with frontal-lobe lesions [2, 12, 14, 29], a further factor to consider was whether these patients would be able to comply with the instruction to vary their selections and not merely perseverate on their initial responses.

METHOD Subjects

The 58 patients tested had each undergone a unilateral cortical excision at the Montreal Neurological Hospital. The operations were carried out for the relief of focal cerebral seizures. In most instances the epileptogenic lesion dated from birth or early life, and was static in nature; however, 11 cases of indolent tumour were also included. Patients who had independent electrographic abnormality arising from both hemispheres, or who had evidence of fast-growing tumours or diffuse cerebral damage, were excluded, as were those known to have right-sided, or bilateral, speech representation, as demonstrated by preoperative intracarotid sodium-amobarbital tests [-3]. Only subjects between the ages of 16 and 60 with Full-Scale Wechsler IQ ratings above 80 were studied.

The hand preference of each subject had been assessed by means of a modified version of the questionnaire developed by Crovitz and Zener [6]. On this scale, a score of 29 or less signifies a strong preference for the right hand [30]; all subjects had a score below thus cut-off point.

In order to ensure that subjects had no motor or sensory deficits that could have interfered with their test performance, the strength and the somatosensory status of the hands were established beforehand. Grip-strength was measured by having the subjects squeeze a wooden handle connected to a Dillon Tensile-Force Gauge graduated in pounds [44]. Three readings were taken for each hand in a balanced order, beginning with the hand ipsilateral to the lesion. For men, scores below 75 lb for the right hand, and 74 lb for the left hand, resulted in exclusion from the study. The corresponding scores for women were 50 lb for the right hand and 49 lb for the left hand. These cut-offscores were defined as being two standard deviations below the mean for normal control groups, matched to the patients with respect to age, education, and socioeconomic status [22]. The sensory status of the hands was determined by quantitative testing [7]; the measures used included two-point discrimination on the palm and sense of passive movement of the fingers. All patients obtained normal scores for both hands on these tests and on a test of tactual object recognition [-7].

Subjects were assigned to eight groups (LTh, LTH, RTh, RTH, LF, RFL, RFS and NC), as shown in Table 1. Also given in Table 1 are the sex distribution, mean age (at time of testing), and mean education for all groups and the mean Wechsler Full-Scale IQ for the patient groups. Separate one-way analyses of variance did not reveal any significant difference between the groups with respect to age [ F = 1.12, d.f. = 7, 61, P > 0.36], or education [ F = 1.39, d.f. = 7, 61, P > 0.22]. Nor did the patient groups differ statistically from one another with respect to Wechsler IQ [ F = 1.22, d.f .=6, 51, P>0.31] .

Temporal-lobe groups The patients within the temporal-lobe groups were subdivided according to the extent of removal from the

hippocampal region. This method of classification resulted in four groups: two left temporal, one (LTh) with small and one (LTH) with large removals from the hippocampal region, and two right temporal-lobe groups, one (RTh) with small and one (RTH) with large hippocampal excisions. The hippocampal removals were classified as small or

SELF-GENERATED ARM MOVEMENTS

Table 1. Main subject groups

613

Sex Age (year) Education (year) Wechsler IQ Group M F Mean Range Mean Range Mean Range

Left temporal --small hippocampal removal 7 5 30 6 19-37 14.6 12 18 106.5 89-135 Left temporal --large hlppocampal removal 5 4 21.7 15-29 12.7 11-16 113.1 98-131 Left frontal 3 2 29.4 14~40 12 2 7-18 95.8 82-112 Right temporal --small hippocampal removal 7 6 30.9 20-48 13.0 9 16 110.8 96-129 Right temporal --large hippocampal removal 1 6 25.7 15 32 12.5 9-19 109.2 100-115 Right frontal --small removal 5 1 32.8 26-49 11.6 7- 14 102.2 94 132 Right frontal --large removal 4 2 33.2 27-43 10.6 3-14 102.6 88-119 Normal control 6 5 24.8 19-29 14.4 12-17 Not assessed

large on the basis of the surgeon's drawing and report at the time of operation. The removal was considered to be small if no more than the pes of the hippocampus was excised; the hippocampal removals were classified as large if the excision encroached further upon the body of the hippocampus or the parahippocampal gyrus. The removals in all cases included the anterior temporal neocortex and the amygdala. Four of these patients had low-grade astrocytomas (two m the left hemisphere and two in the right).

