 |
 |
|
|
1st Joint meeting of
The Institute for Cognitive Science (ISC),
Bron
&
The Institute
of Cognitive Neuroscience (ICN), London
From perception to action representation :
Neural bases and disorders
25-26 septembre 2000
Organised by D. Boussaoud & P. Haggard
ISC, 67 Boulevard Pinel, 69675 Bron cedex
Phone: 33 (0)4 37 91 12 61; Fax: 33 (0)4 37
91 12 10
Contact: Driss Boussaoud (boussaoud@isc.cnrs.fr)
|
|
|
|
|
|
|
|
|
| From perception to action representation
:
Neural bases and disorders |
Monday, september 25
12:00 Registration
13:45 Opening and welcome
Attention and space representation
Chair: D. Wolpert
14:00 P. Haggard (ICN, London) Awareness, binding and causation
14:30 J-R Duhamel (ISC, Bron) Neuronal basis of space representation
15:00 Y. Rossetti (INSERM U534, Bron) Several different “vision for action” systems
15:30 Coffee
Perception and action
Chair: P. Haggard
16:00 M-T Perenin (INSERM U371, Bron) Perception and action in unilateral neglect
16:30 A. Sirigu (ISC, Bron) The parietal cortex and the representation of action
17:00 L. Parsons (San Antonio, Texas) New Findings on the Role of the Cerebellum in Perception and Action.
17:30 C. Tallon-Baudry and O. Bertrand (INSERM U280, Lyon) Gamma oscillations and object representation in humans
18:00
Chair: T. Shallice
18:30 Plenary lecture by M.
Jeannerod
Internal simulation of action as a unified framework for motor cognition
19:30 Session ends
20:00 Diner
| From perception to action representation
:
Neural bases and disorders |
Tuesday september 26
Cognitive functions and their disorders
Chair: P. Jacob
9:00 T. Shallice (ICN, London) Fractionation of the Executive systems
9:30 C. Frith (ICN, London) Consciousness and the functions of the prefrontal cortex
10:00 N. Georgieff (Hôpital Vinatier & ISC, Bron) Agency and its dysfunction
10:30 M. Saoud (Hôpital Vinatier, Bron) Context processing and action in schizophrenia
11:00 Coffee and poster viewing
11:30
12:00 Lunch
Modelling and psychophysics
Chair: C. Prablanc
14:00 J. Driver (ICN, London) Polysensory attention
14:30 Y. Paulignan (ISC, Bron) Perceptual awareness versus action
15:00 D. Wolpert (ICN, London) Computational human sensorimotor control
15:30 P. Burgess (ICN, London) The cognitive and neuroanatomical correlates of multitasking
16:00 Coffee
| From perception to action representation : Neural bases and disorders |
16:30 Discussion Groups
Titles
ACTION THEME
Visuomotor adjustments: differences between motor
performance and awareness of action
Helen Johnson and Patrick Haggard
The coding of movements in primary motor cortex:
a TMS study
A.F. Hamilton, K.E. Jones, Z. Ghahramani*, R.N. Lemon & D.M. Wolpert
State estimation under stochastic and deterministic
time-varying contexts
Philipp Vetter & Daniel
M. Wolpert
Sensorimotor integration of retinal and gaze signals
during smooth pursuit eye movements
Robert J.
van Beers1, Daniel M. Wolpert2, Patrick Haggard1
Prism adaptation for different starting arm postures
Pierre Baraduc and Daniel Wolpert
Reaching and grasping in the monkey: effect of perturbation
of object size and location.
A.C. Roy, Y.Paulignan, D.Boussaoud.
Neuronal activity related to bimanual coordination
in the motor and premotor cortices of rhesus monkeys.
C. Jouffrais, E.M. Rouiller. Institute of Physiology. University of
Fribourg
SPATIAL THEME
Is the Bell Inequality Violated in Visual Perception?
Andrew Duggins, Geraint Rees, Chris Frith
Vision and touch through the looking glass: cognitive
remapping of multisensory information in the normal and damaged brain.
Angelo Maravita1, Charles Spence 2, Clair Sergeant 1, Karen Clarke
1, Masud Husain, & Jon Driver 1
Two systems of spatial representations in human
posterior parietal cortex
S. Clavagnier*, I Schindler, Y. Paulignan, A. Vighetto, MT. Perenin
Convergent and divergent effects of neck proprioceptive
and visual motion stimulation on ego- and allocentric space in neglect
Igor Schindler1, Georg Kerkhoff²
Spatial attention and memory vs motor preparation.
The involvement of the premotor cortices as assessed by fMRI.
Stéphane Simon*, Martine Meunierº, Anna Berardi*, Loÿs
Piettre*, Christoph Segebarth*, Driss Boussaoudº
Beware and be aware: Capture of attention by
emotional stimuli in patients with hemispatial neglect
Patrik Vuilleumier (1,2) & Sophie Schwartz (1,3)
Impaired visual search and extinction following reversible
inactivation of monkey lateral intraparietal area (LIP).
Claire Wardak and Jean-René Duhamel
Non-informative Vision Enhances Tactile Acuity
Steffan Kennett, Marisa Taylor-Clarke & Patrick Haggard
Predicting
the future position of a looming target in 3D space: the role of retino-,
head- and trunk-centered reference frames in man.
Marco Neppi-Mòdona*°, David Auclair,° Jean-René
Duhamel°, Angela Sirigu°
* Facoltà di Psicologia, Università di Torino, Italia;
°Institut des Sciences Cognitives, Bron, France
EXECUTIVE THEME
Agency judgement in schizophrenic patients.
C. Farrer , N. Franck , N.Georgieff , M. Marie-Cardine, J. Daléry,
T. d ’Amato, & M. Jeannerod
In search of a monitoring deficit of action in
schizophrenia : some arguments for a revision of this assumption.
Fourneret P 1., Franck N 1.2., Georgieff N 1.2., Jeannerod M 1
Can you remember the same way twice? Changes in the
structure of autobiographical episodes recalled repeatedly at a short interval.
Laure Coates & Paul Burgess
Important deficit of action generation in schizophrenia:
two experimental studies
Posada, N. Franck, N. Georgieff & M. Jeannerod.
Modelling the Somatoform and Dissociative (Conversion)
Disorders
Richard J. Brown, PhD.
Predicting the sensory consequences of action
S-J.
Blakemore, D.M. Wolpert & C.D. Frith
Planning and action knowledge in schizophrenic patients:
Comparisons with patients with frontal lobe damage.
Tiziana Zalla, Andres Posada, Nicolas Franck, Nicolas Georgieff, Angela
Sirigu
18:00 Meeting ends
| From perception to action representation
:
Neural bases and disorders |
Patrick Haggard, Institute of Cognitive Neuroscience, UCL
The association between our intentions, our actions, and the effects these have on our environment are a key achievement of the human mind. To achieve the status of a conscious agent, my mental processes must mark the external events caused by my intentional actions. We call this process "efferent binding".
