Exp Brain Res (2003) 153: 171–179
DOI 10.1007/s00221-003-1590-6
REVIEW
Yves Rossetti . Laure Pisella . Alain Vighetto
Optic ataxia revisited:
Visually guided action versus immediate visuomotor control
Published online: 12 September 2003
# Springer-Verlag 2003
Abstract Optic ataxia and visual agnosia have been
proposed to constitute a double dissociation which
provides the main argument for the assimilation of the
anatomical distinction between a dorsal and a ventral
visual stream to the functional distinction between
perception and action. In the present review, we argue
that insufficient evidence has been collected to argue for
this double dissociation. Several criteria are reviewed: (1)
exploration of the visuomotor behavior in central versus
peripheral vision has not been matched for the two types
of patients; (2) the temporal constraints of visual processes
that are impaired in the two neurological conditions appear
to play a crucial role in the apparent dissociation; (3) the
necessary reductionism of experimental conditions used to
study action has led to an overconsideration of optic ataxia
as a global deficit for action. Altogether optic ataxia
appears to result from a specific impairment of immediate
visuomotor control rather than of visually guided action as
a whole. These results are discussed in the light of recent
research on optic ataxia and on motor control, and
directions for future research are proposed.
Keywords Optic ataxia . Visual agnosia . Peripheral
vision . Central vision . Visual stream . Visuomotor
transformation
History and definition
Balint (1909), followed by Holmes (1918), described
patients with large lesions in the posterior parietal cortex
(PPC) who exhibited a complex pattern of impairment
(now referred to as Balint syndrome), including characteristic
inaccuracy when they reach or grasp visual objects.
Balint (1909) suggested that this inaccuracy was due to a
visuomotor disconnection, which he called optische
ataxie, whereas Holmes (1918) considered it rather
resulted from a global impairment in spatial perception,
which he called visual disorientation. Subsequent studies
(Garcin et al. 1967; Ratcliff and Davies-Jones 1972;
Vighetto 1980; Vighetto and Perenin 1981; Perenin and
Vighetto 1983, 1988) have progressively argued and
demonstrated that visuomotor deficits for peripheral
targets can occur independently from perceptual disorders,
rather supporting the optic ataxia definition of Balint
(1909). The term ataxie optique was initially used in
French (Garcin et al. 1967) to describe pure cases with
unilateral lesions and deficit restricted to the peripheral
visual field; whereas optische ataxie was used for patients
with Balint syndrome, where praxic aspects are difficult to
disentangle from ataxic symptoms, including visuomotor
problems in central vision. Garcin et al. (1967) have
proposed a number of conditions necessary for diagnosing
optic ataxia: (1) the visual field should be spared in the
area concerned by the visuomotor deficit; (2) proprioception
should be spared; (3) there should be no intrinsic
motor (and oculomotor) and no cerebellar deficit. Interpretation
of optic ataxia as a specific visuomotor disorder
was particularly reinforced by the careful study of
reaching behavior by Vighetto (Vighetto 1980; Perenin
and Vighetto 1988). First, testing auditory-guided movements
showed that optic ataxia appears as a modalityspecific
reaching impairment. Second, although verbal
discrimination of dot position was impaired in some
patients, a direct causal link between these subtle deficits
in visual space perception and the gross misreaching errors
was excluded by the authors. Finally, patients showed
reaching errors related both to a visual field effect and to a
Y. Rossetti . L. Pisella . A. Vighetto
Espace et Action, Unité 534, Institut National de la Santé et de
la Recherche Médicale,
16 avenue Lépine,
Case 13 69676 Bron, France
A. Vighetto
Hôpital Neurologique Pierre Wertheimer,
59 boulevard Pinel,
69003 Lyon, France
Y. Rossetti (*) . L. Pisella . A. Vighetto
IFNL (Institut Fédératif des Neurosciences de Lyon), Bat B13,
Hôpital Neuro-cardiologique,
59 boulevard Pinel,
69394 Lyon cedex 03, France
e-mail: rossetti@lyon.inserm.fr
172
hand effect. This combination of sensory and motor
influences, as well as the localization of the underlying
lesion in-between the visual and the motor areas, supports
the idea that the deficit lies just at the visuomotor interface
rather than solely at either the sensory or motor level.
Superimposition of the lesions of six left brain-damaged
and two right brain-damaged patients with pure optic
ataxia using CT scans (Perenin and Vighetto 1988) has
revealed a symmetrical converging region that includes the
superior parietal lobule, in and above the intraparietal
sulcus and sparing the human inferior parietal lobule. The
localization of the lesion leading to optic ataxia led to this
disorder being considered an impairment of the visual
dorsal stream.
Current dominant theory
In the anatomical context of a dorsal (occipitoparietal)
versus a ventral (occipitotemporal) visual stream, the
visuomotor deficit of optic ataxia has been used as the
main argument to attribute the How function to the dorsal
stream. Indeed, such patients demonstrate an exaggerated
and poorly scaled grip aperture compared with healthy
subjects, although they are not impaired in object recognition
(Jeannerod 1986; Perenin and Vighetto 1988;
Goodale et al. 1991; Jeannerod et al. 1994; Milner and
Goodale 1995). On the other hand, patients with lesion of
the inferotemporal cortex (ventral stream) are impaired in
object recognition (visual agnosia) but remain able to
reach and grasp these same objects that they cannot
describe. The reciprocal patterns of impairment found in
optic ataxia and in visual agnosia have been considered as
a functional double-dissociation (see Rossetti and Revonsuo
2000), leading to the interpretation of these two visual
streams as a dorsal visual pathway for goal-directed action
and a ventral visual pathway for perception (Goodale and
Milner 1992; Jeannerod and Rossetti 1993, Milner and
Goodale 1995; Rossetti 1998). In contrast with these early
interpretations, it is rather important to note that patients
with pure optic ataxia can perform everyday actions,
whereas patients with visual agnosia are strongly impaired
in their daily behavior.