Frontal-lobe groups For the patients with frontal-lobe resection, the extent as well as the side of the removal was considered important,

because, in our previous studies of kinesthetic memory for distance traversed 1-23] and end position [24], only patients with large right frontal-lobe lesions had shown a deficit. Since the left hemisphere was dominant for speech in all cases, the removals from the left frontal lobe tended to be more conservative than those from the right, and only two patients had large removals (their performance was similar to that of the patients with smaller removals); therefore, the subjects who had left frontal-lobe lesions were assigned to one group (LF). For the present study it was decided to maintain the same groupings. There were five patients who had had left frontal-lobe removals, three of whom had large and two who had small removals. The patients with right frontal-lobe excisions had roughly equal numbers of subjects with small (RFS) and with large (RFL) resections of frontal cortex. The classification of lesions into small and large was carried out by an independent judge.

Left frontal-lobe group. The extent of the removals in the five patients in this group is shown in Fig. 1. Two subjects (Ho. Sp., Ke. Ke.) had small removals, and the other three had large removals. Three patients (Lo. La., Dw. Mi., Ho. Sp.) had low-grade astrocytomas, and one patient (Ke. Ke.) had an arteriovenous malformation.

Riyhtfrontal-lobe group with small removals. Figure 2 shows (on the left) the cortical excisions for the six patients in this group, three of whom (Ro. Bo., Br. Fe., and Br. Fo.) had low-grade astrocytomas.

Riohtfrontal-lobe oroup with large removals. Shown on the right side of Fig. 2 are the cortical excisions for the six patients m the RFL group. One patient, Ru. Ma., had a low-grade astrocytoma.

Normal control subjects. A normal control group (NC) of 11 subjects was selected to match the patient groups as far as possible with respect to sex, age and level of education (see Table 1).

Apparatus and procedure

The apparatus used was a slightly modified version of the Manual Lever D, described by Sulzer [45]. The top drawing in Fig. 3 shows the examiner's view of the apparatus, which consisted of a moveable lever (0.9 cm in diameter) mounted on a fiat board. The radius of the lever was 30 cm, as measured from the centre of the shaft to the top of the grip. The total possible range of movement was 140 ° (73.3 cm of arc), and the scale, which was attached to the examiner's side of the instrument, was marked in one-degree steps.

The subject was seated on an adjustable chair, the head being supported by a chin-rest that was mounted on the testing table. Attached to the chin-rest was an opaque screen that prevented the subject from seeing his hand or arm (see Fig. 3).

At the beginning of the first testing session, the subject was informed that the experiment would revolve movements of the arm and hand. It was explained that the test would be run with vision excluded, and that body and head position would be kept constant by the use of the chin-rest.

The subject began the test by placing the hands, palms down, on the table. The position of the apparatus was then

614 G. LEONARD and B. MILNER

Dw Ui Lo. La.

No ~*

Fig. 1. Diagrams based on the surgeon's drawings at the time of operation, showing the estimated extent of the cortical excisions for subjects in the left frontal-lobe group. For all figures illustrating frontal-lobe resections, the medial view (above) and the inferior view (below) will be included

whenever available, together with the lateral view.

adjusted so that no movements of either hand would cross the midline of the subject's body. The examiner guided the subject's hand to the lever, with the instruction that it be held in a full grip (see Fig. 3). The subject was then allowed to move the lever until the complete range of possible movement had been explored. It was required that movements be made smoothly and at a constant speed, as demonstrated by the examiner. The movements were always to be made away from the subject's body, as shown for the right hand in Fig. 3.