This talk will consider the subjective timing aspects of efferent binding. Several converging lines of evidence have shown that the perceived time of discrete actions often precedes the onset of actual muscular movement. We shall argue that psychological representations of actions arise partly from neural premotor processing.
We now report two new experiments on the perceived time of stimuli (beeps) and actions (keypresses) which occur in contingent pairs. In the first experiment, we asked subjects to estimate the perceived time of stimuli and of actions in separate blocks. The contingent relation between stimulus and action could run in either direction according to condition: either the stimuli elicited actions (in an SRT task), or the actions elicited by stimuli (in an operant task). Subjects' estimates of stimulus timing and of action timing were compared with the perceived times of stimuli or actions made in isolation. The method of Libet et al. (1983) was used to obtain timing judgements. The biasing effects on the perceived time of stimuli and of actions of the additional event in the SRT and operant tasks were considered.
Briefly, we found strong perceptual attraction effects between percepts of the two events. These are interpreted as evidence for efferent binding effects processes that influence conscious awareness. In experiment 1, the causal event in each contingent task was perceived to occur at its normal time, whereas the consequent event was biased towards the cause. Thus, in SRT, stimuli were perceptually fixed in time, whereas actions were perceptually attracted towards them. In operance, actions were perceptually fixed in time, whereas stimuli were perceptually attracted towards them. Experiment 2 compared timing judgements between the contingent tasks studied in the first experiment, and non-contingent, sequential tasks (two stimuli occurring in succession, or two actions made in succession). Only contingent tasks showed the pattern of fixed causes and conflated effects.
These results suggest a specific perceptual attraction effect in the perceived time of events occurring in our basic interactions with our world. It applies equally to sensory events and to motor events, and applies equally to operance and reaction. Perceptual attraction may narrow the window of psychological time over which the mind must construct associations, and may allow effects to be individuated by their causes. This could facilitate the process of efferent binding, and could thus underlie our sense of agency.
Acknowledgements: supported by MRC, BBSRC, DAAD and MPG.
Neuronal basis of space representation
Jean-René Duhamel,
Institut des Sciences Cognitives, CNRS, France
The parietal lobe contains multiple specialized areas which contribute
to sensory analysis and movement preparation. Single cell recording evidence
and the analysis of anatomical connectivity patterns in non-human primates
reveal both similarities and differences in the core functional characteristics
of each of the currently recognized parietal subdivisions. One of the main
common features is that parietal areas integrate multiple sources of signals
which are combined, presumably through population coding, in order to provide
the spatial coordinates of directed actions. Investigation of one
parietal subdivision, the ventral intraparietal area (VIP), demonstrates
in particular that the activity of multisensory neurons is influenced by
eye position in the orbit and that their receptive field, i.e. the region
of the environment for which such neurons demonstrate spatial selectivity,
is encoded in coordinates which can be intermediate between a retinal and
a head-centered reference frame. Neural network modeling suggests that
such a complex form of coding could serve to perform multidirectional coordinate
transformations across different sensory and motor modalities. Preliminary
evidence from human psychophysical experiments will also be presented in
support of the concept of multiple reference frames for encoding spatial
locations.
Several different ‘vision for action’ systems
Yves Rossetti and Laure Pisella
Espace et Action, Institut National de la Santé Et de la Recherche
Médicale, Unité 534
16 avenue Lépine, Case 13, 69676 Bron - France
There is a well-established argument for a double dissociation between
vision for action and vision for conscious identification1. However, the
complex neuroanatomical networks involved in vision, perception and action
does not allow for a strict segregation between two cortical visual pathways,
since numerous direct and indirect interconnections and convergence can
be observed between the dorsal and the ventral stream2. Accordingly, evidence
for a double interaction between the two visual systems have been documented.
The residual sensori-motor performance of blindsight and numbsense patients
is fully disrupted when a verbal representation of the stimulus is activated.
This cognitive ? motor interference suggests that action can rely on the
semantic representation (that is missing in these patients)3. Reciprocally,
sensori-motor adaptation to a visual shift has been shown to affect cognitive
representation of space in both neglect patients and normals4. This longer
term sensori-motor ?cognitive interference suggests that higher-level representations
of space can be re-organised by the plastic modification of sensori-motor
correspondence.
The best way to isolate the sensori-motor mode of vision is to apply
time constraints to the response being made by normal subjects. During
a pointing experiment, subjects have been requested to immediately stop
their action in case of a target jump. For speeded movements, they nevertheless
performed numerous (counter-intentional) corrections toward the secondary
target position. This automatic movement control was disrupted in a patient
with a bilateral posterior parietal lesion . The posterior parietal cortex
appears as an automatic pilot able to track a selected spatial target and
drive the hand to it irrespective of the subject’s own intention5. The
most typical function of the dorsal stream appears to be the on-line control
of an ongoing goal-directed action.
It is concluded that, depending on the time scale considered, no interaction,
or two way interactions between the dissociated vision for action and vision
for identification can be observed6. The temporal dimension appears as
one of the keys to the understanding of complex interconnected networks
such as the visual brain.
1. Milner A.D., Goodale, M.A. 1995. The visual brain in action (Oxford
Psychology Series 27). Oxford: Oxford University Press. 248pp.
2. Rossetti Y., Pisella L., Pélisson (2000) Eye blindness and
hand sight: temporal aspects of visuo-motor processing. Visual Cognition,
7, in press.
3. Rossetti Y. (1998) Implicit short-lived motor representation of
space in brain-damaged and healthy subjects. Consciousness and Cognition,
7, 520-558.
4. Rossetti Y., Rode G., Pisella L., et al. (1998) Prism adaptation
to a rightward optical deviation rehabilitates left hemispatial neglect.
Nature, 395, 166-169.
5. Pisella L, Gréa H, Tiliket C, et al. (2000) An automatic
pilot for the hand in the human posterior parietal cortex toward a reinterpretation
of optic ataxia. Nature Neuroscience, 3, 7, 729-736
6. Pisella L., Rossetti Y. (2000) Interaction between conscious identification
and non-conscious sensori-motor processing: temporal constraints. in: Rossetti
& Revonsuo (eds): Beyond dissociation: Interaction between dissociated
implicit and explicit processing. Benjamins Amsterdam. pp. 129-151.
Perception and action in unilateral neglect
Marie-Thérèse Perenin
Cerveau et Vision, INSERM Unité 371, Bron
The multiform aspect of the neglect syndrome has often been ascribed
to the several ways space perception and representation may be altered.
Some authors have stressed the prevalent role of a displacement and/or
distortion of an egocentric frame of reference while other have pointed
out a disturbance of object-centred coordinates or even of world-centred
coordinates. Variations according to the prevalence of either a ‘perceptual’
or a ‘motor’ component have also been described in a number of studies.
However, in most of them, non-natural, incompatible tasks were used, which
prevent a clear interpretation of the results.