The neuroanatomical basis for a dissociation between
perception and action can also be questioned. A recent
review of anatomical networks between primary visual
area V1 and primary motor area M1 has underlined that
the ventral stream, as well as the dorsal stream, projects
onto the frontal motor areas, and consequently the dorsal
stream has no private pathway toward M1 (see Fig. 1;
Rossetti et al. 2000; Rossetti and Pisella 2002). Furthermore,
no strict segregation between two visual streams
appears within the complex and very interconnected
visuomotor network. It is a temporal dissociation that
can allow dissociation of the two visual streams on the
basis of functional rather than anatomical criteria (Rossetti
and Pisella 2002). Parietal areas exhibit shorter response
latencies to visual stimuli than ventral areas (fastconducting
magnocellular dorsal pathway versus mainly
slow-conducting parvocellular ventral pathway), and the
dorsal pathway has numerous shortcuts and is organized in
a more parallel manner than the ventral stream (justifying
the name proposed by Nowak and Bullier, 1997: the fast
brain).
At the behavioral level, recent analyses of visuomotor
performances of patients with bilateral optic ataxia have
led to a distinct interpretation of optic ataxia and thus of
the function of the dorsal stream in human. Pisella et al.
(1999, 2000) and Gréa et al. (2002) have shown that a
patient with bilateral optic ataxia shows a specific
impairment of on-line visuomotor control rather than a
global impairment of goal-directed actions (see Fig. 3).
These results are discussed below together with those of
other studies that indicate that visuomotor impairments in
optic ataxia are restricted to specific experimental
conditions in which fast and immediate actions are
directed into the contralesional peripheral visual field. In
addition, important clinical observations of the difficulties
faced by patients with optic ataxia and visual agnosia in
Fig. 1 Overview of the visual-to-motor network. Cortical neuronal
networks allowing visual inputs to be transformed into motor
output: the possible substrates for dissociation and interactions
between ventral and dorsal pathways driving information from
primary visual cortex (V1) to primary motor cortex (M1). The dorsal
and the ventral streams are depicted here in green and red,
respectively, as well as their efferences. Blue arrows arise from areas
receiving convergent dorsal and ventral inputs, either directly or
indirectly. Further projections from areas receiving these mixed
convergent inputs have also been represented in blue. Even though
the posterior parietal cortex and the inferior temporal cortex receive
a single direct projection from each other, they were not considered
as mixed recipient areas. By contrast, areas in the frontal lobe
receive parallel dorsal, ventral, and mixed projections. (AIP Anterior
intraparietal area, BS brainstem, Cing. cingulate motor areas, d
dorsal, FEF frontal eye field, FST floor of the superior temporal
sulcus, Hipp. hippocampus, LIP lateral intraparietal area, MIP
mesial intraparietal area, PIP posterior intraparietal area, MST
medial superior temporal area, MT mediotemporal area, PF
prefrontal cortex, PM premotor cortex, SC superior colliculus,
SEF supplementary eye field, SMA supplementary motor area, STS
superior temporal sulcus, STP superior temporal polysensory area,
TE temporal area, TEO temporo-occipital area, v ventral, VIP ventral
intraparietal area.) (Adapted from Rossetti et al. 2000—derived from
Morel and Bullier 1990; Schwarz 1994; Schall et al. 1995; Tanné et
al. 1995; Van Hoesen 1982)
everyday life will be considered. Altogether this review
will shed new light on the interpretation of optic ataxia and
leads us to reconsider and to modify the dominant theory
about the functions of the dorsal and the ventral visual
streams.
Foveal vision versus peripheral vision
The concept of two visual streams was mostly developed
in the late 1960s by Schneider (1969) and Trevarthen
(1968; for previous accounts see Jeannerod and Rossetti
1993). It was based on the functional and phylogenetic
distinction of two visual pathways processing visual
information received on the retina. On the one hand, an
older tectal pathway for ambient vision is involved in
localization of stimuli in space and orientation behavior
(see also Goodale 1983). On the other hand, a more
recently evolved geniculostriate pathway for central vision
participates in visual stimulus identification. The distinction
between these two visual functions was initially
interpreted as a subcortical versus cortical dissociation and
then evolved toward the dissociation within the cortex of
the dorsal versus ventral visual streams (Ungerleider and
Mishkin 1982; reviews by: Goodale and Milner 1992;
Jeannerod and Rossetti 1993; Milner and Goodale 1995;
Pisella and Rossetti 2000; Rossetti and Pisella 2002).
Since subcortical visual structures are now known to
project onto cortical areas, the two types of interpretation
are not necessarily mutually exclusive.
In the early writings about the ventral versus dorsal
anatomical dissociation, specialization of the dorsal stream
for peripheral vision and of the ventral stream for central
vision was suggested and even emphasized (review by
Paillard, 1982). In the context of motor control, the
dissociation between the two streams was correlated with
the functional segregation between position and movement,
i.e., between static and kinetic visual cues.
Consequently, they were supposed to process two kinds
of error signals and to be involved in fast versus slow
corrective visuomotor loops (Bard et al. 1985; review by
Paillard, 1996).