For the reproduction of both distance and location, four recall-conditions were employed: 15 sec and 30 sec unfilled interval; 15 sec and 30 sec with an interpolated counting task. In both experiments the four conditions were randomly distributed over the testing sessions, with the constraint that two of the same recall-conditions could not follow each other more than once.

For the distance task, subjects were required to generate arm movements of different lengths, and then, following a delay, to reproduce the distance traversed. For end-position, the subject selected end-points for different arm movements, and then recalled these "self-selected" stopping points following a delay.

Instructions for the distance task were as follows:

I am going to place your hand on the lever. Before you move the lever, you have to decide what distance you want to move, then make this movement. Remember the distance moved, and replace your hand on the table. You should try to vary your choices as much as possible.

Following a delay, the lever was placed in the starting-position for that trial, and the subject's hand was replaced on the lever with the instruction to move the lever the same distance as before.

Subjects were informed that returning to the end-point of the movement would not result in accurate performance, as the starting-points for reproduction trials would always be different from those for the original movements. For the location experiment, the instructions were the same as above, except that now the subjects were asked to choose a location, then move to it, and remember it.

For these tasks the experimenter does not determine either the extent of the movement or the location of the stopping-point. The starting-point for each movement, however, was chosen by the experimenter, and these starting-points ranged over most of the scale. This arrangement rendered it difficult for subjects to make exactly the same movement on successive trials, and also discouraged the strategy of increasing successive movements by a constant amount. Displacements were permitted in one direction only: towards the right, when using the right hand, and vice versa for the left. For preselected distance, the subjects had to be allowed enough of the scale to be able to reproduce the movement they had chosen; therefore, all starting-points for the left hand had to commence at a point on the scale to the right of the original one, and to the left for the right hand (see Fig. 3).

Two blocks of trials were constructed, each consisting of 12 different positions. Blocks of starting-positions were presented either in the order ABBA (beginning with the hand ipsilateral to the lesion) or BAAB (beginning with the

SELF-GENERATED ARM MOVEMENTS 615

Br.~

La. ~ _ ~

Small Removals

Ro. Be.

Br. Fe. ~ ---.,

GI..Ji.

a(" ~ r

Jo t:d

Large Removals

Ru. Ma. - ,,

Ra. Mi. Je. -"

Lo. Th.

Fig. 2. Diagrams showing the estimated extent of cortical excision for the patients in the right frontal-lobe groups; small removals are shown on the left, and large removals on the right.

Table 2. Extent of removal for temporal-lobe groups and time of testing for all patient groups

Group

Time of testing Extent of neocortical excision (cm)

Post- Follow- Sylvian fissure Base operative up Mean Range Mean Range

Left temporal --small hippocampal removal 7 Left temporal --large hlppocampal removal 4 Left frontal 3 Right temporal --small bippocampal removal 3 Right temporal --large hippocampal removal 3 Right frontal --small removal 2 Right frontal --large removal 1

6 5.1 4.5-65 5.2 4.~6.5

5 4.9 4.56.0 5.5 4.56.4 2 . . . .

10 50 4.5-55 5.7 4.5~.3

4 5.3 4.5~.5 6.1 4.58.5

4 . . . .

5 . . . .

616 G. LEONARD and B. MILNER

Fig. 3. Modified version of the Manual Lever D. Above: Examiner's view. Below: subjects's position with respect to the apparatus. The arrow indicates the direction of movement for the right hand.

hand contralateral to the lesion). These orders were alternated within groups for successive subjects. Each subject received 48 trials, 24 with each hand.

In order to classify movement extents or end positions, a novel response was operationally defined, as having to differ by four degrees from its closest neighbour. Testing was carried out in a single session, and successive subjects, within each group, were alternated with respect to whether they performed the distance or the location experiment first.

