Intact mirror-tracing and - Faculty Virginia

advertisement
Copyright 1997 by the American Psychological Association, Inc.
0894-4105/97/$3.00
Neuropsychology
1997, Vol. II, No.2, 272-281
Intact Mirror-Tracing and Impaired Rotary-Pursuit Skill Learning
in PatientsWith Huntington's Disease:Evidence for Dissociable
Memory Systemsin Skill Learning
John D. E. Gabrieli
Glenn T. Stebbins, laswinder Singh
Rush-Presbyterian-St.Luke's Medical Center
Rush-Presbyterian-St.Luke's Medical Center,
and StanfordUniversity
Daniel B. Willingham
Universityof VIrginia
Christopher
G. Goetz
Rush-Presbyterian-St. Luke's Medical Center
Skill learning in early-stageHuntington'sdisease(HD) patientswas comparedwith that of
nonnal controls on 2 perceptual-motortasks,rotary pursuit and mirror tracing. HD patients
demonstrateda dissociationbetweenimpaired rotary-pursuit and intact mirror-tracing skill
learning. These results suggestthat different fonns of perceptual-motorskill learning are
mediatedby separableneuralcircuits.A striatalmemorysystemmay be essentialfor sequence
or open-loopskill learning but not for skills that involve the closed-looplearning of novel
visual-responsemappings.It is hypothesizedthat working memory deficits in HD resulting
from frontostriataldamagemay accountbroadly for intact and impaired long-tenn learning
andmemoryin HD patients.
Dissociations in patients with brain injuries between
intact and inlpaired fonDSof learning provide evidencefor
the existenceof memorysystems,particularneuralnetworks
that mediate specific mnemonic processes.Patients with
global amnesiahave bilateral lesionsin medial-temporalor
diencephalic brain regions, and they exhibit a relatively
circumscribeddeficit in the learning and rememberingof
new events and facts, a fonD of memory that has been
temled declarative(Cohen & Squire, 1980). The deficit is
evident in explicit testsof recall and recognitionfor recent
episodesor materialsencounteredin thoseepisodes(Oraf &
Schacter, 1985). Amnesic patients, however, have shown
intact memory perfonnanceon two broad classesof tasks
where memory is measuredimplicitly or indirectly by a
changein perfonnanceandwithout explicit referenceto any
prior study or training episode: skill-learning tasks and
repetition-primingtasks.
In skill-learning tasks,memory is measuredas inlproveJohn D. E. Gabrieli, Department of Neurological Sciences,
Rush-Presbyterian-St.Luke's Medical Center,Chicago, illinois,
and Department of Psychology, Stanford University; Glenn T.
Stebbins,JaswinderSingh,and ChristopherG. Goetz,Department
of Neurological Sciences,Rush-Presbyterian-St.Luke's Medical
Center, Chicago, illinois; Daniel B. Willingham, Departmentof
Psychology,University of Virginia. JaswinderSingh is now at the
Departmentof Psychiatryand BehavioralSciences,Northwestern
University.
This researchwas supportedby grantsfrom the Office of Naval
Research(NOOOI4-92-J-184)
and from the McDonnell-PewProgram in Cognitive Neuroscience.
Correspondenceconcerningthis article should be addressedto
John D. E. Gabrieli, Departmentof Psychology,StanfordUniversity, Stanford, California 94305. Electronic mail may be sent via
Internet to gabrieli@psych.stanford.edu.
272
ment with practice in perceptual-motor, perceptual, or
cognitive performance. Amnesic patients have shown intact
skill learning on a number of tasks, including mirror tracing
(Gabrieli, Corkin, Mickel, & Growdon, 1993; Milner, 1962),
rotary pursuit (Cermak, Lewis, Butters, & Goodglass, 1973;
Corkin, 1968; Heindel, Butters, & Salmon, 1988), serial
reaction time (Nissen & Bullemer, 1987; Nissen, WIllingham, & Hartman, 1989), and reading mirror-reversed text
(Cohen & Squire, 1980; Martone, Butters, Payne, Becker, &
Sax, 1984). In repetition-priming tasks, memory is measured
as a change in the speed or accuracy of processing a
stimulus, usually a word or picture, due to prior exposure to
that stimulus. Amnesic patients have shown intact repetition
priming on a number of tasks involving words (e.g., Graf,
Shimamura, & Squire, 1985; Graf, Squire, & Mandler, 1984;
Warrington & Weiskrantz, 1970), representational pictures
(Cave & Squire, 1992), and novel, nonrepresentational
pictures (Gabrieli, Milberg, Keane, & Corkin, 1990; Musen
& Squire, 1992; Schacter, Cooper, Tharan, & Rubens, 1991).
The sparing of skill learning and repetition priming in
amnesia indicates that those forms of memory do not depend
upon the medial-temporal and diencephalic brain structures
that are vital for declarative memory, but it does not indicate
how many different forms of memory are involved in either
skill learning or repetition priming or what neural networks
subserve those forms of memory. Studies of patients with
neurodegenerative diseases,especially Huntington's disease
(HD) and Alzheimer's disease (AD), have provided evidence about the functional neural architecture of skill
learning and repetition priming. HD patients have shown
impaired skill learning on tasks of rotary pursuit (Heindel et
al., 1988; Heindel, Salmon, Shults, Walicke, & Butters,
1989), reading mirror-reversed text (Martone et al., 1984),
and, more variably, serial reaction time (Koopman & Nissen,
273
INTACf AND IMPAIRED SKll..L LEARNING IN HD
1991; Wlllingham & Koroshetz, 1993). HD patients, however, often show intact repetition priming (Heindel et al.,
1989; Salmon, Shimamura, Butters, & Smith, 1988). AD
patients have shown the reverse dissociation, exhibiting
intact skill learning on tasks of rotary pursuit (Eslinger &
Damasio, 1986; Heindel et al., 1988, 1989), mirror tracing
(Gabrieli et al., 1993), and serial reaction time (Koopman,
1991; Koopman & Nissen, 1987) and impaired repetition
priming (e.g., Gabrieli et al., 1994; Heindel et al., 1989;
Keane, Gabrieli, Fennema, Growdon, & Corkin, 1991;
Monti et al., 1996; Salmon et al., 1988; Shimamura, Salmon,
Squire, & Butters, 1987).