In recent decades, the dorsal stream has been mostly
considered as the vision-for-action (or How) system and it
is rarely emphasized that most patients with optic ataxia
are able to guide accurate actions toward objects in central
vision. This pattern of result has, however, been confirmed
by all studies which have compared movements aimed at
central and peripheral targets in cases of pure optic ataxia.
Out of the ten patients studied by Vighetto (1980; Perenin
and Vighetto 1983, 1988), only three had a deficit in
central vision. This pattern of results should not be
neglected, because it is not as easy to explain as it appears.
In patients with unilateral optic ataxia, this could still be
interpreted in terms of visual fields: since central vision is
analyzed by both hemispheres, the good performance at
fixation could be attributed to the spared hemisphere, as is
the case for the visual field following occipital lesions.
However, patients with pure bilateral optic ataxia also
exhibit normal visuomotor performance in central vision
in many experimental conditions involving either a
pointing or a grasping response (Milner et al. 1999,
Pisella et al. 2000, Gréa et al. 2002). Figure 2 depicts A.T.s
pointing accuracy reported by Milner et al. (1999). The
accuracy of this bilateral patient in central vision is close
to that of control subjects. Figure 3 shows I.G.s grasping
trajectories (from Gréa et al. 2002), which are not
kinematically different from controls for stationary objects
in free vision. In the case of primary visual areas, bilateral
lesions are known to suppress the contralesional macular
sparing that is observed with unilateral lesions, which
demonstrates that central vision is processed in the two
occipital lobes. But, in the case of optic ataxia, it is rather
difficult to propose a satisfactory explanation of the
sparing of visuomotor performance in central vision
following a bilateral lesion of the PPC. To explain this
simple fact, one has to argue that a different system has to
be recruited for action in central vision, or at least that
actions directed to centrally viewed objects must be
processed through visuomotor channel(s) that bypass(es)
the dorsal stream. As a consequence, the specific
implication of the dorsal stream in action would be
restricted to movements guided in peripheral vision. Under
natural conditions, objects are most often first captured
visually, since central vision allows them to be perceptually
analyzed, identified, and selected, before it is
decided to guide an action toward them. The only
exceptions can be found in particular cases of familiar
objects in a familiar environment. But it is reported in the
literature that patients with optic ataxia exhibit far less
visuomotor deficits when they reach and grasp familiar
objects (Jeannerod et al. 1994; Milner et al. 2001). Given
that in everyday life it is rather exceptional to direct
actions toward unfamiliar objects that have never been
seen in central vision, the claim that the dorsal stream is
the main support to our actions clearly has to be
reconsidered.
173
Fig. 2 Central versus peripheral vision. Pointing accuracy of a
patient with a bilateral optic ataxia (A.T.) compared with normal
performance. Mean pointing error is plotted (in millimeters) for
various target positions (in degrees). As is seen in most patients with
pure optic ataxia, pointing accuracy in central vision falls in the
same range as normals. (Redrawn from Milner et al. 1999)
174
Programming versus real-time visuomotor guidance
Since Perenin and Vighetto (1988), who argued that optic
ataxia consists of a general impairment of visuomotor
transformations, many arguments have raised the necessity
to determine more precisely the exact visuomotor process
that is impaired in these patients and thus the precise role
of the human dorsal stream in goal-directed action. From a
kinematic analysis of ataxic grasping movements, Jakobson
et al. (1991) have proposed that the basic deficiency of
optic ataxia is a programming deficit. In a more recent
study, we tried to determine what processing specifically
characterizes action guided toward peripheral targets and
whether we could reveal a related impairment in central
vision (Pisella et al. 2000). A patient with bilateral optic
ataxia (I.G.) due to a bilateral PPC lesion was shown to be
able to reach accurately to targets presented in free vision
(not only for pointing, but also for grasping; Milner et al.
2001; Gréa et al. 2002). However, I.G. was strongly
impaired in the same free-vision condition, when these
targets or objects were experimentally displaced at the
onset of her pointing or grasping movement (see Fig. 3).
She could eventually reach the final target position only
after completing her initially programmed movement and
using slow and intentional corrective processes (Pisella et
al. 1999, 2000; Gréa et al. 2002). In addition, the fast,
unwilled corrections produced by normal subjects in
response to the target jumps (i.e., by what has been called
the hands automatic pilot by Pisella et al. 2000) were not
observed in I.G. Results in normal subjects using
transcranial magnetic stimulation also support the idea
that the dorsal stream is necessary for on-line motor
control (Desmurget et al. 1999). To summarize, converging
results (Desmurget et al. 1999; Pisella et al. 2000;
Gréa et al. 2002) now demonstrate the crucial role of the
dorsal stream in real-time automatic adjustments but not in
movement programming, nor in slow intentional motor
control.
We will thus argue that a specific impairment in realtime
motor control may explain the pattern of deficit
observed in optic ataxia with respect to peripheral/central
vision (Pisella et al. 2000) and with respect to closed-loop
versus open-loop conditions. When motor programming is
realized on the basis of foveal visual information, on-line
visuomotor guidance participates in goal-directed actions
only for minor final adjustments. Conversely, when an
action is programmed on the basis of imprecise peripheral
visual information, on-line visuomotor control appears
essential to quickly adjust the parameters of the ongoing
action in order to succeed in reaching or grasping the goal
(Pélisson et al. 1986; Goodale et al. 1986). Target jumps
occurring at movement onset similarly engage early and
important on-line corrective control. Such on-line processes
have also been described for pointing responses to
stationary targets in visual open-loop (i.e., without visual
feedback) conditions (Prablanc et al. 1986; Pélisson et al.