In the present study, by recording natural, reaching and exploration
hand movements, we have shown evidences for perception/action dissociations
in both neglect patients and normal subjects with a neglect-like syndrome
following optokinetic or vestibular stimulation. Despite a misperception
of their sagittal midplane, all subjects performed virtually as normals
when they had to point rapidly at visual targets. On the contrary, tactile
exploratory movements and 'indirect’ pointing (after a delay or at a distance
from the visual stimulus, i.e. in a different plane) were both influenced
by the shift of the subjective straight ahead.
These results provide further arguments for the hypothesis of
two separate systems of spatial representations. ‘Motor’ representations
dedicated to action are short-living, unconscious and effector-specific
processes; they require a strict temporal and spatial correspondence between
the goal to be reached and the response directed at it. ‘Perceptual’ representations
are long-lasting and more elaborate processes which do not require such
a stimulus-response correspondence. They can be memorized and give rise
to delayed motor responses, or to responses at a distance from the stimulus,
or to verbal responses concerning the perceived, explicit, localization
of the stimulus. They would also take part in building a supramodal and
conscious representation of space, on which rely deliberate exploratory
behaviours.
The parietal cortex and the representation of action
Angela Sirigu
Institute for Cognitive Science, CNRS, Lyon, France.
For over a century, the posterior parietal regions have been known to be involved in arm and hand movements, from reaching and grasping movements to symbolic gestures and the manual use of tools and objects. Although the nature of this contribution is not fully understood, a recurring hypothesis is that the parietal cortex maintains some aspects of the internal models of the learned limb postures and motor programs necessary for the guidance of skilled hand movements. In support of this hypothesis, I will present data: (1) on different patients with parietal cortex lesions who show selective impairments in mental movement rehearsal, using a variety of tasks which can be used to establish the parallelism, or lack thereof, between imagined and executed movement sequence (Sirigu et al. , Neuroreport 1995, Sirigu et al. Science 1996; (2) recent fMRI results obtained in normal subjects who were asked to either imagine or to execute auditory-cued hand movements. Direct comparison between the two experimental conditions showed that specific cortico-subcortical areas were more engaged in mental simulation, including bilateral premotor, supplementary motor and left posterior parietal areas thus suggesting that specific brain areas are more involved and weigh more importantly on imagined movements rather than executed movements (Gerardin, Sirigu et al., in press, Cerebral Cortex, 2000).
Gamma oscillations
and object representation in humans
C. Tallon-Baudry, O. Bertrand and C. Fischer.
INSERM U 280 Lyon
We experience objects as entities irrespective of whether they are perceived
by our sensory system or recalled from memory. Since parts and properties
of objects are encoded in many distinct brain areas, how does a coherent
percept emerge? Rhythmic synchronization of neural discharges in the gamma
band (around 40 Hz) could provide a spatial and temporal dynamic link between
and within the areas involved in the same network, thus solving the so-called
"binding problem" for object features [1]. We suggest that this mechanism
could be used more generally for the construction of an object representation,
driven either by sensory inputs or by internal top-down processes. If this
hypothesis is correct, gamma-band oscillatory activity should appear in
any task requiring the activation of an object representation.
Significant enhancements of induced activity (e.g. activity appearing
with a jitter in latency from one trial to the next) in the gamma-band
were repeatedly observed in human scalp EEG [2], in different tasks involving
the activation of an object representation (a feature binding task, an
hidden-object detection task, a delayed-matching-to-sample task). In addition,
the different topographies of the induced gamma-band oscillatory activity
suggest that depending on the task to be performed, different functional
areas can be recruited to participate in an oscillatory ensemble. Intra-cranial
EEG recordings in epileptic patients supported this interpretation: they
revealed the existence in the extrastriate cortex of both within- and between-area
oscillatory synchronization. Both local and long-distance synchronizations
were task-dependent, and could account for the results obtained at the
scalp level.
Scalp and intra-cranial recordings in humans thus suggest that the
neural correlate of the activation of an object representation is the synchronization
of a distributed cell assembly in the gamma-band.
1- Singer, W. & Gray, C.M. Annu. Rev. Neurosci. 18, 555-586 (1995).
2- Tallon-Baudry, C. & Bertrand, O. Trends Cogn. Sci. 3, 151-162
(1999)
| From perception to action representation
:
Neural bases and disorders |
Tuesday Abstracts
Fractionating the Supervisory System
Tim Shallice
Institute of Cognitive Neuroscience, UCL.
A theoretical framework for fractionating the Supervisory System within the Norman/Shallice model of executive functions will be derived from the AI work of Fox and Dhas. How adequately the framework relates to the subfunctions of regions of the frontal cortex will be considered from the prespectives of two types of empirical evidence. One type are functional imaging studies of the control processes involved in encoding and retrieval of episodic memories. The second are neuropsychological studies of problem-solving and intention realisation.
Consciousness and the functions of prefrontal cortex
Chris Frith
Wellcome Department of Cognitive Neurology, Institute of Neurology, UCL.
Awareness of selecting one from a number of possible actions and then initiating that action forms an important component of consciousness. This awareness of selection arises especially for novel situations in which it is not immediately obvious which action is most appropriate. There is considerable evidence, from lesion and imaging studies, that dorso-lateral prefrontal cortex (DLPFC) has a major role in response selection in these circumstances. I propose that this selection is achieved by biasing an arbitrary, but appropriate subset of responses (sculpting the response space). If this is the case then the final initiation is achieved by some bottom-signal and not directly by activity in DLPFC.
It is not clear whether awareness is required for this kind of selection
and initiation or whether activity in DLPFC is always associated with awareness.
Electrophysiological studies suggest that awareness of selection is associated
with the precise specification of the response to be made. Since this occurs
rather late in the selection process controlled by the DLPFC, I would conclude
that activity in DLPFC is not sufficient for awareness of response selection.
Imaging studies directly examining awareness of selection implicate medial
frontal structures including anterior cingulate cortex. Awareness of initiation
precedes actual initiation and is therefore likely to be based on predicted
rather than actual sensory consequences of action. The supplementary motor
area (SMA) seems to have a critical role in the initiation of an action.
This is probably achieved by inhibiting action until the appropriate moment.
Thus SMA may have a critical role in the timing of awareness of action
initation.
Context processing and action in schizophrenia
Mohammed Saoud
Hôpital Le Vinatier, Lyon
Patients with schizophrenia present a wide variety of deficits in the realm of cognitive functions. There is a challenge to understand such schizophrenia cognitive deficits in terms of a common unifying hypothesis. A single unifying hypothesis has difficulty explaining such disparate deficits, and, consequently, reference has been made to a variety of underlying models of normal cognitive functioning. One of the most coherent suggestions relates to impairments in context processing (extraction, representation and/or use of contextual information) (Servan-Shreiber et al., 1996). Such impairments were often explored in high levels of cognition, for instance, memory or language. Aside from the presence of various abnormalities, movement disorder is one characteristic of schizophrenia that is often described regardless of the theoretical framework used to define the disease. Context processing may apply to very simple pointing task as well as complex cognitive tasks. We will present several studies with different task difficulties ranging from simple speed-accuracy trade-off for target size to delayed pointing in a spatio-temporal context. Our results show that while the simplest context levels are processed normally in patients with schizophrenia, the extraction and use of more complex contextual information is impaired even in the case of goal directed action. We suggest that analysis of a low level of cognition, such as the natural tendency to adapt pointing movement to the task difficulty, could provide the basis for understanding higher level processing.