The double dissociation between skill learning and repetition priming in HD and AD indicates that at least some
forms of skill learning and repetition priming are mediated
by different memory systems. Further, knowledge of the
prominent pathology in the early stages of the two diseases
(in later stages of both diseases patients are untestable)
provides evidence about the neural networks that mediate
skill learning and repetition priming. Postmortem studies
(Vonsattel et al., 1985), in vivo structural imaging (Starkstein et al., 1989), and in vivo functional imaging (Kuhl et ai,
1982; Young et al., 1986) indicate that the caudate and
putamen, structures that are part of the basal ganglia or
striatum, are especially affected in HD. Conversely, the
striatum is relatively spared in AD (Arnold, Hyman, Flory,
Damasio, & Van Hoesen, 1991; Brun & Englund, 1981).
Thus, it appears that the basal ganglia are part of a memory
system that mediates various forms of skill learning but not
repetition priming.
AD patients have marked hippocampal lesions (Arnold et
al., 1991; Brun & Englund, 1981; Hyman, Van Hoesen,
Damasio, & Barnes, 1984), but it is unlikely that those
lesions account for their repetition-priming deficits because
amnesic patients with hippocampal lesions exhibit preserved
repetition priming. AD patients, unlike most amnesic patients, have marked pathology in frontal, temporal, and
parietal association areas, with relative sparing of primary
sensory, cerebellar, and striatal regions. Therefore, cortical
areas damaged in AD, but not in HD or amnesia, may
constitute memory systems that mediate repetition priming.
Repetition priming, however, can be further decomposed
into perceptual priming, reflecting implicit memory for
stimulus fonn, and conceptual priming, reflecting implicit
memory for stimulus meaning (e.g., Blaxton, 1989; Roediger, Weldon, & Challis, 1989). Both perceptual and conceptual priming are spared in amnesia, but the two forms of
priming dissociate in AD patients who are often intact on
perceptual priming tasks (e.g., Gabrieli et al., 1994; Keane et
al., 1991) but are impaired on conceptual priming tasks (e.g.,
Monti et al., 19%; Salmon et al., 1988). Therefore, perceptual forms of priming may be mediated by modality-specific
perceptual cortices, which are relatively spared in AD, and
conceptual forms of priming may be mediated by association cortices, which are severely compromised in AD. This
interpretation is supported by evidence from lesion and
imaging studies indicating that visual perceptual priming is
mediated by occipital cortex and that conceptual priming is
mpti;"t..J1 hv left frontal and temporal-parietal cortices
(Blaxton et al., 1996; Buckner et al., 1995; Demb et al.,
1995; Fleischman et al., 1995; Gabrieli, Fleischman, Keane,
Reminger, & Morrell, 1995; Gabrieli, Singh, Stebbins, &
Goetz, 1996; Keane, Gabrieli, Mapstone, Johnson, & Corkin, 1995; Raichle et al., 1994; Squire et al., 1992; Swick &
Knight, 1996).
In the present study, we aimed to examine whether two
forms of perceptual-motor skill learning that are spared in
amnesia are further dissociable. Such a dissociation would
indicate that various forms of skill learning, like various
forms of repetition priming, are mediated by separable
memory systems. Therefore, we examined the skill learning
of one group of early-stage HD patients on both rotarypursuit and mirror-tracing tasks. We hypothesized that HD
patients would show impaired learning of the rotary-pursuit
skill as other HD patients have done before (Heindel et al.,
1988, 1989). The performance of HD patients on mirror
tracing has not been examined previously. Cerebellar lesions
impair mirror-tracing skill learning (Sanes, Dimitrov, &
Hallett, 1990), but it is not known whether the striatum is
essential for such skill learning.
One difficulty in assessingmotor skill learning in HD is
the inherent confound between baseline performance ability
and learning in a group of patients with a motor disease. If
patients are tested under conditions identical to those for
normal control participants, then any deficit could reflect
initial performance difficulties, or learning difficulties, or
both. This is not just a nuisance issue in measurement,
because it is plausible that changes in the same neural
network that mediates initial, unskilled performance constitute the basis of later, skilled performance. To minimize the
confounding effects of initial performance and skill learning,
some investigators have equated initial performance between HD and control groups by testing the HD patients
under easier conditions (Heindel et al., 1988, 1989). We
chose the same approach in our study and also explored the
implications of that approach. For rotary pursuit, HD and
controls were tested at speeds of rotation that resulted in
equivalent initial performance. In addition, 3 other HD
patients were tested under conditions that resulted in superior initial performance, so that we could examine what
consequences such manipulations of initial performance
have for rotary-pursuit skill learning. For mirror tracing, HD
and controls were tested with one of three patterns of
descending difficulty (star, diamond, square).
Method
Participants
There were 21 participants:12 HD patients(5 male, 8 female)
and 9 (3 male, 6 female)healthynormal control participants(NC;
seeTable 1). The diagnosisof HD was madeby a board-certified
neurologist based on a positive family history of HD and the
presenceof choreatic movements.Severity of HD was assessed
with the Huntington's DiseaseFunctionalCapacityScale(Shoulson & Fahn, 1979),which stagesdiseaseseverityfrom I (minimal
disability) to 5 (completedisability). The HD patientswere all in
the early stagesof the disease:4 patientswere ratedat Stage2, 6
274
GABRIEU, SlEBBINS, SINGH, Wll..LINGHAM, AND GOE1Z
Table 1
Participant Characteristics
Durationb
2
3
4
5
6
7
8
9
10
11
12
M
SD
38
40
52
31
47
40
41
41
67
37
42
48
43. .7
9..2
14
12
14
12
12
16
12
14
14
12
12
14
13.2
1.3
26
26
30
28
25
29
28
26
26
27
25
26
27.0
1,.6
6
7
2
4
5
7
4
7
6
5
7
3
5.3
1.7
RP, MT
RP, MT
RP, MT
RP, MT
RP, MT
RP, MT
RP
RP
RP
RP. t
RP. t
RP. t
NC
3
2
2
.
3
-
4
-
2
3
2
2.1
1.0
HDFCS
4
4
2
3
3
2
2
3
3
3
3
2
2.8
0.7
42.8
14.6
29.9
13.8
2.1
0.3
Note. MMSE = Mini-Mental StateExamination;Ham-D = Hamilton DepressionRating Scale.
RP = rotary pursuit; MT = mirror tracing; RP+ = rotary pursuit with 10-sbaseline;HDFCS =
Huntington's DiseaseFunctional Disability Scale; HD = Huntington's disease;NC = nonna!
control.