1986; Goodale et al. 1986). Vighetto (1980) has tested
three patients with unilateral optic ataxia in open-loop
versus closed-loop pointing conditions. He has shown that
lack of vision of the hand during movement execution
dramatically affects pointing accuracy in these patients
(see also Fig. 4). Vighetto (1980) has also reported that, in
the closed-loop condition, patients spontaneously increase
their movement duration. It could be suggested that
slowing the movement allows patients to use slow
feedback loops based on the visual reafference of hand
position, which consequently improves the final pointing
performance. Vighetto (1980) has also tested the effect of
viewing the hand at starting position. Pointing accuracy is
improved significantly in conditions where vision of the
hands starting position is provided, probably again
because this condition allows a better movement programming
ad hence reduces the need for on-line corrective
processes (Rossetti et al. 1994). Interestingly the view of
Paillard (1996) leads to the same hypothesis, that the
dorsal system is specialized for the fast corrective loop and
would predict as well that its need increase when action is
guided to peripheral targets. Indeed the dorsal system
allows quick adjustment of movement trajectory based on
the kinetic feedback of the moving arm seen in peripheral
vision, which becomes more important during reaching to
peripheral targets, since action is directed away from eye
fixation.
It is interesting to reconsider the arguments found in the
literature in favor of a programming deficit in optic ataxia
in the light of these recent results. We will see that most
Fig. 3 Online correction deficit.
When a grasping movement is
directed to objects presented in
free vision in a central (C) or
right (R) location, a patient with
a bilateral optic ataxia can reach
accurately to and grasp them.
However, she exhibits strong
difficulties in updating her trajectories
on-line in response to a
target displacement from C to R
(CR) triggered at movement
onset. In this situation, artificially
raising the need for online
corrections, I.G. completed
her initial movement before
initiating a second movement to
the final position of the object
data interpreted as reflecting a programming deficit can
also be seen as consequences of a deficit in real-time
motor control. For example, the abnormally long deceleration
period and large grip aperture described in optic
ataxia patients (Jeannerod 1986; Jakobson et al. 1991)
could be considered as compensation for the lack of online
control processes, which are applied toward the end of
the movement. Similarly, the increase in hand movement
latency observed after a lesion of area 7 in monkeys
(Faugier-Grimaud et al. 1985) and in man (Vighetto 1980;
Jeannerod 1986; Jakobson et al. 1991) may be explained
by the need to refine the programming of movements in
order to compensate for the on-line control deficit. This is
consistent with the recent evidence detailed below that
delaying action can improve visuomotor performance in
optic ataxia.
Altogether the above data provide strong evidence for a
deficit of real-time visuomotor guidance, but the definitive
demonstration of an additional programming deficit
remains to be done. Recent results may indeed suggest
that the early kinematics of reaching movements are
impaired in some patients with optic ataxia (Roy et al.
2002; Milner et al. 2003). These results could be related to
the impairment of reaching observed following parietooccipital
(PO) lesion in the monkey (Galletti et al. 2003).
In the same way as it is specifically the fast (versus the
slow), on-line motor corrections that are impaired in optic
ataxia, it remains to be determined whether an impairment
of visuomotor programming might only be observed when
temporal constraints are specifically imposed on movement
initiation (see Milner et al. 2003; and below).
Immediate versus delayed movement
The classical distinction between programming versus online
control referred to in the previous section is currently
being reconsidered, and these two aspects of motor control
are progressively conceived as more and more interdependent
(review by Desmurget et al., 1998; see also Figs.
4–15 in Rossetti and Pisella 2002). Accordingly, they
might be alternatively described as a distinction between
fast immediate visuomotor transformations and slow
immediate visuomotor transformations (Rossetti and
Pisella 2003). As mentioned earlier, the microscopic and
functional dissociation between the two main visual
streams are more important to consider than only their
relative anatomical segregation. This functional dissociation
is mainly based on the temporal properties of transfer
of visual information along the two streams (Nowak and
Bullier 1997). These properties allow the prediction that,
unlike healthy controls, patients with lesion of the PPC
would perform delayed action better than immediate
action and that the residual visuomotor abilities of patients
with visual agnosia should be dramatically affected by a
memory delay. Both of these predictions have been
supported by neuropsychological observations. Milner et
al. (1999) have shown that patients with bilateral optic
ataxia are more accurate in delayed pointing than in
immediate pointing to peripheral targets. The same
paradoxical improvement of visuomotor performance in
these patients (contrary to what happens with delay in
normal subjects) was also observed for the grasp component
(Milner et al. 2001, 2003). Note that these results
strongly challenge the view that the integrity of the dorsal
stream is necessary for performing actions. They suggest
that the impairment found in the peripheral visual field of
patients with optic ataxia is specific to fast visuomotor
processes, as already suggested by Pisella et al. (2000).
Interestingly this paradoxical pattern of results is clearly
dissociated from the disruption of visuomotor functions
found in visual agnosia in delayed tasks (Goodale et al.
1994). This clear dissociation could be interpreted as the
most demonstrative argument for a double dissociation
between visual agnosia and optic ataxia (see Milner et al.
2003). However, the visuomotor performance of the two
types of patients have not yet been tested in similar
conditions, especially with respect to the distinction
between central and peripheral vision.