R E S U L T S

T h e ana lys i s o f v a r i a n c e for sub jec t - se lec ted d i s t ance and tha t for sub jec t - se lec ted l o c a t i o n

i nc luded f o u r fac tors ; g r o u p , de lay (15 sec vs 30 sec), in te r fe rence , a n d hand . G r o u p was the

on ly i n d e p e n d e n t f ac to r a n d the d e p e n d e n t va r i ab l e was a b s o l u t e e r ro r .

S E L F - G E N E R A T E D ARM M O V E M E N T S 617

12

A

..Q <

10

9

8

7

6

5

o[

Wtthout Counting With Counting

/ 15 30 30

Delay (s)

1 15

Fig. 4. Subject-selected distance: main effect of delay (without counting) and interference (with counting). Recall accuracy is significantly influenced both by the passage of time and by the

performance of an interpolated task in the recall-interval.

Subject-selected distance Main effect of delay. There was a main effect of delay evident for the absolute-error analysis

I F = 40.39, d.f. = 1, 61, P < 0.0001], performance after 15 sec delay being more accurate than after 30 sec (see left side of Fig. 4).

Main effect of interference. The presence of an interpolated task resulted in less accurate recall than was seen in the absence of interference I F = 15.25, d.f .= 1, 61, P<0.001] ; this effect is illustrated to the right in Fig. 4. No other main effects or interactions reached significance.

Subject-selected location Main effect of delay. For subject-selected location, there was again a main effect of delay

(see left side of Fig. 5); recall accuracy was again seen to be a deteriorating function of time [F-- 65.54, d.f .= 1, 61, P<0.0001] .

Main effect of interference. Figure 5 on the right also illustrates the fact that when subjects were not required to count backwards in the recall interval, accuracy was significantly greater than when an interpolated task was present [ F = 9.76, d.f. = l, 61, P < 0.01].

Diversity of movement-selection Figure 6 shows the mean number of novel distances and novel locations chosen by the

individual groups. Analysis of variance did not reveal any significant differences between the groups, patients with frontal-lobe lesions selecting as many different exemplars as the other groups.

618 G. LEONARD and B. MILNER

12

11

~5

g

0!

Wtthout Counhng Wdh Counting

15 30 15 Delay (s)

30

Fig. 5. Subject-selected location: recall accuracy deteriorates with the passage of time, and is also affected by the presence in the intratrial delay of an interpolated task.

16

,-. 14

12

W Q 10 .-. 0

0 8

~ 8 E z 4 C

~ 2

• Distance

0 LTh LTH LF RTh RTH RF$ RFI. NC

Fig. 6. Mean number of choices made by the various groups on the distance and location tasks.

D I S C U S S I O N

In experiments examining subject-selected movements in normal subjects [9, 25, 35-37, 42-1, the consensus, a m o n g investigators, is that distance and location cues are both coded centrally and retained over an unfilled interval, and that both show significant decay when

SELF-GENERATED ARM MOVEMENTS 619

the interval is taken up with an interpolated task. In addition, there is a superiority of recall for preselected movements over that for examiner-defined movements, and both Roy [36], and Kelso and Wallace [16] have suggested that this advantage derives from the fact that the subject has the opportunity to develop an image, just before the movement is generated, which enhances the encoding process. Schmidt [39] has proposed that the preselection effect arises because the subjects are in fact producing a movement from a store of familiar ones and therefore reproducing this known movement is easier than when the subject has no prior information about the extent or final location of a movement (as is the case for examiner- defined movements). Many other authors have invoked Sperry's corollary discharge notion [41 ] to explain the preselection phenomenon. Teuber [46] describes corollary discharge as a concomitant discharge from motor to sensory cortex that takes place at the same time as the efferent command to move the musculature. The sensory system is thereby tuned or preset for changes in the input that are "the anticipated consequences of the willed movement" [43]. Finally, Lee and Gallagher [I 8] have drawn a parallel between the preselection effect and the generation effect in verbal memory studies [13, 27, 28, 40]. McFarland et al. [28] have suggested that the preselection effect is directly related to the greater amount of effort required of subjects, when they have to produce the stimuli themselves rather than having them supplied by the experimenter. An interesting parallel exists between this idea and the finding of McAndrews and Milner [26] that a memory deficit for temporal order, demonstrated by patients with frontal-lobe lesions, is no longer observed when these subjects are given the opportunity to perform actions with the objects whose order is to be recalled. This finding has been confirmed and extended recently by Butters et al. [5], who demonstrated that naming, visual imagery, experimenter-performed tasks or verbal elaboration were not sufficient to overcome the frontal-lobe deficit in recency recall, the performance of actions by the subjects themselves being the critical factor in overcoming the deficit.