"In years. bDurationofHD in years.
M
SD
patientsat Stage3, and2 patientsat Stage4. Presenceor absenceof
dementiawas assessedusing the Mini-Mental State Examination
(MMSE; Folstein,Folstein,& McHugh, 1975).All patientsscored
above 24, the standardcut-off score for dementia. Presenceor
absenceof depressionwas assessed
using the Hamilton Depression
Rating Scale (Hamilton, 1960). All patients scored below 10,
indicating that none was depressed.Nine HD participants completed the rotary-pursuittask, and 6 of those also performed the
mirror-tracingtask.An additional3 HD participantsperformedthe
rotary-pursuit task at a starting point superior to the NC participants.
The 9 NC participants were hospital volunteers, community
volunteers,and spousesof the HD participantsin good health by
self-reportandnot dementedby MMSE scores(Table 1). The HD
and NC groups did not differ significantly in regard to age or
education.The NC group had significantly higher MMSE scores,
1(19) = 5.13,p < .001. All 9 participantsperformed the rotarypursuit task, and 8 participantsalso performedthe mirror-tracing
task.
Materials
The rotary-pursuitapparatuspresenteda O.75-in.-diameter(1.91
cm) circular target,located3.25in. (8.26 cm) away from the center
of the apparatus,that rotatedclockwise in a circular path. Participantsheld a stylus,anda photoelectricdevicetriggereda clock that
measuredhow long the stylus was held directly over the targeton
each trial. The mirror-tracing apparatusconsistedof an aluminum
plate on which nonconductingtape was placed in one of three
patterns:a six-pointedstar,a four-sideddiamond,or a square.One
patternat a time was mountedon a woodenboard that was hidden
from the participants' view by a near-vertical metal barrier.
Participantshadto reacharoundthe metalbarrier and could seethe
pattern and their handonly in a mirror, mountedon the far side of
the woodenboard.For eachtrial, a counterrecordedthe numberof
times participantswent off the pattern (errors), and a stopwatch
recordedthe time to completetracing of the pattern.
Procedure
For rotary pursuit, participantsheld a stylus in their preferred
hand and were told to try to maintaincontactwith a rotating spot.
The apparatuscould be adjustedto move the target at various
rotationsper minute for each20-s trial. An initial set of five trials
wasusedto determine,individually for eachparticipant,how many
rotations per minute yielded an initial performanceof about 5 s.
That rotation rate was then used for the skill-learning session,
which consistedof three blocks, with each block comprisedof
eight 20-s trials. There was a 20-s intertrial interval. The three
blocks were separatedby intervalsof about 10 min, during which
participantsperformedothertests.
For Inirror tracing, participantsheld the stylus in their preferred
hand and were told to trace around the pattern, from a starting
mark, working as quickly and accuratelyas possible.All participantswereinitially presentedwith the starpattern.If theparticipant
could not initiate a movementon the star,the diamondfigure was
mounted;if the participant could not initiate a movementon the
diamond, the squarewas mounted.All skill-learning trials were
conductedwith onepatternperparticipant.Testingconsistedof two
blocks with five trials per block. Therewere no intervals between
trials, andthe two blockswere separated
by intervalsof 10-15 min
during which participantsperformedothertests.
Results
Rotary Pursuit
The skill-learning measurewas time-on-targetper trial
(maximumof 20 s). The HD and NC groupswere testedat
meanrotationsper minute,respectively,of28.3 s (SD = 9.3)
and 44.4 s (SD = 9.8); this difference was significant,
1(12)= 3.10,p < .01.The two groupswerewell matchedon
initial time-on-targetperformancewith the HD and NC
INTACT AND IMPAIRED SKll..L LEARNING IN HD
groups scoring, respectively, means of 5.6 s and 6.3 s
(p > .70). Time-on-target scores were analyzedin a repeatedmeasuresanalysisof variance (ANOYA) with variablesof group (HD, NC), block (Block 1, Block 2, Block 3),
and trial (eight trials per block); trials were nestedwithin
blocks. Participants showed skill learning by increasing
time-on-target across blocks, F(2, 32) = 54.55, MSE =
3.04,p < .001,andtrials,F(7, 112)= 4.55,MSE= 1.54,
p < .01.The HD group,however,showedlessskill learning
than the NC group as evidencedby a significant effect of
group, F(I, 16) = 16.87,MSE = 63.46,p < .001, and a
significantinteractionbetweengroup andblock, F(2, 32) =
38.31, p
< .001. A separate analysis of the data from the 6
HD and 8 NC participants who also performed mirror
tracing showed a reliable skill-learning deficit in the HD
patients,with a significanteffect of group,F(I, 12) = 12.96,
MSE = 62.45,p < .01, anda significantinteractionbetween
group andblock, F(7, 84) = 3.14,MSE = 3.08,p < .01 (see
Figure 1).
The HD patientsdemonstratedlessskill learningthanthe
NC group, but the question remained as to whether they
275
showed any skill learning at all. A repeatedmeasures
ANaYA on the meantime-on-targetper block for the lID
patients did not reveal a reliable change across blocks
(p > .26). The lID patients who performed the task at
rotation ratesthat yielded an initial time-on-targetof about
10 s (superiorto that of the NC and other lID group) also
showedno rotary-pursuitskill learning(seeFigure2).
Mirror Tracing
In the HD group, 4 participants traced the star, 1
participant traced the diamond, and 1 participant traced the
square. In the NC group, 7 participants traced the star, and 1
participant traced the square. The skill-learning measures
were completion time (number of seconds taken to trace a
pattern) and errors (number of contacts with the metal plate
signifying departures from the pattern; see Figure 3). Mean
performance on Trial 1 was not matched as closely as in the
rotary-pursuit task. On Trial 1, the HD and NC groups did
not differ reliably for completion time (lID: M = 93.8,
SD = 69.6;NC: M = 60.6,SD = 32.1),t(12) = 1.2,p =
Trial
Figure 1. Rotary-pursuitresultsfor patientswith Huntington'sdisease(HD) and normal control
(NC) participants.Top: Mean time on targetfor 9 HD and9 NC participantsin threeblocks of eight
trials. Bottom: Mean time on target for 6 HD and 8 NC participants who also performed the
mirror-tracin2task.Bracketsarestandarderrorsof the means.
~
GABRIEU, STEBBINS, SINGH, WllLINGHAM,
276
12.