If one considers the ecological usefulness of delayed
sensorimotor reactions, then another pattern of dissociation
may be proposed. In many areas of life within a social
environment, it may be more productive not to react
immediately to attractive stimuli. Therefore the immediate
reaction systems have to be supervised by more complex
levels of cognition. That is exactly what patients with
environment-dependency syndrome (Lhermitte 1986)
appear to lack. This syndrome results in the release of
immediate reactions to visual stimuli and is consecutive to
frontal lesion. Therefore the issue of immediate versus
delayed action suggests that optic ataxia might be better
dissociated from patients with frontal lesions than with
175
Fig. 4 Open loop versus closed loop. Pointing accuracy of patient
O.K. with unilateral optic ataxia following a lesion of the right
posterior parietal cortex. This patient exhibits errors due to a
combination of a hand effect and a field effect (see Revol et al.
2003). When pointing with the right (ipsilesional) hand, ataxic errors
in the left visual field due to the field effect mainly expressed in
terms of variable errors (not represented here). Lateral error is
plotted here when pointing to targets at different eccentricities with
the right hand in experimental conditions where visual feedback of
the hand is available during movement execution (Closed-Loop) or
not (Open-Loop). The open-loop condition reveals the deficit of the
patient in terms of lateral errors in the left visual field (including
central vision)
176
lesion of the ventral stream (see Rossetti and Pisella 2003).
In this framework, optic ataxia would be considered as a
specific deficit for immediate (i.e., nonmediated) visuomotor
responses.
Everyday life versus experimental conditions
Action, in contrast to movement or behavior, refers to the
ability to produce a deliberate sequence of motor activation
that produces a desired effect on the body–environment
relationship. Therefore action should be associated
with intention (see Revonsuo and Rossetti 2000), and
immediate motor control can be regarded as a local
function of the whole system (Fig. 5). This definition of
action is to be contrasted with its frequent use in
neuroscience and experimental psychology. The frequent
association of voluntary with action should be regarded as
a pleonasm, but it also results from the overuse of action
terminology. Action is broadly used whenever a motor
response is required from the subject. However, it is rather
obvious that simple key-pressing tasks used in most
laboratories cannot be considered as action. The only use
of reaction time measurements in such experiment reflects
the fact that reactions rather than actions are processed in
these tasks. Most experimental conditions designed to
explore voluntary actions actually investigate reactions to
an expected stimulus. These reactions are most often
overlearned in everyday life (e.g., pointing or grasping,
key-pressing) or are investigated after a learning session in
the case of more complex tasks. Therefore the intention of
the subject may be restricted to the will to perform the task
or to comply to the instructions. Clinical observation and
examination may provide a less constrained way to
explore intentional aspects of behavior.
The most important and relevant aspects to mention
about everyday consequences of neurological deficit are
the spontaneous complaints from patients with optic ataxia
and visual agnosia. Most of the patients with unilateral
optic ataxia do not complain of any action deficit in
everyday life. In the group study by Vighetto and Perenin
(1981), half of the patients had never complained of
functional deficit before the systematic search for movement
guidance in peripheral vision, and the other half of
the patients included cases with other symptoms such as
mild sensory deficit or praxic problems. Patients with a
bilateral lesion often complain about functional deficits,
from which it is difficult to disentangle the impact of pure
optic ataxia and that of symptoms due to other parietal or
more remote damage. This is evident in the case of Balint
syndrome.
Reports about the functional deficit of patients with
optic ataxia in everyday life are usually brief. Little
mention of functional deficit has been provided even for
patients with a bilateral lesion (e.g., Milner et al. 1999,
2001; Pisella et al. 2000). Jeannerod et al. (1994)
described in more detail the case of the bilateral patient
A.T.:
Visual recognition of shape, texture, depth, colour,
faces had always been normal. AT never complained
of any difficulty concerning perception of movement....Despite
some recovery with the passing years
of rehabilitation programs, AT still presented a severe
disorientation. Indeed she was hampered in her
everyday life for actions like dressing, cooking,
ironing, sewing or drivingHowever, it is not clear
whether the whole deficit exhibited by A.T. can be
accounted for by her optic ataxia, since she also
exhibited several problems with more perceptual
functions (which is also compatible with her rather
large lesion; see Michel and Hénaff, in press). The
observation made by Jeannerod et al. (1994) was that
A.T. was mostly impaired in peripheral vision when
she had to grasp unfamiliar objects. It was later
shown that another bilateral optic ataxia patient (I.G.)
could learn to size her grip to unfamiliar objects
across several experimental sessions (Milner et al.
2001). These two observations show that the grasping
deficit observed in peripheral vision should not be
related to a gross visuomotor deficit but rather to a
deficit in determining the grip size appropriate to
unfamiliar objects, i.e., when only metric cues are
available in the visual environment.
In our clinical experience, patients such as I.G. with
bilateral lesions (Pisella et al. 2000) may, for example,
exhibit difficulties in taking the first step on a staircase.
They also have problems navigating in a complex dynamic
environment (e.g., a train station), in that they have to stop
repeatedly and slow down their walking speed considerFig.
5 Visually guided action versus immediate visuomotor control.
Action refers to the ability to produce a deliberate sequence of motor
activation. Therefore action should be associated with both intention
and perceptual input, both factors contributing to initial decision
process and to the planning of action. Immediate motor control can
thus be regarded as local regulation of the whole system
ably. One other typical complaint of bilateral patients is of
a slowness and clumsiness in writing. I.G.s deficit thus
appears to be mostly confined to situations where she has
to react to visual information in real time or when facing a
complex environment.
In contrast, patients with visual agnosia spontaneously
complain of huge difficulties in everyday life. They cannot
find familiar objects or select the appropriate tool for daily
tasks. They have to rely on specific cues such as context,
touch, and noise, and strategically put objects in fixed,
predetermined places so that they can find them without
recognizing them.