The findings for the self-selected tasks of the present study revealed that patients with large right frontal-lobe lesions did not differ significantly from the normal control subjects in recalling the extent of movements, or their end positions, that they had chosen for themselves. In light of the difficulties experienced by patients with large right frontal-lobe excisions on the examiner-defined distance and location tasks [23, 24], it is particularly interesting that these patients were unimpaired on the present tasks. The results show that, when the need to monitor feedback is reduced, these subjects can accurately recall the extent of their limb movements and the terminal position of these movements. Patients with small right frontal-lobe excisions, or with left frontal-lobe removals, were also normal in their performance of these self-selected tasks. The accurate recall of the frontal-lobe groups was not secondary to an abnormally repetitive choice of exemplars, as these patients were just as varied as the other groups in their selection of novel movements and novel end positions.

Temporal-lobe resection, irrespective of the medial extent of the removal, did not interfere with the ability to generate and recall distance traversed, a result that is consistent with the findings for examiner-defined distance [23]. On the subject selected end-position task, patients with temporal-lobe lesions also performed as well as control subjects. This finding is notable in the case of the subjects with extensive medial removals, and is in contrast to the impaired scores obtained by such patients on the recall of examiner-defined location following 30 sec during which an interference task was performed [24].

The normal performance of the temporal- and frontal-lobe groups on the preselected tasks might be explained in two ways, not necessarily mutually exclusive. The first is that these

620 G. LEONARD and B MILNER

tasks were easier than the examiner-defined ones, and therefore a restricted unilateral lesion may not have been sufficient to produce a deficit; however, it was noted that the interference conditions were associated with greater error on these subject-generated tasks than was delay per se, and therefore it would seem reasonable to suppose that some cognitive effort must have been required to maintain the information in memory [4]. Another possible explanation of the normal scores obtained by all the patient groups on these tasks might be that the cortical areas critical to performance are not encroached upon by a temporal- or frontal-lobe removal. In terms of Sperry's [41] notion of corollary discharge (i.e. that there is a concomitant discharge from motor to sensory cortex, which takes place at the same time as the efferent command to move), it could be that area 4, in conjunction with SI and areas 5 and 7, may be sufficient to effect and subsequently reproduce, centrally generated movements. This would be especially true for those movements that do not place demands on feedback systems, either for their initiation, or for their accuracy. It is interesting to note that, although many of the frontal-lobe excisions intruded into the supplementary as well as into cingulate motor areas (regions hypothesized to play a facilitative role in the selection of appropriate motor responses [32]), nonetheless, in the present studies, subjects with these lesions were just as accurate as normal control subjects in the reproduction of limb movements that they themselves had selected.

The view that parietal areas 5 and 7, in cooperation with motor cortex and basal ganglia, might subserve a movement system such as that described above, has been expressed by Mountcastle et al. [31]. These authors have postulated that the posterior parietal cortex, in addition to receiving information descriptive of the position and movement of the body in space, contains a command centre for the operation of the limbs, hands, and eyes within immediate extrapersonal space [11, 15]. In keeping with this interpretation are the findings that monkeys with parietal-lobe lesions show a neglect of the contralateral limb, a reduction in spontaneous movements, and errors in reaching with the contralateral arm into either half of extrapersonal space [8, 10]. The cerebellum, with its dense projections (via the pontine nuclei) from SI, and areas 5, 7, and 4 might be a key player in the process of activating established motor patterns [1, 17, 19-21, 33, 34, 38].