AND GOETZ
--,
I
'II
~
~
~
"OJ 10
C)
..
10
I-
E>
c
G---~e-
0
GI
E
8
~
i=
c
10
GI
/
7
--E1
~
.L
Figure 2. Rotary-pursuitresultsfor 9 normal control (NC) participants;9 patientswith Huntington's disease(HD) whoseinitial performancewas equatedwith that of NC participantsby slowing
the speedof rotation,and3 patientswith Huntington'sdisease(HD+) whoseinitial performancewas
madesuperiorto that ofNC participantsby further slowing the speedof rotation.Meantime on target
per block.
.25, but the HD patients made more errors (lID: M = 29.8,
SD = 19.5; NC: M = 12.8, SD = 8.4),1(12) = 1.2,p < .05.
Completion-time scores and error scores (see Figure 3) were
analyzed in separate repeated measures ANOVAs with
Trial
Trial
Figure 3. Mirror-tracing resultsfor 6 patientswith Huntington's
disease(HD) and 8 normal control (NC) participants.Top: Mean
numberof errorsfor two blocksof five trials. Bottom:Meantime to
tracepatternsfor the sameparticipantson the sametrials. Brackets
are standarderrorsof the means.
variables of group (RD, NC), block (Block 1, Block 2), and
trial (five trials per block); trials were nested within blocks.
Participants showed skill learning by tracing the patterns
progressively faster acrossblocks, F(I, 12) = 15.44, MSE =
1125.31, P < .01, and trials, F(4, 48) = 15.70, MSE =
129.12,p < .001, and by making fewer errors across blocks,
F(I, 12) = 8.86, MSE = 107.80, p < .05, and trials, F(4,
48) = 8.85, MSE = 22.30,p < .001. For both skill-learning
measures, there were Block X Trial interactions, indicating
that within-block improvement across trials decreasedacross
blocks: for completion time, F(4, 48) = 4.57, p < .01; for
errors, F(4, 48) = 3.38,p < .05. There was no main effect of
group on either skill-learning measure and, most important,
no interaction between group and either block or trial,
indicating that the HD and NC groups improved similarly on
mirror tracing.
Discussion
This studyexaminedwhetherHD patientswould show a
skill-learning deficit on mirror tracing, as they have on
skill-learning tasks of rotary pursuit, serial reaction time,
and mirror reading.The HD patients in the presentstudy
were impaired relative to control participants on rotarypursuit skill learning. In fact, HD patients showed no
reliable rotary-pursuit skill learning whether their initial
level of performancewas equal to or superior to that of
control participants.From the first to the last trial, HD and
NC groups improved their rotary-pursuit performanceby
16% and 48%, respectively.The HD impairment is consistent with prior findings (Heindelet al., 1988, 1989).In stark
contrast,the sameHD patientsshowedmirror-tracing skill
learningthat was statisticallyindistinguishablefrom that of
control participants.HD mirror-tracing skill learning was
robust. From the first to the last trial, HD and NC groups
improved the accuraciesof their performancesby 70% and
48%, respectively,and increasedthe speedof their perfor-
INTACT AND IMPAIRED SKILL LEARNING IN HD
mances by 56% and 67%, respectively. Thus, there was a
clear-cut dissociation in a group of HD patients between
impaired rotary-pursuit skill learning and intact mirrortracing skill learning.
The preservation of mirror-tracing skill learning in HD
must be considered with some caution for several reasons.
First, initial HD performance was not equated fully with that
of the NC group. On the first trial, HD patients were not
significantly slower than NC participants, but they were
significantly less accurate. The finding, however, that HD
patients had intact mirror-tracing skill learning on the speed
measure where there was no reliable initial difference favors
the interpretation that the HD group showed normal learning. Second, the interpretation of preserved mirror-tracing
skill learning rests on the absenceof any statistically reliable
interaction between groups and learning measures, and this
absence occurred with small groups of participants. The
study, however, had sufficient statistical power to detect the
rotary-pursuit deficit. Further, inspection of the HD learning
curves on the mirror-tracing task shows that their skill
learning closely paralleled that of the NC group. Thus, there
is no apparent evidence that HD patients had any deficit that
would become significant in a larger sample. It will be
assumed, therefore, for purposes of the remaining discussion, that HD patients had intact mirror-tracing skill learning
that was complicated by their motor and cognitive disease.
There are no published studies of mirror tracing in HD,
but there is a study of prism adaptation in HD (paulsen,
Butters, Salmon, Heindel, & Swenson, 1993). In that study,
participants pointed toward a target while wearing distorting
prisms that shifted objects 20° laterally. HD patients, relative
to control and AD participants, failed in adapting to the
prisms after visuomotor feedback. The prism and mirrortracing tasks share some apparent features: Both tasks
involve visually guided motor learning in an environment
with transposed visuomotor relations. Paulsen et al., however, found that there was a correlation between dementia
severity and impaired prism adaptation in the HD patients
(but not in the AD patients): Less demented HD patients
were unimpaired. The dementia severity of HD patients in
the present study resembles that of the less demented HD
group in the Paulsen study. The two studies are actually in
accord, therefore, in finding that mildly demented HD
patients retain the ability to acquire motor skills in altered
visuomotor environments.
Other studies with HD patients provide further evidence
that dementia severity or other patient variables can have
major consequenceson HD patients' skill-learning abilities.
Heindel et al. found correlations between dementia severity
and both rotary pursuit learning (1988, 1989) and weight
biasing (Heindel, Salmon, & Butters, 1991). More demented
HD patients have shown impaired cognitive skill learning on
the Tower-of-Hanoi and Tower-of- Toronto puzzles, but HD
patients with less dementia have shown intact cognitive skill
learning on those tasks (although this latter finding was more
variable for the Tower-of- Toronto study; Butters, Wolfe,
Martone, Granholm, & Cermak, 1985; Saint-Cyr, Taylor, &
Lang, 1988). Dementia severity appears to be an important
variable in determinin~ HD patients' skill-learning perfor-
277
mance, but that variability does not apply to the dissociation
in the present study that occurred in a single group of
patients.
There are several ways to consider the dissociation
between HD performance on two skill-learning tasks. One
constraint is that it is single, not a double, dissociation. Thus,
it could be that learning on both tasks involves the same
memory system, but that rotary pursuit has greater demands
upon that system than does mirror tracing. It was, however,
more difficult to equate HD patients on the mirror-tracing
than the rotary-pursuit task, and more HD patients could
perform the rotary-pursuit than the mirror-tracing task.