Perceptual awareness versus action ?
Y Paulignan, A.C. Roy and M. Jeannerod
Institut des Sciences Cognitives, CNRS UMR 5015, 67 boulevard Pinel,F69675
Bron cedex
In 1991 Castiello et al. presented an experiment showing a dissociation
between a motor response to a perturbation and the corresponding awareness
of the same event. Estimation of the delays was made by a double task paradigm
: a motor and a vocal responses to the same stimuli. The results showed
a large difference between latencies of motor and vocal (conscious) responses.
The possible interactions between these tasks was a matter of debate.
The aim of the present study is to replicate this experiment with only
one task.
Three objects were placed on a table at -10, 0 and 10 degrees of subject's
sagittal axis. These objects were translucent dowels 10 cm height, 1.5
cm in diameter. In 80% of the trials the illumination of the central dowel
was the signal for the subject to grasp it. On the remaining 20% of trials,
the illumination was moved from the central object to one of the others
at the beginning of hand movement. Movements were recorded by an OPTOTRAK
system. After each perturbed trial the subjects were asked to show where
the hand was when they detected the perturbation. This hand location was
recorded. We computed the minimum of distance between the reported position
and the wrist trajectory during the preceding perturbed trial. Then, we
obtained a point on the wrist movement trajectory of the perturbed trial.
This point determines the moment at which the subjects became aware of
the perturbation.
Subject's awareness initiation occurs 410 ms after the perturbation.
The motor response occurs later than for the previous studies of Castiello
et al. (1991) and Paulignan et al. (1991).
The results confirm the Castiello's estimations of time of consciousness
by giving the same values. They demonstrate that having to recall accurately
the previous motor acts, hence being aware of them, has an influence on
the motor performance.
Castiello, U., Paulignan, Y., & Jeannerod, M. (1991) Temporal dissociation of motor responses and subjective awareness : a study in normal subjects. Brain, 114, 2639-2655
Paulignan Y, MacKenzie CL, Marteniuk RG, and Jeannerod M (1991) Selective perturbation of visual input during prehension movements. 1. The effects of changing object position. Exp Brain Res 83: 502-512
Computational human sensorimotor control
Daniel Wolpert
Sobell Department of Neurophysiology, Institute of Neurology, UCL, UK.
The talk will focus on the computations underlying planning, prediction
and learning in sensorimotor control. First, I will present our work
on motor planning which provides a unifying theory of eye and arm movement
control. This theory suggest that trajectories are selected so as to minimize
the consequences of signal-dependent noise. I will then present work on
predictive internal models, neural circuits that predict the consequences
of motor commands, focusing on their uses, how they are learned and their
neural basis. Finally, I will describe work on modular leaning in
which we examine how humans learn to generate accurate and appropriate
motor behaviour under many different and often uncertain environmental
conditions. This work focuses on a novel model, the multiple paired
predictor-controller model, to deal with such selection and learning.
The cognitive and neuroanatomical correlates of multitasking
Paul W. Burgess
Institute of Cognitive Neuroscience, UCL, London
One of most curious groups of patients are those who are normal or even
supra-normal on traditional neuropsychological tests, including those sensitive
to frontal lobe damage, but who show deficits in planning, organisation
and prospective memory in everyday life. These patients show a circumscribed
deficit in situations that involve multitasking: i.e. situations that involve
dovetailing different activities when returns to task are not directly
signalled. This suggests that there are cognitive resources that are dedicated
to this function.
This view receives support from a recent study of the Six Element Test
(SET; the most commonly used clinical test of multitasking originally described
by Shallice and Burgess, 1991). The carers of 92 mixed aetiology neurological
patients were asked to rate the patients using a questionnaire measuring
20 of the commonest symptoms of the dysexecutive syndrome. factor analysis
of these ratings revealed 5 orthogonal factors, with factor 2 related specifically
to planning and organisation in everyday life. When correlations between
the factor scores and the patients' performances on a range of clinical
tests were examined, the scores for factor 2 showed only significant correlations
with the SET, and not with WAIS-R IQ or a range of memory, language and
other tests.
Further support is found in Burgess et al (2000). This study involved
a componential analysis of a complex multitasking test using 60 patients
with circumscribed cerebral lesions and 60 age- and IQ-matched controls.
The tests was given in such a way that the relative contributions of rule
learning, planning, plan following, autobiographical and retrospective
memory to overall task performance could be assessed. Analysis of relations
between lesion sites and test performance showed that damage to left medial
posterior regions, including the posterior cingulate and occipital lobe,
caused deficits on all aspects of the procedure except planning. Circumscribed
impairments were however found in other lesion groups. Lesions to the right
dorsolateral prefrontal cortex caused isolated planning impairment. The
left anterior cingulate and surrounding regions were damaged in people
who showed problems in remembering the task rules after a delay; and poor
performance on the multitasking component of the test (measured as task
switches minus rule-breaks) were associated with lesions to the left frontal
pole.
Analysis of the relations between the different performance stages using
structural equation modeling suggested that three primary are involved
in multitasking. The first is retrospective memory, which is involved in
rule learning and remembering. The second is required for planning, and
the third is involved in the prospective memory demands of the task, such
as remembering to switch tasks, and to follow one's plan.
Most recently, this methodology has been applied to the original version
of the SET, with similar results, highlighting the role of anterior polar
and superior medial regions in task switching after filled delays. Furthermore,
two recent PET studies of this function have shown rCBF changes in this
region during the maintenance of a delayed intention. The results could
not be explained as increases in attentional demands. These results of
this series of studies converge on the conclusion that there is a circumscribed
set of cognitive processes that support human multitasking, and that the
most likely supporting structures for the most dedicated processes are
located in the medial and polar prefrontal regions.
| From perception to action representation : Neural bases and disorders |
Prism adaptation for
different starting arm postures
Pierre Baraduc and Daniel Wolpert
Sobell Dpt of Neurophysiology, Institute of Neurology, London
Prism adaptation is generally viewed as the learning of a congruence
between two perceptual dimensions (Harris 1965; Bedford 1993). In accordance
to this view, some authors found that the adaptation transfers from one
type of movement to another (Yachzel and Lackner 1977; Freedman et
al. 1965). This view of adaptation as an alignment problem has been more
recently challenged with the discovery that the magnitude of prism adaptation
depends on the kinematics of the pointing movement (Kitazawa
1997).
These aftereffects had been formerly described as a sum of three components:
(1) a visual shift; (2) a change in felt limb position; (3) an “assimilated
corrective response”, that depends on the experimental context (Welch 1974).
This third component has however not been properly characterised, and could
correspond to the learning of a new sensorimotor coordination. It would
therefore logically depend on all the parameters of the movement. This
idea was checked by manipulating the initial posture of the arm.