Acknowledgements--The research was supported by Grant MT 2624 and a career investigatorship from the Medical Research Council of Canada to Brenda Milner. We thank Drs T. Rasmussen, W. Feindel, A. Olivier and J. G. Villemure of the Montreal Neurological Hospital for the opportunity to study their patients, and for providing detailed descriptions of the surgical removals.

REFERENCES 1. Barker, W. W., Yoshii, F., Loewenstein, D. A., Chang, J. Y., Apicella, A., Shlomo, P., Boothe, T. E., Ginsberg,

M. D. and Duara, R. Cerebrocerebellar relationship during behavioral activation: A PET study. J. Cereb. Blood Flow Metab. 11, 48-54, 1991.

2. Benton, A. L. Differential behavioral effects in frontal lobe disease. Neuropsychologia 6, 53~50, 1968. 3. Branch, C., Milner, B. and Rasmussen, T. Intracarotid sodium amytal for the lateralization of cerebral speech

dominance: Observations in 123 patients. J. Neurosurg. 21, 399-405, 1964. 4. Broadbent, D. E. Perception and Communicatton. Pergamon Press, London, 1958. 5. Butters, M. A., Kaszniak, A. W., Glisky, E. L., Eslinger, P. J. and Schacter, D. L. Recency discrimination

deficits in frontal lobe patients. Neuropsychology 8, 343 353, 1994. 6. Crovitz, H. F. and Zener, K. A group test for assessing hand- and eye-dominance. Am. J. Psychol. 75, 271 276,

1962. 7. Corkin, S., Milner, B. and Rasmussen, T. Somatosensory thresholds: Contrasting effects of postcentral gyrus

and posterior parietal-lobe excisions. Archs Neurol. 22, 41-58, 1970. 8. Deuel, R. K. Loss of motor habits after non-sensorimotor cortical lesions in monkeys. J. Neurol. 19, 294, 1969.

SELF-GENERATED ARM MOVEMENTS 621

9. Diewert, G. L. and Roy, E. A. Coding strategy for memory of movement extent information. J. exp. Psychol." Hum. Learn. Mere. 4, 666~75, 1978.

10. Ettlinger, G. and Kalsbeck, J. E. Changes m tactile discrimination and in visual reaching after successive and simultaneous bilateral posterior parietal ablations in the monkey. J. Neurol. Neurosur#. 25, 256-268, 1962.

11. Georgopoulis, A. P., Kalaska, J F. and Caminiti, R. Relations between two-dimensional arm movements and single cell discharge in motor cortex and area 5: Movement direction versus movement endpoint. Exp. Brain Bes. 10, 175-183, 1985.

12. Jason, G. W. Gesture fluency after focal cortical lesions. Neuropsychologia 23, 463-481, 1985. 13. Jacoby, L. L. On interpreting the effects of repetition: Solving a problem versus remembermg a solution. J.

Verb. Learn. Verb. Behav. 17, 649-667, 1978. 14. Jones-Gotman, M. and Milner, B. Design fluency: the invention of nonsense drawings after focal cortical

lesions. Neuropsychologia 15, 653-674, 1977 15. Kalaska, J. K., Caminm, R. and Georgopoulos, A. P. Cortical mechanisms related to direction of

two-dimensional arm movements: Relations in parietal area 5 and comparison with motor cortex. Exp. Brain Res. 51,247-260, 1983.

16. Kelso, J. A. S., and Wallace, S. A. Conscious mechanisms in movement. In Information Processing in Motor Control and Learning, G. E. Stelmach (Editor). Academic Press, New York, 1978.

17. Krupta, D. J., Thompson, J. K. and Thompson, R. F. Localization of a memory trace in the mammalian brain. Science 260, 989-991, 1993.

18. Lee, T. D. and Gallagher, J. D. A parallel between the preselection effect in psychomotor memory and the generation effect in verbal memory. J. exp. Psychol.. Hum. Learn. Mere. 7, 77-78, 1981.