These findings do not favor the notion that mirror tracing is
less demanding of a common memory system than is rotary
pursuit. Alternatively, skill learning on the two tasks may be
mediated by memory systems that differ only by a single
component that is essential for rotary pursuit but not mirror
tracing. In either case, no double dissociation would be
expected.
A third possibility is that the two skill-learning tasks
depend upon separable brain regions. The integrity of the
striatum appears essential for rotary-pursuit skill learning,
because such learning is consistently impaired in HD
patients and also in patients with Gilles de la Tourette's
syndrome who have abnormal striatal anatomy (Stebbins,
Singh, Weiner, Goetz, & Gabrieli, 1995). Conversely, the
cerebellum may be essential for mirror-tracing skill learning. Patients with lesions in some, but not all, locations in the
cerebellum have been reported to show impaired mirrortracing skill learning (Sanes et al., 1990). Cerebellar patients
have also been reported to show impaired motor adaptation
in other altered visual environments (Gauthier, Hofferer,
Hoyt, & Stark, 1979; Weiner, Hallett, & Funkenstein, 1983)
and impaired visually guided tracking (Beppu, Suda, &
Tanaka, 1984; Beppu, Nagaoka, & Tanaka, 1987). Therefore, it is possible that a double dissociation between the two
kinds of learning could occur in patients with striatal or
cerebellar lesions. The performance, however, of cerebellar
patients on rotary-pursuit skill learning remains to be
determined.
Rotary pursuit and mirror tracing differ in many ways, and
it is unknown as to what critical difference between the tasks
renders skill learning impaired or intact in HD. Two studies
have reported single dissociations in HD on skill-learning
tasks that were more matched in perceptual and motor
demands. In a serial reaction time task, HD participants were
impaired in learning a sequence-specific skill for pressing
four keys in response to targets appearing in four locations in
a recurring sequence (Willingham & Koroshetz, 1993). HD
patients had an impairment in using the recurring sequence
to predict where the next target would appear.The same HD
patients showed intact skill learning when the task was to
press a key one position to the right of the target. In this task,
there was no sequence to be learned. Rather, participants
learned a new mapping between visual stimuli (targets in a
particular location) and motor responses (pressing the key
one position to the right). In a task where participants
tracked a cursor with a joystick, HD patients showed a
normal pattern of skill learning when the cursor moved
278
GABRIEU. STEBBINS. SINGH. Wll..LINGHAM. AND GOETZ
randomly but impaired learning when the cursor moved in a
repeating pattern (Wlllingham, Koroshetz, & Peterson, 1996).
Willingham proposed that the two within-task dissociations could be explained by a distinction between the
learning of repetitive motor sequences,which depends upon
the striatum, and the learning of new mappings between
visual cues and motor responses, which does not depend
upon the striatum. Thus, HD patients were impaired for
learning repeated sequences of movements in both studies
but were intact for learning new, constant mappings between
visual cues and motor responses.The same distinction could
apply to the present study. Skill learning on rotary pursuit
may involve the learning of a pattern-specific series of
repetitive movements. Indeed, the rotary-pursuit task in the
present study and the tracking of a repeating sequence
(Willingham et al., 1996) are quite similar in this regard.
Rotary-pursuit skill learning does not require any new
mapping between visual cues and motor responses. Conversely, skill learning on mirror tracing appears straightforwardly to reflect the learning of a new mapping between
reversed visual inputs and motor outputs. Further, mirror
tracing may not involve sequence-specific learning because
both control and AD groups demonstrated nearly perfect
transfer of training across two different patterns that involved different sequences of movements (Gabrieli et al.,
1993). Thus, Willingham's distinction between two kinds of
skill learning is consistent with the HD dissociation between
impaired rotary-pursuit learning and intact mirror-tracing
learning.
A second distinction that may apply to the HD dissociation is that between closed-loop and open-loop skillieaming. Skill learning for mirror tracing may be classified as
closed-loop skillieaming because participants are provided
continuous external, visual feedback about errors in movement. Participants can pause at any moment to use current
perceptual feedback about present position and direction of
movement to correct any error in the motor programming
controlling their movements. Knowledge of results is continuous and immediate. In rotary pursuit, however, the target
moves too rapidly and continuously for participants to use
current visual feedback as a cue for motor programming or a
source for error correction. Rather, skill learning for rotary
pursuit may be classified as open-loop learning where good
performance requires participants to program movements on
the basis of a prediction about where the target will be by the
time a motor command is executed. Thus, learning a skill for
rotary pursuit, like learning a skill for serial reaction time
responses to a recurrent sequence, involves responding to
future target locations that are sequentially predicted upon
the basis of currently perceived target locations. Further,
knowledge of results is delayed in rotary pursuit. Participants receive visual feedback that their hand has left the
target after a faulty prediction has been followed. Thus,
neither error correction nor successful motor programming
can be driven by present perceptual information. Rather,
participants must construct a model that uses present perceptual feedback as a predictor for motor programming and
update that model after an error has occurred.
Th~ idea that the striatum has a specific role in open-loop
learning has both supportive and contradictory evidence.
Studies with Parkinson's disease (PD) patients have indicated that the striatum may be critical for open-loop motor
movements that are guided by prediction rather than by
response to current sensory information (Cooke, Brown, &
Brooks, 1978; Flowers, 1978; Stem, Mayeux, Rosen, &
llson, 1983; but see, Hockerman & Aharon-Peretz, 1994).
When performing tracking tasks, PD patients were abnormally dependent upon visual feedback and did not show the
anticipatory motor planning that normal participants did.
Another study, however, reported a normal advantage for PD
patients when they tracked a target moving in a repeating
versus a random pattern (Bloxham, Mindel, & Frith, 1984).
PD patients without dementia, however, exhibit intact
rotary-pursuit learning (Heindel et al., 1989), so it appears
that motor and skill-learning deficits vary considerably
among PD, as well as RD, patients. The sequencingmapping and open-closed loop distinctions need not be
contradictory. For example, animal studies indicate that
striatal cells code open-loop movements only during the
performance of sequencesof movements (Kimura, Aosaki,
Hu, Ishida, & Watanabe, 1992). Thus, under many circumstances, the learning of sequencesmay involve open-loop
learning, whereas the learning of mappings may involve
closed-loop learning.