Methods
Subjects were required to point with their right hand from a fixed
starting point to a static target. The orientation of the arm at the beginning
of the movement was restricted to a set of three postures that differed
by the degree of adduction (see figure; for clarity, only 2 postures have
been represented). During the pre-test period, movements with vision were
interleaved with movements without vision. The visual feedback was only
given when the starting position was the most adducted (pos. A). It was
limited to the representation of the finger endpoint. A horizontal
visuomotor discrepancy was then progressively introduced, that reached
a maximum of 10 cm. This discrepancy varied as a function of arm
extension, as a prism shift. During this exposure phase, the starting position
was always A. Finally, a post-test phase identical to the pre-test assessed
the amount of adaptation.
Results and discussion
These are preliminary results. Maximum adaptation is achieved for the
starting position used during exposure. The magnitude of the aftereffects
decreases sharply as a function of the initial abduction, from 80% adaptation
for pos. A to 40% adaptation for pos. C. This suggests that a major component
of the rapid visuomotor remapping is the acquisition of a new sensorimotor
coordination. In accordance to this, a modelling study showed that two
distinct learning mechanisms could be necessary to explain the whole range
of aftereffects (Guigon, personal communication): (1) a corrective supervised
learning, analogous to the correction based on visual feedback alone; (2)
a slower Hebbian mechanism, analogous to the realignment between proprioception
and vision.
As the end posture varies as a function of the starting posture, it
cannot however be excluded that these results are due to a difference in
final proprioception. The variability of final arm posture does not lend
a strong support to this hypothesis. Nevertheless, a control experiment
will have to also constrain the final orientation of the arm.
References
Bedford F (1993) in The psychology of learning and motivation pp 1-60.
Freedman SJ, Hall SB, Rekosh JH (1965) Percept Motor Skills 20:1054-1056.
Harris CS (1965) Psychol Rev 72:419-444.
Yachzel B and Lackner J (1977) Percept Psychophys 22:147-151.
Welch RB (1974) J Exp Psychol 103:700-705.
Predicting the sensory consequences of action
S-J. Blakemore, D.M. Wolpert & C.D. Frith
Wellcome Department of Cognitive Neurology, University College London
We have investigated how the brain distinguishes between self-produced and externally produced sensations. 'Forward models' predict the sensory feedback from self-produced movements thereby enabling us to recognise the sensory consequences of our own actions. It is proposed that an impairment of this predictive mechanism might give rise to certain symptoms experienced in schizophrenia (1,2). If self-produced sensations are interpreted as being generated by an external source, then thoughts might be interpreted as external voices (auditory hallucinations) and self-produced movements might be interpreted as externally generated (delusions of control or passivity phenomena). Psychophysical studies have shown that self-produced tactile stimulation is rated as less intense and 'tickly' than externally produced stimulation, and we have proposed that such an attenuation of self-produced tactile stimulation is due to the sensory predictions made by a forward model. Functional neuroimaging studies have demonstrated that such attenuation is mediated by somatosensory cortex and the anterior cingulate cortex: these areas are activated less by a self-produced tactile stimulus than by the same stimulus when it is externally produced. Furthermore, evidence suggests that the cerebellum might be involved in generating the prediction of the sensory consequences of movement (3). Finally, we have recently demonstrated that psychotic patients with auditory hallucinations and passivity phenomena show no attenuation of self-produced sensory stimulation (4). This supports the proposal that these symptoms are associated with an impairment of the functioning of the forward model.
1. Frith, CD. (1992). The Cognitive Neuropsychology of Schizophrenia.
Lawrence Erlbaum Associates, UK.
2. Frith, CD, Blakemore S-J & Wolpert, DM. Abnormalities of the
Perception and Control of Action. In Press in Philosophical Transaction
of the Royal Society of London: Biological Sciences
3. Blakemore, S-J, Wolpert, DM & Frith, CD (1998). Central cancellation
of self-produced tickle sensation. Nature Neuroscience 1(7), 635
- 640
4. Blakemore, S-J, Smith, J, Steel, R, Johnstone, E & Frith, CD.
The perception of self-produced sensory stimuli in patients with auditory
hallucinations and passivity experiences: Evidence for a breakdown in self-monitoring.
In press in Psychological Medicine.
Modelling the Somatoform
and Dissociative (Conversion) Disorders
Richard J. Brown, PhD.
Institute of Neurology, UCL.
Traditional models have emphasized the role of motivational factors in the pathogenesis of the Somatoform and Dissociative (conversion) disorders. However, such factors are absent in many instances of somatoform and dissociative illness. Based on an integrative model of cognition, somatoform and dissociative symptoms are reconceived here as distortions in subjective experience arising from the chronic activation of symptom representations during the generation of consciousness. By this view, somatoform and dissociative illness are essentially cognitive phenomena rather than the product of emotional trauma or the expression of unconscious conflict as is traditionally conceived. Several factors that moderate the occurrence of these conditions via their effect on the activation of symptom representations are identified, including symptom monitoring, the activation of complementary representations, high-level rumination and low-level attentional sensitivity.
Two systems of spatial
representations in human posterior parietal cortex
S. Clavagnier*, I Schindler, Y. Paulignan, A. Vighetto, MT. Perenin
INSERM U 371, ISC, 69500 Bron, Hôpital Neurologique 69003 Lyon,
France
Although often associated, spatial disorders from parietal origin may
occur in isolation. This is the case for optic ataxia and unilateral neglect,
which result from lesions respectively centered on the dorsal and the ventral
part of posterior parietal cortex. This neurological dissociation as well
as psychophysical findings has led to the hypothesis of two separate systems
of spatial representations. "Motor" short-lived are dedicated to action
and subtend goal-directed movements. "Perceptual", long-lasting representations
are dedicated to space perception; they can be memorized and used in motor
responses at a distance from the stimulus or on exploratory behavior. These
two kinds of representations would be subserved by the superior and inferior
parietal lobules respectively.
Optic ataxia patients would thus be improved when required to perform
either delayed or displaced responses, while neglect patients would perform
worse. In the present study, visuo-motor performances of two patients,
one with optic ataxia (IG) and another one with unilateral neglect (FP),
were compared to that of normal control subjects. Subjects were required
to point at visual targets presented randomly in their peripheral field.
Pointing responses of the right hand were recorded by using an optoelectronical
motion analyzer (OPTOTRAK?) in three conditions: "direct" pointing, "indirect"
pointing with either a delay of five seconds between stimulus offset and
movement onset, or a spatial separation between stimulus and responses
planes. While normal subjects and FP pointing movements became less accurate
in the “indirect” pointing conditions, disturbances of IG performances
decreased significantly in these conditions. These results indicate that
optic ataxia patients still have access to a cognitive mode of spatial
representations, the one that is impaired in neglect patients. This new
dissociation confirms the idea of a functional subdivision in posterior
parietal cortex.
Can you remember the same way twice? Changes in the structure of autobiographical episodes recalled repeatedly at a short interval.