The patients difficulties are so obvious that descriptions
often focus on the few residual functions that patients can
perform in everyday life. The famous case D.F., who was
initially described by Milner and Heywood (1989), had a
constructional apraxia. However, she could manage some
everyday activities such as shopping and cooking by
2 years postonset (Milner et al. 1991). As in most patients
with visual agnosia, one ultimate consequence of the
deficit was a loss of spontaneity in her behaviour (Milner
et al. 1991, p. 410). A more recent description by Lê et al.
(2002) emphasized the fact that their patient (S.B.) could
move in space without apparent difficulties and grasp
moving objects (table tennis balls), i.e., when no goal
selection was involved. But, for example, S.B. could not
even choose his food at the cafeteria, as is typically
observed in other patients with visual agnosia. Therefore
the cluster of residual functions maintained in visual
agnosia appears to be mostly related to intact, on-line
motion-processing abilities and their use in controlling
movement (e.g., Lê et al. 2002).
Taken altogether these clinical considerations allow us
to draw the conclusion that, for everyday-life actions,
patients with visual agnosia are impaired at least as much
as patients with optic ataxia. Of course this does not
contradict the fact that under experimental conditions
where so-called actions are limited to reactions to a visual
target or object—presented in peripheral vision—patients
with optic ataxia are more impaired than patients with
visual agnosia. But one has to be aware that it is an
obvious constraint of these experimental conditions that
they emphasize the specific need for on-line visuomotor
guidance whereas the contribution of other aspects of
action control is artificially reduced.
As mentioned earlier, another line of thought has been
to propose a contrast between optic ataxia and the
environment-dependency syndrome observed following
frontal lesions (Lhermitte 1986; Pisella et al. 2000;
Rossetti and Pisella 2003). This loss of autonomy is
observed both for social stimuli (eliciting imitation
behavior) and for physical stimuli (eliciting utilization
behavior; Lhermitte et al. 1986). Rather, the patients
behavior appears as though implicit in the environment
was an order to respond to the situation in which the
patients find themselves (Lhermitte 1986). In other words
these patients appear to be reacting immediately to
external stimuli without being able to shift their action
in time or in space, which may appear to be the opposite
pattern to that described in optic ataxia. As already
proposed by Pisella et al. (2000), Frith et al. (2000), and
Rossetti and Pisella (2003), it may be that optic ataxia can
be better dissociated from this environmental-dependency
syndrome than from visual agnosia.
Discussion
We have reviewed several axes along which dissociations
between optic ataxia and visual agnosia have been
proposed, and arguments for a dissociation between
perception and action have been made. The issues of
foveal versus peripheral vision, the temporal constraints of
visual information processing, and the necessary reductionism
of experimental conditions used to study action
are the key factors that should be kept in mind before
arguing for a double dissociation between optic ataxia and
visual agnosia, as well as between vision for action and
vision for perception. It is also obvious that many visual
streams can contribute to action, none of them disrupting
all types of action. This is compatible with the variety and
the complexity of projection streams that can be described
between the visual input and the visual output (Rossetti
and Pisella 2002).
In contrast to local sensorimotor processes, a true action
involves many aspects of perception. Complex actions
imply a continuous perceptual control of the movement
sequence and of possible side effects. Simple actions
imply that an appropriate goal is selected in a rich
environment. In this framework the postulated double
dissociation between optic ataxia and visual agnosia can
obviously not be supported by a simple distinction
between global perception and global action. The former
idea of a double dissociation between perception and
action was based on the assumption that all visual
perception was altered in visual agnosia (which is mostly
the case in at least D.F.) but also considered optic ataxia as
being an action disorder. More recent studies have shown
that the aspects of action that are impaired in optic ataxia
are quite restricted. Optic ataxia patients can perform all
types of natural actions and are impaired only when a time
constraint is imposed: their on-line motor updating is
altered and their reaches toward peripheral targets are
impaired when they have to produce immediate responses
toward unknown objects. The strongest argument for a
dissociation between optic ataxia and visual agnosia is
now grounded on temporal parameters. Our main argument
here is that optic ataxia does not appear as a general
deficit of action but rather as a specific deficit localized at
a restricted level of action organization that is immediate
visuomotor control. A crucial role appears to be played by
real-time visual processing of both the goal and the arm
for this limited aspect of action that is impaired in optic
ataxia. Several higher levels of action organization can be
affected by other syndromes such as apraxias, unilateral
neglect or frontal syndromes, resulting from the unilateral
lesion of functions with high hemispheric lateralization. In
contrast to optic ataxia, affecting a low-level of the
177
178
interface between visual input and motor output, these
cognitive disorders often produce deficits for both hands
and in both hemifields.
Three clear lines of investigation emerge from the
present review. First, to build a strong argument for a
double dissociation between optic ataxia and visual
agnosia, further studies should either test visuomotor
performance of optic ataxia patients in central vision (as is
done for visual agnosia) or study visuomotor performance
of visual agnosia patients in peripheral vision (as is done
for optic ataxia). The visuomotor performance of patients
with visual agnosia are never explored in peripheral vision
as is the case in most reports about optic ataxia, and the
perceptual performance of optic ataxia patients have not
been described in peripheral vision to allow comparison
with their motor impairment. A second crucial area of
investigation should explore the important role of time in
the dissociations found within the visual system. This area
of research is currently growing fast. The third line of
investigation concerns the cluster of characteristics that are
impaired in optic ataxia: it would be interesting for future
studies to investigate the functional links between fast, online
visual processing, immediate visuomotor control, and
peripheral vision. One possible link could be that these
situations are mainly realized using automatic control
processes.