Finally, it is of interest to consider whether there are
similarities between the perceptual-motor skill-learning deficits seen in RD and more nonmotoric learning and memory
deficits also seen in early RD. RD patients are impaired
often on Tower cognitive skill-learning tasks (Butters et al.,
1985; Saint-Cyr et al., 1988). These cognitive tasks are
open-loop in nature because they require the selection of
movements in response to a mental model of a problem
solution; there is no perceptual information that provides
definitive feedback to guide performance (except for the
final step of the problem). The patients in the present study
also exhibited a characteristic RD pattern of verbal memory
deficits comprised of relatively intact recognition performance and impaired performance on tests of free recall,
self-ordering pointing, and recency judgments (Singh, Gabrieli, Stebbins, & Goetz, 1994). Gabrieli et al. (1996)
proposed that this pattern of intact and impaired verbal
memory performance reflects the severe impairment of
working memory in RD, PD, and Tourette's syndrome.
Verbal memory tasks are intact or impaired to the extent they
make demands upon the reduced working memory capacities exhibited by RD, PD, and Tourette's patients (Gabrieli et
al., 1996; Singh et al., 1994; Stebbins et al., 1995).
Open-loop skill learning that involves the planning of
sequences of movements may also have much higher
demands on working memory than closed-loop skillieaming that uses visual cues to guide. Indeed, Goldman-Rakic
(1987) noted that a fundamental feature of working memory
is that it guides behavior on the basis of internal, symbolic
models as opposed to external, perceptual feedback.
Thus, it is plausible that damage to frontostriatal working
memory systems in RD, PD, and Tourette's syndrome
diminishes long-term learning and memory on tasks that
require internally guided, open-loop thought, including
INTACT AND IMPAIRED SKILL LEARNING IN HD
certain perceptual-motor skills (rotary pursuit, serial response time with sequences), cognitive skills (Tower problems), and strategic verbal memory (recall, recency judgments). Other learning and memory tasks that are less
dependent on working memory capacity may be relatively or
fully intact in HD, PD, and Tourette's, including perceptualmotor skills (mirror tracing) and nonstrategic verbal memory
tasks (recognition). It is likely that the diverse forms of
long-term learning and memory impaired in striatal patients
do not depend on a single frontostriatal system but rather on
multiple, anatomically distinct frontostriatal loops (Alexander, DeLong, & Strick, 1986). The loops mediate working
memory processesin different domains (WIlson, 0' Scalaidhe,
& Goldman-Rakic, 1993), but they may mediate the same
kinds of working memory contributions in each of those
domains. This hypothesis, although speculative, would provide an integrated account for the various forms of learning
and memory that do and do not depend upon the integrity of
the basal ganglia.
References
Alexander, G. E., DeLong, M. R., & Strick, P. L. (1986). Parallel
organization of functionally segregated circuits linking basal
ganglia and cortex. Annual Review of Neuroscience, 9, 357-381.
Arnold, S. E., Hyman, B. T., Flory, J., Damasio, A. R., & Van
Hoesen, G. W. (1991). The topographical and neuroanatomical
distribution of neurofibrillary tangles and neuritic plaques in the
cerebral cortex of patients with Alzheimer's disease. Cerebral
Cortex, 1, 103-116.
Beppu, H., Nagaoka, M., & Tanaka, R. (1987). Analysis of
cerebellar motor disorders by visually-guided elbow tracking
movement. 2. Contribution of the visual cues on slow ramp
pursuit. Brain, 110, 1-18.
Beppu, H., Suda, M., & Tanaka, R. (1984). Analysis of cerebellar
motor disorders by visually-guided elbow tracking movement.
Brain, 107, 787-809.
B1axton, T. A. (1989). Investigating dissociations among memory
measures: Support for a transfer-appropriate processing framework. Journal of Experimental Psychology: Learning, Memory,
and Cognition, 15, 657-668.
Blaxton, T. A., Bookheimer, S. Y., Zeffiro, T. A., Fig10zzi, C. M.,
William, D. G., & Theodore, W. H. (1996). Functional mapping
of human memory using PET: Comparisons of conceptual and
perceptual tasks. Canadian Journal of Experimental Psychology,
50, 42-56.
Bloxham, C. A., Mindel, T. A., & Frith, C. D. (1984). Initiation and
execution of predictable and unpredictable movements in Parkinson's disease.Brain, 107, 371-384.
Brun, A., & Englund, E. (1981). Regional pattern of degeneration
in Alzheimer's disease: Neuronal loss and histopathological
grading. Histopathology, 5, 549-564.
Buckner, R. L., Petersen, S. E., Ojemann, J. G., Miezin, F. M.,
Squire, L. R., & Raichle, M. E. (1995). Functional anatomical
studies of explicit and implicit memory retrieval tasks. Journal
of Neuroscience, 12, 12-29.
Butters, N., Wolfe, J., Martone, M., Granholm, E., & Cermak, L. S.
(1985). Memory disorders associated with Huntington's disease:
Verbal recall, verbal recognition, and procedural memory. Neuropsychologia, 6, 729-744.
Cave, C. B., & Squire, L. R. (1992). Intact and long-lasting
repetition priming in amnesia. Journal of Experimental Psychology: Learning, Memory, and Cognition, 18,509-520.
279
Cennak, L. S., Lewis, R., Butters, N., & Goodglass,H. (1973).
Role of verbal mediation in perfonnanceof motor tasks by
Korsakoffpatients.Perceptionand Motor Skills, 37, 259-263.
Cohen, N. J., & Squire, L. R. (1980). Preservedlearning and
retention of pattern-analyzingskill in amnesia:Dissociationof
knowing how andknowing that.Science,210, 207-210.
Cooke, J. D., Brown, J. D., & Brooks, V. B. (1978). Increased
dependenceon visual infonnation for movement control in
patientswith Parkinson'sdisease.CanadianJournal of Neurological Sciences,5, 413-415.
Corkin, S. (1968).Acquisition of motor skill after bilateral medial
temporal-lobeexcision.Neuropsychologia,6, 255-265.
Demb,J. B., Desmond,J. E., Wagner,A. D., Vaidya,C. J., Glover,
G. H., & Gabrieli, J. D. E. (1995). Semanticencoding and
retrieval in the left inferior prefrontalcortex:A functional MRI
study of task difficulty and process specificity. Journal of
Neuroscience,15, 5870-5878.