Laure Coates & Paul Burgess
Institute of Cognitive Neuroscience, UCL (University College London)
Using verbal protocol analysis of healthy participants' recollections, Burgess and Shallice (1996) proposed a model of the processes involved in remembering events that have happened to you. They found that in addition to memories themselves, participants produced many protocol elements that were evidence for the operation of cognitive control processes in autobiographical recollection. The model proposes three broad categories of control process. Descriptor processes are assumed to produce a specification of the type of trace that would satisfy the demands of the retrieval task. Memory Editor processes are involved in checking that the output fits with previously retrieved memory elements in the episode being retrieved and also with the overall task requirement. Mediator processes control strategic and problem-solving operations concerning the adequacy or plausibility of retrieved memory elements.
If the operation of these control processes corresponds to the process of recollection rather than the products of it, then subsequent recollection of them should be weaker than for the memories that were recalled. We therefore investigated the proportional change in control elements when people are asked to recall the same autobiographical episode twice. 16 participants were asked 12 questions about everyday events that happened to them, such as "Describe the last time you had dealings with the police". Immediately after they finished answering all 12 questions, they were asked half of these questions again, and had to repeat exactly what they said the first time they answered them.
The results showed a significant proportional decrease in all control
elements (Editor, Mediator and Descriptor) in the second recollection episode,
with a corresponding proportional increase in the recollection of memories.
This pattern even remained when subjects were asked not merely to recall
the episode again, but also when they were asked to recall the exact words
they had used in giving their first answer. This pattern also appeared
in recollection of other people’s verbal protocols. The results support
the hypothesis that cognitive control processing in recollection has a
different status from the activation of memory representation.
Is the Bell Inequality Violated in Visual Perception?
Andrew Duggins, Geraint Rees, Chris Frith
Wellcome Department of Cognitive Neurology, University College London
The perceptual latency of a direction change in an array of oscillating squares was shown by Moutoussis and Zeki to be significantly greater than the perceptual latency of a colour change . This has led to the suggestion that there is a ‘microconsciousness’ generated at each of several anatomically separate and functionally specialised nodes . Furthermore, anatomical evidence suggests that there is relative segregation of inputs to the different nodes. From these theories follows a further experimental prediction, that the probability of perceptual pairing of colour and motion oscillations must not violate the Bell Inequality : We will review a standard derivation of the Inequality, and explain how it can be applied to the Moutoussis/Zeki paradigm. Surpisingly, the results of the original experiment suggest that the Inequality is violated, and thereby cast doubt on the ‘microconsciousness’ theory. Alternatively, it is possible that the critical assumption of perceptual pairing, which would allow application of the Bell Inequality in this context, is not valid. In a new experiment, we replicate the Moutoussis/Zeki paradigm in a way that allows these two alternative interpretations to be resolved, and which allows statistical test of the ‘microconsciousness’ theory. Even isolating those trials in which perceptual pairing is confirmed by subject report, we find that the Inequality is violated. We conclude that the activity of cells in V5 must indeed depend partly on colour input and the activity of cells in V4 partly on motion. Visual consciousness must then be considered non-local and inseparable: the microconsciousness does not exist.
Agency judgement in schizophrenic patients.
C. Farrer , N. Franck , N.Georgieff , M. Marie-Cardine, J. Daléry, T. d ’Amato, & M. Jeannerod .
One of the central questions about the problem of self-consciousness
and other-consciousness is to understand the experience of agency or how
a subject attributes an action to himself or to an other agent. One approach
for understanding the underlying mechanisms is to study disturbances of
self-consciousness in schizophrenic patients. Indeed, one class of symptoms
displayed by these patients, the so-called positive symptoms (e.g: delusion
of control, hallucinations, ...) are suggestive of an alteration
of the awareness of one’s own action and of the recognition of actions
performed by others.
One hypothesis postulates that an attribution of action relies on a
discrimination between central representations activated from within and
those activated by external cues (Georgieff and Jeannerod, 1998). A first
experiment by Daprati et al (1997) has tempted to study attribution
of action in a group of schizophrenic patients. The results have shown
that schizophrenic patients with hallucinations and delusion of control
tended to overattribute to themselves actions produced by others and in
conditions where the cues for discriminating the origin of an action were
degraded.
In order to analyse more precisely the influence of the perceptive
information on this agency judgement we realised another experiment
where the parameters of the visual information were controlled using
an electronic device. Subjects hold a joystick with their right hand. The
movements of the joystick were fed into a virtual hand holding a
joystick. The virtual image was projected on a mirror overlying the subject’s
hand. This device allowed us modifying the apparent direction and/or velocity
of the movement actually performed by the subject. A manual motor task
and an agency judgement about this task were performed by 29 normal subjects
and 29 schizophrenic subjects
The results obtained in schizophrenic patients showed a clear
deficit for all the patients when their movements were delayed up to 250-280ms.
Only the influenced patients were severely impaired for angular biases
up to 30°. Those results could not be explained by deficits such as
perceptual or attentional mechanisms impairment, indeed all the patients
performed well at the BORB which tests those abilities. An impairment in
the detection of visual cues indicating the intention of the movement
(as it has been shown in our group of influenced schizophrenic patients)
can lead to a misunderstanding of the action of an agent. So
a deficit at this level could explain the fact that influenced patients
tended to self-attribute movements they observed. This result is reinterpreted
in the theoretical framework of Georgieff and Jeannerod.
Daprati E, Franck N, Georgieff N, Proust J, Pacherie E, Dalery J, Jeannerod
M, (1997) : Looking for the agent: an investigation into consciousness
of action and self-consciousness in schizophrenic patients. Cognition;
65 : 71-86.
Georgieff N. and Jeannerod M, (1998). Beyond consciousness of external
reality: a “who” system for consciousness of action and self-consciousness.
Consciousness and Cognition; 7: 465-477
In search of a monitoring deficit of action in schizophrenia : some arguments for a revision of this assumption
Fourneret P 1., Franck N 1.2., Georgieff N 1.2., Jeannerod M 1
1 : Institut des Sciences Cognitives. 67 Bd Pinel – 69675 Bron
2 : CHS Le Vinatier. 95 Bd Pinel – 69677 Bron
Abstract
The first-rank symptoms (Schneider, 1959), among which delusions of
control, thought insertion, verbal hallucinations, are certainly the most
impressive clinical signs met in schizophrenia and the most detrimentalto
the social functioning of the patients. According to Frith (2000), these
symptoms result from a lack of awareness of certain aspects of motor control
; this deficit is, itself, in relation with a failure in the cognitive
mechanism by which the predicted consequences of an action can be derived
from a forward model based on the intended sequence of motor commands.
This assumption of an impairement in the central monitoring of their own
actions by the schizophrenic patients, is derived from several experimental
studies, which highlighted a greater difficulty for the patients, compared
with controls, of being able to correct erroneous movements in the absence
of visual feedback (Malenka, 1982 ; Frith & Done,1989 ; Mlakar et al,
1994).