19. Leiner, H. C., Leiner, A. L. and Dow, R. S. Does the cerebellum contribute to mental skills? Behav. Neurosci 100, 443-454, 1986.

20. Lelner, H. C., Leiner, A. L. and Dow, R. S. Cerebro-cerebellar learning loops in apes and humans, hal. J. Neurol. Sct. 8, 425-436, 1987.

21. Leiner, H. C., Leiner, A. L. and Dow, R. S. The human cerebro-cerebellar system: lts computing, cognitive, and language skills. Behav. Brain Res. 44, 113 128, 1991.

22. Leonard, G., Jones, L. and Milner, B. Residual impairment in handgrip strength after unilateral frontal-lobe lesions. Neuropsychologia 26, 555-564, 1988.

23. Leonard, G., and Milner, B. Contribution of the right frontal lobe to the encoding and recall of kinesthetic distance information. Neuropsychologia 29, 47-58, 1991a.

24. Leonard, G., and Milner, B. Recall of the end-position of examiner defined arm movements by patients with frontal- or temporal-lobe lesions. Neuropsychologia 29, 629-640, 1991b.

25. Marteniuk, R. G. Retention characteristics of motor short-term memory cues. J. Motor Behav. 5, 249 259, 1973.

26. McAndrews, M. P. and Milner, B. The frontal cortex and memory for temporal order. Neuropsychologta 29, 849-859, 1991.

27. McElroy, L. A. and Slameeka, N. J. Memorial consequences of generating nonwords: Implications for semantic-memory interpretations of the generation effect. J. Verb. Learn. Verb. Behav. 21, 249-259, 1982.

28 McFarland, C. E., Frey, T. J. and Rhodes, D. D. Retrieval of internally versus externally generated words in episodic memory. J. Verb. Learn. Verb. Behav. 19, 210-225, 1980.

29. Mdner, B. Someeffectsoffrontallobectomyinman. In The Frontal Granular Cortex and Behawor, J. M.Warren and K. Akert (Editors), pp. 313-334. McGraw-Hill, New York, 1964.

30. Milner, B. Psych•••gica•aspects•ff••a•epi•epsyanditsneur•surgica•management.InAdvancesinNeur•l••y• D. P. Purpura, J. K. Penry and R. D. Walter (Editors), Vol. 8, pp. 299 321. Raven Press, New York, 1975.

31. Mountcastle, V. B., Lynch, J. C., Georgopoulos, A., Sakata, H. and Acuna, C. Posterior parietal association cortex of the monkey: Command functions for operations within extrapersonal space. J. Neurophysiol. 38, 871-908, 1975.

32. Paus, T., Petrides, M., Evans, A. C. and Meyer, E. Role of the human anterior cingulate cortex in the control of oculomotor, manual, and speech responses: A positron emission tomography study. J. Neurophystol. 70, 453-469, 1993.

33. Rlspal-Padel, L., Ciclrata, F. and Pons, C. Cerebellar nuclear topography of simple and synergistic movements in the alert baboon (Papio papto). Exp. Brain. Res., 47, 365-380, 1982.

34. Rispal-Padel, L., Ciclrata, F. and Pons C. Neocerebellar synergies. In Neural coding of motor performance, J. Massion, J. Paillard, W. Schultz and M. Weisendanger (Editors), pp. 213-223 Springer, Heidelberg; and Exp. Brain Res., Suppl. 7, 1983.

35. Roy, E. A. Spatial cues in memory for movement. J. Mot. Behav. 9, 151-156, 1977. 36. Roy, E A. Role ofpreselectlon in memory for movement extent. J. exp. Psychol. Hum. Learn. Mere. 4, 397-405.

1978. 37 Roy, E. A. and Kelso, J. A. S. Movement cues in motor memory: Precuing versus postculng. J. Hum. Move.

Stud. 3, 232 239, 1977. 38. Schmahmann, J. D. The cerebellar contribution to higher function. Arch. Neurol. 48, 1178 -1187, 199l