Acknowledgements The authors wish to thank Aarlenne Kahn,
Rob McIntosh, Hisaaki Ota, Patrice Revol, Caroline Tilikete, and
Gilles Rode for fruitful discussions. This work was supported by
grants from the McDonnell-Pew Foundation and the Leverhulme
trust (F/00128/0).
References
Balint R (1909) Seelenlähmung des Schauens, optische Ataxie,
raümliche Störung der Aufmerksamkeit. Monats Psychiatr
Neurol 25:51–81
Bard C, Hay L, Fleury M (1985) Role of peripheral vision in the
directional control of rapid aiming movements. Can J Psychol
39:151–161
Desmurget M, Pélisson D, Rossetti Y, Prablanc C (1998) From eye
to hand: planning goal-directed movements. Neurosci Biobehav
Rev 22:761–788
Desmurget M, Epstein CM, Turner RS, Prablanc C, Alexander GE,
Grafton ST (1999) Role of the posterior parietal cortex in
updating reaching movements to a visual target. Nat Neurosci
2:563–567
Faugier-Grimaud S, Frenois C, Peronnet F (1985) Effects of
posterior parietal lesions on visually guided movements in
monkeys. Exp Brain Res 59:125–128
Frith CD, Blakemore SJ, Wolpert DM (2000) Abnormalities in the
awareness and control of action. Philos Trans R Soc Lond B
Biol Sci 29:1771–1788
Galletti C, Kutz DF, Gamberini M, Breveglieri R, Fattori P (2003)
Role of the medial parieto-occipital cortex in the control of
reaching and grasping movements. Exp Brain Res DOI
10.1007/s00221-003-1589-z (this issue)
Garcin R, Rondot P, Recondo de J (1967) Optic ataxia localized in
two left homonymous visual hemifields (clinical study with
film presentation). Rev Neurol (Paris) 116:707–714
Goodale MA (1983) Neural mechanisms of visual orientation in
rodents: targets versus places. In: Hein A, Jeannerod M (eds)
Spatially oriented behavior. Springer, New York, pp 34–61
Goodale MA, Milner AD (1992) Separate visual pathways for
perception and action. Trends Neurosci 15:20–25
Goodale MA, Pélisson D, Prablanc C (1986) Large adjustments in
visually guided reaching do not depend on vision of the hand or
perception of target displacement. Nature 320:748–750
Goodale MA, Milner AD, Jakobson LS, Carey DP (1991)
Perceiving the world and grasping it. A neurological dissociation.
Nature 349:154–156
Goodale MA., Jakobson LS, Keillor JM (1994) Differences in the
visual control of pantomimed and natural grasping movements.
Neuropsychologia 32:1159–1178
Gréa H, Pisella L, Rossetti Y, Desmurget M, Tilikete C, Prablanc C,
Vighetto A (2002) A Lesion of the posterior parietal cortex
disrupts on-line adjustments during aiming movements. Neuropsychologia
40:2471–80
Holmes G (1918) Disturbances of visual orientation. Br J
Ophthalmol 2:449–506
Jakobson LS, Archibald YM, Carey DP, Goodale MA (1991) A
kinematic analysis of reaching and grasping movements in a
patient recovering from optic ataxia. Neuropsychologia
29:803–809
Jeannerod M (1986) Mechanisms of visuomotor coordination: a
study of normal and brain-damaged subjects. Neuropsychologia
24:41–78
Jeannerod M, Rossetti Y (1993) Visuomotor coordination as a
dissociable function: experimental and clinical evidence. In:
Kennard C (ed) Visual perceptual defects. (Baillières clinical
neurology: International practice and research) Baillière
Tindall, London, pp 439–460
Jeannerod M, Decety J, Michel F (1994) Impairment of grasping
movements following a bilateral posterior parietal lesion.
Neuropsychologia 32:369–380
Lê S, Cardebat D, Boulanouar K, Hénaff M-A, Michel F, Milner
AD, Dijkerman C, Puel M, Démonet J-F (2002) Seeing, since
childhood, without ventral stream: a behavioural study. Brain
125:58–74
Lhermitte F (1986) Human autonomy and the frontal lobes. II.
Patient behavior in complex and social situations: the
environmental dependency syndrome. Ann Neurol 19:335–343
Lhermitte F, Pillon B, Serdaru M (1986) Human autonomy and the
frontal lobes. I. Imitation and utilization behavior: a neuropsychological
study of 75 patients. Ann Neurol 19:326–334
Michel F, Hénaff M-A (In press) To see without occipito-parietal
cortex: simultanagnosia as an attentional visual field shrinking.
Behav Neurol
Milner AD, Goodale MA (1995) The visual brain in action. Oxford
University Press, Oxford
Milner AD, Heywood CA (1989) A disorder of lightness discrimination
in a case of visual form agnosia. Cortex 25:489–494
Milner AD, Perrett DI, Johnston RS, Benson PJ, Jordan TR, Heeley
DW, Bettucci D, Mortara F, Mutani R, Terazzi E, Davidson
DLW (1991) Perception and action in visual form agnosia.