In our study, 19 schizophrenic patients (10 with Schneiderian symptoms
and 9 without Schneiderian symptoms) versus 19 paired control subjects
were subjected to a sensorimotor adjustment task. Using a magnetic stylet,
they had to connect two visual targets with a trajectory as straight as
possible. The subjects could not see their hand during the test and could
not see the result of their action before the last third of the distance
separating the
two targets. In the experimental condition, the computer generated
a conflict between the action planned and the sensory-motor feedback by
a linear directional bias (15°) to the right. Thus, to succeed the
task, subjects had to modify their motor program and deviate their hand
by 15° in the opposite direction. All subjects were naive at the beginning
of the test and performed twenty trials. Sensory-motor adjustment to the
bias was evaluated by the surface between the line traced and the ideal
line to compensate for the deviation (square medium ratio). Awareness of
the conflict was appreciated by the explicit report of a left manual correction.
Our results show that only the patients aware of the manual correction
(n= 9) succeeded the task with a performance not significantly different
from the control subjects. Nevertheless and surprisingly, most of them
presented Schneiderian symptoms. For the others (mainly patients without
Schneiderian signs), although the degree of adjustment was significantly
lower than in controls, it showed evidence for a broadly preserved central
monitoring of action. Only the ability to become aware of the motor intention
appeared to be defective in these subjects.
These results do not support the assumption of a deficit in central
monitoring of action in schizophrenic patients, especially those suffering
from Schneiderian symptoms. Furthermore, they reinforces the idea, defended
by some authors, of a functional dissociation between the consciousness
of action and the executive control of action.
The coding of movements in primary motor cortex: a TMS study
A.F. Hamilton, K.E. Jones, Z. Ghahramani*, R.N. Lemon & D.M. Wolpert
Sobell Dept of Neurophysiology, Institute of Neurology
*Gatsby Computational Neuroscience Unit.
Whether movements are represented in Cartesian, joint, muscle or in
some other co-ordinate system in primary motor cortex is still unclear.
Almost all evidence comes from studies carried out in non-human primates.
To examine movement coding non-invasively in humans we have used Transcranial
Magnetic Stimulation (TMS) over the arm area of motor cortex to produces
a force pulse generated by the muscles in the arm. While the coil
remained stimulating the same area of M1, the arm was moved over a range
of joint angles covering the workspace. We recorded the force vectors at
the wrist and the EMG in four muscles of the upper arm produced by the
TMS pulse. The array of force vectors and EMG over the workspace showed
a systematic pattern of change. We will present preliminary results and
discuss their implications for the coding debate.
Visuomotor adjustments: differences between motor performance and awareness of action
Helen Johnson and Patrick Haggard
Institute of Cognitive Neuroscience, UCL.
Previous studies using the double-step pointing paradigm have suggested that visuomotor adjustments to a target shift can be dissociated from conscious perceptual awareness of the target shift. This study investigated the action awareness subjects have of their own visuomotor adjustments, by asking subjects to repeat in their own time the movement paths made in immediately preceding double-step pointing trials.
Subjects were asked to reach out and touch a central illuminated LED
as quickly and as accurately as possible. On 33% of trials the LED
‘jumped’ 10 cm to left or right 25 ms after movement onset. Subjects
were instructed before each block to respond to the target jump in one
of four ways, according to condition:
1) Point: if the LED jumped the subject had to follow it
2) Antipoint: if the LED jumped the subject had to point to the equivalent
opposite location
3) Overpoint: if the LED jumped, the subject had to point to a location
twice the distance of the jump in the same direction
4) Overantipoint: if the LED jumped, the subject had to point to a
location twice the distance in the opposite direction
In all conditions subjects simply pointed to the central LED if it
did not jump. Immediately after completing each trial subjects repeated
the path of their original movement as accurately as possible. The
LED remained at its final location throughout this repetition period.
Visuomotor adjustments were more rapid in the basic point condition than in the other conditions. We then compared the original movements with the repetitions to measure the extent to which subjects were aware of the path of their original movement, and in particular, their awareness of any visuomotor adjustments. In the basic point condition subjects believed that they began to adjust later and to a lesser extent than they actually did. In the antipoint condition, in contrast, they overestimated the extent of their adjustment. That is, they thought they deviated further in the intended direction, opposite to that of the LED, than they actually did. In the overpoint condition subjects initially underestimated the extent of their adjustment, as in the basic point condition, and then, later in the movement, overestimated it, as in the antipoint condition. In the overantipoint condition subjects overestimated only the end of the deviation.
These findings provide evidence that visuomotor adjustments may be dissociated from awareness of action. When the visuomotor adjustment directly corresponds to the perceptual target shift, as in the basic point condition, visuomotor adjustment is mediated by an automated response system of which subjects are at least partly unaware. Thus, subjects believe that they performed basic double-step adjustments slower and less accurately than they actually do. In the other conditions the required adjustment requires some indirect transformation of the perceptual target shift. These adjustments are mediated by a slower, controlled response system that not only enters awareness but is also slightly exaggerated, leaving subjects believing that they performed the task better than they actually did. Interestingly, the overpoint condition may reflect an amalgamation of the two systems, in which an automated initial response to follow the LED is subsequently modified.
Acknowledgement: Helen Johnson is supported by a BBSRC Special Committee
Studentship. Additional funding was provided by the Royal Society.
NEURONAL ACTIVITY RELATED TO BIMANUAL COORDINATION IN THE MOTOR AND PREMOTOR CORTICES OF RHESUS MONKEYS
C. Jouffrais, E.M. Rouiller.
Institute of Physiology. University of Fribourg
As movements of the two arms are naturally coupled, there must be a
neural mechanism which can possibly influence this bimanual coupling. To
investigate the role of motor cortical areas in the control of bimanual
spatial and temporal coordination, we recorded single unit activity in
the primary motor and premotor cortices of a monkey trained to execute
a conditional reach and grasp drawer task. Four different delayed movements
-unimanual right, bimanual right-left (reaching with the right hand, grasping
with the left hand), unimanual left or bimanual left-right (reaching with
the left hand, grasping with the right hand) - were instructed depending
on the color of a visual cue. The gaze position of the monkey was controlled
using the scleral search coil technique. This design allowed us to examine
the signal-, set- and movement-related activity of individual cells involved
in specific unimanual or bimanual movements. We classified the cells in
three different categories possibly contributing to the coding of bimanual
movements: (1) the “combination neurons” exhibited a spike rate modulation
during bimanual movements which was the combination of the discharges observed
during unimanual movements; (2) the “invariant neurons” showed signal-
and/or set-related activity which was the same whatever the task the monkey
was performing; (3) the “interaction neurons” exhibited discharge rate
modulations which specifically reflected the interaction of the two arms
during at least one of bimanual tasks. These preliminary results suggest
a possible role for premotor and motor cortices in bimanual coordination.
Non-informative Vision
Enhances Tactile Acuity
Steffan Kennett, Marisa Taylor-Clarke & Patrick Haggard
Institute of Cognitive Neuroscience, UCL (University College London),
UK.