Eslinger,P.J., & Damasio,A. R. (1986).Preservedmotor learning
in Alzheimer's disease:Implicationsfor anatomyand behavior.
JournalofNeuroscience,6,3006-3009.
Fleischman,D. A., Gabrieli, J. D. E., Reminger,S., Rinaldi, J.,
Morrell, F., & Wilson, R. (1995). Conceptual priming in
perceptualidentification for patientswith Alzheimer's disease
anda patientwith a right occipital lobectomy.Neuropsychology,
9, 187-197.
Flowers,K. (1978). Lack of prediction in the motor behaviourof
Parkinsonism.Brain, 101, 35-52.
Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975).
"Mini-mental state."A practical methodfor gradingthe cognitive state of patients for the clinician. Journal of Psychiatric
Research,12, 189-198.
Gabrieli, J. D. E., Corkin, S., Mickel, S. F., & Growdon, J. H.
(1993). Intact acquisition and long-tenD retention of mirrortracing skill in Alzheimer's diseaseand in global amnesia.
BehavioralNeuroscience,107, 899-910.
Gabrieli, J. D. E., Fleischman,D. A., Keane, M. M., Reminger,
S. L., & Morrell, F. (1995). Double dissociation between
memorysystemsunderlyingexplicit andimplicit memoryin the
humanbrain. PsychologicalScience,6, 76-82.
Gabrieli,J. D. E., Keane,M. M., Stanger,B. Z., Kjelgaard,M. M.,
Corkin, S., & Growdon, J. H. (1994). Dissociations among
structural-perceptual,lexical-semantic,and even-fact memory
systemsin amnesic,Alzheimer's, and nonnal subjects.Cortex,
30,75-103.
Gabrieli,J. D. E., Milberg, W., Keane,M. M., & Corkin, S. (1990).
Intact priming of patternsdespiteimpaired memory.Neuropsychologia,28, 417-427.
Gabrieli, J. D. E., Singh,J., Stebbins,G. T., & Goetz,C. G. (1996).
Reducedworking memory span in Parkinson's disease:Evidence for the role of frontostriatal system in working and
strategicmemory.Neuropsychology,
10, 322-332.
Gauthier,G. M., Hofferer, J. M., Hoyt, W. F., & Stark,L. (1979).
Visual-motoradaptation:Quantitativedemonstrationin patients
with posterior fossa involvement. Archives of Neurology,36,
155-160.
Goldman-Rakic,P.S. (1987).Circuitry of primateprefrontalcortex
and regulation of behavior by representationalmemory.In F.
Plum (Ed.), Handbookof physiology-the nervoussystemV (pp.
373-417).New York: Oxford University Press.
Graf, P.,& Schacter,D. L. (1985).Implicit andexplicit memoryfor
new associationsin nonnal and amnesic subjects.Journal of
ExperimentalPsychology:Learning, Memory, and Cognition,
11,501-518.
Graf, P.,Shimamura,A. P., & Squire,L. R. (1985).Priming across
modalities and priming acrosscategory levels: Extending the
280
GABRIEU, STEBBINS, SINGH, WILLINGHAM, AND GOE1L
domainof preservedfunction in amnesia.Journal of Experimental Psychology:Learning,Memory,and Cognition,11,386-396.
Oraf, P., Squire,L. R., & Mandler,G. (1984).The informationthat
amnesicpatientsdo not forget.Journal of ExperimentalPsychology: Learning,Memory,and Cognition,10, 164-178.
Hamilton, M. (1960). A rating scale for depression.Journal of
Neurology,Neurosurgery,and Psychiatry,23, 56-62.
Heindel, W. C., Butters, N., & Salmon, D. P. (1988). Impaired
learning of a motor skill in patientswith Huntington'sdisease.
BehavioralNeuroscience,102, 141-147.
Heindel,W. C., Salmon,D. P.,& Butters,N. (1991).The biasingof
weight judgments in Alzheimer's and Huntington's disease:A
priming or programmingphenomenon?Journal of Clinical and
ExperimentalNeuropsychology,13, 189-203.
Heindel, W. C., Salmon,D. P., Shults, C. W., Walicke, P. A., &
Butters, N. (1989). Neuropsychologicalevidencefor multiple
implicit memorysystems:A comparisonof Alzheimer's,Huntington's, andParkinson'sdiseasepatients.TheJournal of Neuroscience,9, 582-587.
Hockerman, S., & Aharon-Peretz,J. (1994). Two-dimensional
tracingandtrackingin patientswith Parkinson'sdisease.Neurology,44,III-116.
Hyman,B. T., Van Hoesen,G. W., Damasio,A. R., & Barnes,C. L.
(1984).Alzheimer's disease:Cell-specificpathologyisolatesthe
hippocampalformation.Science,225, 1168-1170.
Keane,M. M., Gabrieli, J. D. E., Fennema,A. C., Orowdon,J. H.,
& Corkin, S. (1991). Evidence for a dissociation between
perceptual and conceptual priming in Alzheimer's disease.
BehavioralNeuroscience,105,326-342.
Keane,M. M., Gabrieli, J. D. E., Mapstone,H., Johnson,K. A., &
Corkin, S. (1995). Double dissociationof memory capacities
after bilateral occipital-lobe or medial temporal-lobelesions.
Brain, 118,1129-1148.
Kimura, M., Aosaki, T., Hu, Y., Ishida,A., & Watanabe,K. (1992).
Activity of primateputamenneuronsis selectiveto the modeof
voluntary movement:Visually guided,self initiated, or memoryguided.ExperimentalBrain Research,89, 473-477.
Knopman,D. (1991). Long-term retention of implicitly acquired
learningin patientswith Alzheimer's disease.Journal of Clinical
and ExperimentalNeuropsychology,
13(6),880-894.
Knopman, D. S., & Nissen, M. J. (1987). Implicit learning in
patients with probable Alzheimer's disease.Neurology, 37,
784-788.
Knopman, D., & Nissen, M. J. (1991). Procedural learning is
impaired in Huntington's disease:Evidence from the serial
reactiontime task.Neuropsychologia,29, 245-254.
Kuhl, D. E., Phelps,M. E., Markham, C. H., Metter, E. J., Riege,
W. H., & Winter, J. (1982). Cerebralmetabolismand atrophyin
Huntington'sdiseasedeterminedby FDG and computedtomographicscan.Annalsof Neurology,12, 425-434.
Martone,M., Butters,N., Payne,M., Becker,J. T., & Sax, D. S.