Brain 114:405–28
Milner AD, Paulignan Y, Dijkerman HC, Michel F, Jeannerod M
(1999) A paradoxical improvement of optic ataxia with delay:
new evidence for two separate neural systems for visual
localization. Proc R Soc Lond B Biol 266:2225–30
Milner AD, Dijkerman HC, Pisella L, McIntosh RD, Tilikete C,
Vighetto A, Rossetti Y (2001) Grasping the past: delay can
improve visuomotor performance. Curr Biol 11:1896–1901
Milner AD, Dijkerman HC, McIntosh RD, Rossetti Y, Pisella L
(2003) Delayed reaching and grasping in patients with optic
ataxia. Prog Brain Res 142:225–242
Nowak L, Bullier J (1997) The timing of information transfer in the
visual system. Cereb Cortex 12:205–241
Paillard J (1982) The contribution of peripheral and central vision to
visually guided reaching. In: Ingle DJ, Goodale MA, Mansfield
RJW (eds) Analysis of visual behavior. MIT press, Cambridge,
pp 367–385
Paillard J (1996) Fast and slow feedback loops for the visual
correction of spatial errors in a pointing task: a reappraisal. Can
J Physiol Pharmacol 74:401–417
Pélisson D, Prablanc C, Goodale MA, Jeannerod M (1986) Visual
control of reaching movements without vision of the limb. II.
Evidence of fast unconscious processes correcting the trajectory
of the hand to the final position of a double step stimulus. Exp
Brain Res 62:303–311
Perenin M-T, Vighetto A (1983) Optic ataxia: a specific disorder in
visuomotor coordination. In: Hein A, Jeannerod M (eds)
Spatially oriented behavior. Springer, New York, pp 305–326
Perenin M-T, Vighetto A (1988) Optic ataxia: a specific disruption
in visuomotor mechanisms. I. Different aspects of the deficit in
reaching for objects. Brain 111:643–674
Pisella L, Rossetti Y (2000) Interaction between conscious identification
and non-conscious sensori-motor processing: temporal
constraints. In: Rossetti Y, Revonsuo A (eds) Beyond dissociation:
interaction between dissociated implicit and explicit
processing. Benjamins, Amsterdam, pp 129-151
Pisella L, Tilikete C, Rode G, Boisson D, Vighetto A, Rossetti Y
(1999) Automatic corrections prevail in spite of an instructed
stopping response. In: Grealy M, Thomson JA (eds) Studies in
perception and action. Erlbaum, New York, pp 275–279
Pisella L, Gréa H, Tiliket C, Vighetto A, Desmurget M, Rode G,
Boisson D, Rossetti Y (2000) An automatic pilot for the hand in
the human posterior parietal cortex toward a reinterpretation of
optic ataxia. Nat Neurosci 3:729–736
Prablanc C, Pélisson D, Goodale MA (1986) Role of retinal
feedback of target position in guiding the hand. Exp Brain Res
62:293–302
Ratcliff G, Davies-Jones GA (1972) Defective visual localization in
focal brain wounds. Brain 95:49–60
Revol P, Rossetti Y, Vighetto A, Rode G, Boisson D, Pisella L
(2003) Pointing errors in immediate and delayed conditions in
unilateral optic ataxia. In: Special issue, Vision and real-time
control of action. Spat Vis 16:347–364
Revonsuo A, Rossetti Y (2000) Dissociation and interaction:
windows to the hidden mechanisms of consciousness. In:
Rossetti Y, Revonsuo A (eds) Beyond dissociation: interaction
between dissociated implicit and explicit processing. Benjamins,
Amsterdam, pp 351–366
Rossetti Y (1998) Implicit short-lived motor representation of space
in brain-damaged and healthy subjects. Conscious Cogn 7:520–
558
Rossetti Y, Pisella L (2002) Several vision for action systems: a
guide to dissociating and integrating dorsal and ventral
functions. In: Prinz W, Hommel B (eds) Common mechanisms
in perception and action. (Attention and performance XIX)
Oxford University Press, New York, pp 62–119
Rossetti Y, Pisella L (2003) Mediate responses as direct evidence for
intention: neuropsychology of not to-, not now- and not theretasks.
In: Johnson SH (ed) Taking action: cognitive neuroscience
perspectives on intentional acts. MIT Press, Cambridge,
MA, pp 67–105
Rossetti Y, Revonsuo A (2000) Beyond dissociations: recomposing
the mind-brain after all? In: Rossetti Y, Revonsuo A (eds)
Beyond dissociation: interaction between dissociated implicit
and explicit processing. Benjamins, Amsterdam, pp 1–16
Rossetti Y, Stelmach GE, Desmurget M, Prablanc C, Jeannerod M
(1994) Influence of viewing the static hand prior to movement
onset on pointing kinematics and accuracy. Exp Brain Res
101:323–330
Rossetti Y, Pisella L, Pélisson D (2000) Eye blindness and hand
sight: temporal aspects of visuomotor processing. Vis Cognit
7:785–809
Roy AC, Gallese, V, Pavesi G, Gentilucci M (2002) Earliest
kinematics parameters of prehension movement are not scaled
to the object properties in ataxia: a case study. Soc Neurosci
Abstr 268.1
Schneider GE (1969) Two visual systems. Science 163:895–902
Trevarthen CB (1968) Two mechanisms of vision in primates.
Psychol Forschung 31:99–337
Ungerleider LG, Mishkin M (1982) Two cortical visual systems. In:
Ingle DJ, Goodale MA, Mansfield RJW (eds) Analysis of
visual behavior. MIT press, Cambridge, MA, pp 549–586
Vighetto A (1980) Etude neuropsychologique et psychophysique de
lataxie optique. Thèse Université Claude Bernard Lyon I
Vighetto A, Perenin M-T (1981) Optic ataxia: analysis of eye and
hand responses in pointing at visual targets. Rev Neurol (Paris)
137:357–72
179