(1984).Dissociationsbetweenskill learningandverbalrecognition in amnesiaand dementia.Archives of Neurology, 41,
965-970.
Milner, B. (1962).Les troublesde la memoire accompagnantdes
lesions hippocampiquesbilaterales [Memory problems with
bilateral hippocampallesions].In Psychologiede l'hippocampe.
Paris:CentreNationalde la RechercheScientifique.
Monti, L. A., Gabrieli, J. D. E., Reminger,S. L., Rinaldi, J. A.,
Wilson, R. S., & Fleischman,D.A. (1996).Differential effectsof
aging and Alzheimer's diseaseupon conceptualimplicit and
explicit memory.Neuropsychology,
10, 101-112.
Musen, G., & Squire,L. R. (1992).Nonverbalpriming in amnesia.
Memory and Cognition,20, 441-448.
Nissen, M. J., & Bul1emer, P. (1987). Attentional requirements of
learning: Evidence from perfonnance measures. Cognitive Psychology, 19, 1-32.
Nissen, M. J., Willingham, D., & Hartman, M. (1989). Explicit and
implicit remembering: When is learning preserved in amnesia?
Neuropsychologia, 27,341-352.
Paulsen, J. S., Butters, N., Salmon, D. P., Heindel, W. C., &
Swenson, M. R. (1993). Prism adaptation in Alzheimer's and
Huntington's disease. Neuropsychology, 7,73-81.
Raichle, M. E., Fiez, J. A., Videen, T. 0., MacLeod, A. K., Pardo,
J. V., Fox, P. E., & Petersen, S. E. (1994). Practice-related
changes in human brain functional anatomy during nonmotor
learning. Cerebral Cortex, 4, 8-26.
Roediger, H. L., Weldon, M. S., & Challis, B. H. (1989).
Explaining dissociations between implicit and explicit measures
of retention: A processing account. In H. L. ROediger & F. I. M.
Craik (Eds.), Varieties of memory and consciousness: Essays in
honour of Endel Tulving (pp. 3-14). Hillsdale, NJ: Erlbaum.
Saint-Cyr, J. A., Taylor, A. E., & Lang, A. E. (1988). Procedural
learning and neostriatal function in man. Brain, 111,941-959.
Salmon, D. P., Shimamura, A. P., Butters, N., & Smith, S. (1988).
Lexical and semantic deficits in patients with Alzheimer's
disease. Journal of Clinical and Experimental Neuropsychology,
10, 477-494.
Sanes, J. N., Dimitrov, B., & Hallett, M. (1990). Motor learning in
patients with cerebellar dysfunction. Brain, 113, 103-120.
Schacter, D. L., Cooper, L. A., Tharan, M., & Rubens, A. B. (1991).
Preserved priming of novel objects in patients with memory
disorders. Journal of Cognitive Neuroscience, 3, 117-130.
Shimamura, A. P., Salmon, D. P., Squire, L. R., & Butters, N.
(1987). Memory dysfunction and word priming in dementia and
amnesia. Behavioral Neuroscience, 101, 347-351.
Shoulson, I., & Fahn, S. (1979). Huntington's disease: Clinical care
and evaluation. Neurology, 29, 1-3.
Singh, J., Gabrieli, J. D. E., Stebbins, G. T., & Goetz, C. G. (1994).
Impairment of working memory in patients with Huntington s
disease. Manuscript submitted for publication.
Squire, L. R., Ojemann, J. G., Miezin, F. M., Petersen, S. E.,
Videen, T. 0., & Raichle, M. E. (1992). Activation of the
hippocampus in normal humans: A functional anatomical study
of memory. Proceedings from the National Academy of Sciences,
USA, 89, 1837-1841.
Starkstein, S. E., Fo1stein, S. E., Brandt, J., Pearlson, G. D.,
McDonnell, A., & Fo1stein, M. (1989). Brain atrophy in Huntington's disease. Neuroradiology, 31, 156-159.
Stebbins, G. T., Singh, J., Weiner, J., GOetz, C. G., & Gabrieli, J. D.
E. (1995). Selective impairments of memory functioning in
unmedicated adults with Gilles de la Tourette's syndrome.
Neuropsychology, 9, 329-337.
Stem, Y., Mayeux, R., Rosen, J., & llson, J. (1983). Perceptual
motor dysfunction in Parkinson's disease: A deficit in sequential
and predictive voluntary movement. Journal of Neurology,
Neurosurgery, and Psychiatry, 46, 145-151.
Swick, D., & Knight, R. T. (1996). Is prefrontal cortex involved in
cued recall? A neuropsychological test of PET findings. Neuropsychologia, 34, 1019-1028.
Vonsattel, J. P., Myers, R. H., Stevens, T. J., Ferrante, M. S., Bird,
E. D., & Richardson, E. P. (1985). Neuropathological classification of Huntington's disease. Journal of Neuropathology and
Experimental Neurology, 44, 559-577.
Warrington, E. K., & Weiskrantz, L. (1970). The amnesic syndrome: Consolidation or retrieval? Nature, 228, 628-630.
Weiner, M. J., Hallett, M., & Funkenstein, H. H. (1983). Adaptation
to lateral displacement of vision in patients with lesions of the
central nervous system. Neurology, 33, 766-772.
INTACf AND IMPAIRED SKll..L LEARNING IN HD
Willingham, D. B., & Koroshetz, W. J. (1993). Evidence for
dissociable motor skills in Huntington's diseasepatients. Psychobiology, 21, 173-182.
Willingham, D. B., Koroshetz, W. J., & Peterson, E. W. (1996).
Motor skills have diverse neural bases: Spared and impaired skill
acquisition in Huntington's disease. Neuropsychology, 10, 315321.
Wilson, F. A. W., O'Scalaidhe, S. P., & Goldman-Rakic, P. S.
(1993). Dissociation of object and spatial processing domains in
primate prefrontal cortex. Science, 260, 1955-1957.
281
Young,A. B., Penney,J. B., Startosa-Rubenstein,
S., Markel, D. S.,
Berent,S., Giordani, B., Ehrenkaufer,R., Jewett,D., & Hichwa,
R. (1986). PET scan investigationsof Huntington's disease:
Cerebral metabolic correlates of neurological features and
functionaldecline.Annalsof Neurology,20, 296-303.
ReceivedSeptember13, 1994
RevisionreceivedOctober22,1996
AcceptedOctober23,1996 .
Download