Previous work has revealed interdependence between the visual and tactile systems. For example, vision and touch appear to be intimately linked during both reflexive and voluntary shifts of spatial attention [1, 2]. Furthermore, single-unit recording has shown multimodal cells with spatially congruent receptive fields in both modalities [3]. Sight of a hand has been shown to reduce detection times to tactile targets presented to it [4]. However, no previous work has investigated whether vision affects the informational content of tactile signals, and in particular the spatial resolution of the tactile map of the body surface. Such evidence would add weight to suggestions that crossmodal links may arise due to feedback from populations of multimodal neurons to primary sensory cortex [e.g., 5]. Such feedback may modulate receptive field size or even lead to short-term cortical plasticity. Previous behavioural studies cannot distinguish between crossmodal modulations in the salience of a tactile stimulus (which could arise at any of several stages in processing pathway) and modulations in informational content (acuity), which are likely to arise at the level of the primary cortical projection. In this study, we directly measured effects of vision on tactile acuity by using a staircase procedure to estimate tactile two-point discrimination thresholds (2PDTs) on the arm. Participants always gazed towards their arm throughout the tactile stimulation session but, via use of a half-silvered mirror, the visual scene they viewed was varied across four conditions. In the first condition, participants gazed at their arm in total darkness (DARK condition). In the second condition, participants had clear vision of the stimulated part of their arm, except during a 700ms interval when the tactile stimulators were moving to strike the arm (VISION OF ARM condition). In the third condition, participants saw a cylindrical object at the location of the stimulated part of the arm (VISION OF OBJECT). In the fourth condition, participants viewed their arm through a 2.5? magnifying glass (MAGNIFIED VISION OF ARM). The dark interval around the time of tactile stimulation in conditions 2 and 4 ensured that vision could not be directly informative for tactile discrimination judgements.
Tactile acuity was improved in the VISION OF ARM condition relative to both DARK and VISION OF OBJECT conditions. The latter difference rules out spatial attention to the stimulated location as the sole explanation of the improved performance. Interestingly, acuity was further improved in the MAGNIFIED VISION OF ARM condition relative to the VISION OF ARM condition. This provides the first direct evidence that crossmodal links modulate somatosensory spatial representations.
Figure: Schematic view of the experimental set-up in VISION OF ARM condition. Participants placed their right arm inside a box that had a semi-silvered mirror in one wall (pale shading). When lights inside the box were illuminated the arm was clearly visible, but the retracted tactile stimulators (tactors) were hidden by the opaque walls (dark shading). Lights were extinguished during tactile stimulus delivery (either one or two stimuli) to preclude any informative vision. (The tactors are shown at the point of contact with the arm; they would not have been visible at this position during the experiment.) Participants wore a patch over their right eye.
1. Spence, C., Nicholls, M.E.R., Gillespie, N. & Driver, J. (1998).
Perception & Psychophysics, 60, 544-557.
2. Spence, C., Pavani, F. & Driver, J. (in press). Journal of Experimental
Psychology: Human Perception and Performance.
3. Graziano, M.S.A. & Gross, C.G. (1993). Experimental Brain Research,
97, 96-109.
4. Tipper, S.P., Lloyd, D., Shorland, B., Dancer, C., Howard, L.A.
& McGlone, F. (1998). Neuroreport, 9, 1741-1744.
5. Hahnloser, R., Douglas, R.J., Mahowald, M. & Hepp, K. (1999).
Nature Neuroscience, 2, 746-752.
Vision and touch through the looking glass: cognitive remapping of multisensory information in the normal and damaged brain.
Angelo Maravita1, Charles Spence 2, Clair Sergeant 1, Karen Clarke 1, Masud Husain, & Jon Driver 1
1 -Institute of Cognitive Neuroscience, University College London
2 -Department of Experimental Psychology, University of Oxford
3 - Imperial College School of Medicine, Charing Cross Hospital, London
Integration of visual and tactile stimuli is particularly effective
when visual stimuli are delivered near to tactually stimulated body parts.
Studies on monkeys show that response of cross-modal visual-tactile neurons
in several regions of the cortex is stronger for visual stimuli presented
in the vicinity of the hands. This makes functional sense, as related visual
and tactile information will normally come from similar locations, near
the skin surface. However, there is one common situation in which stimuli
on our body produce the visual appearance of being placed far away. This
happens when we observe ourselves in mirrors, while acting on our own body
(shaving, making-up, etc.). Tactile-proprioceptive cues then signal stimuli
on the body, while visual information suggests that related visual stimuli
are placed at a distance (“through the looking glass”).
We have shown in both normal subjects, and in a right-hemisphere patient
with crossmodal extinction, that visual information delivered close
to the hand, but observed only via its mirror reflection in far space,
is treated by the brain as a stimulus near the body in peripersonal space.
Normal subjects were tested in a visual-tactile interference paradigm.
Crossmodal interference with tactile judgements is usually stronger for
visual distractors close to the tactually stimulated hands, than for visual
distractors in far space. In one of our conditions (Mirror) visual distractors
were delivered from lights near the hands, but the lights and hands could
only be observed as reflections in a mirror placed in front of the subjects.
In another condition (Box), the half-silvered mirror was illuminated so
as to act like a window into the box on which it was mounted. Subjects
now observed the contents of the box, where visual distractors and two
stuffed rubber hands were placed far away from the subject, at the exact
positions consistent with the reflections of visual distractors and subject’s
hands in the Mirror condition. Interference from visual distractors on
tactile judgements was stronger in the Mirror than the Box condition. We
therefore suggest that visual afference from the mirror, although optically
appearing as far away (“through the looking glass”) is computed by the
normal brain as being near to the corresponding body part. A similar finding
was observed in a neurological case for whom right visual events could
extinguish awareness of tactile events on the left hand, especially if
the visual event was close to the right hand. Such crossmodal extinction
was stronger for a distant right visual stimulus in the Mirror condition,
than in the Box condition.
These results suggest that mirror reflections projecting the image
of distant visual events, can be recoded as arising from peripersonal rather
than extrapersonal space.
Predicting the future position of a looming target in 3D space: the role of retino-, head- and trunk-centered reference frames in man.
Marco Neppi-Mòdona*°, David Auclair,° Jean-René
Duhamel°, Angela Sirigu°
* Facoltà di Psicologia, Università di Torino, Italia;
°Institut des Sciences Cognitives, Bron, France.
Avoiding or intercepting looming objects implies not only a precise
estimate of time to contact but also of impact location. We present an
experiment designed to investigate the relative contribution of retino-,
head-, and trunk-centered reference frames in a trajectory prediction task.
Normal subjects have to predict, in complete darkness, the impact location
on their face (left or right of their head midline) of a LED fixed to a
robotic arm moving in 3D space with a looming trajectory originating from
a leftward, straight-ahead or rightward spatial location. Manipulating
subjects' gaze and trunk orientation, produced important and systematic
errors in the prediction of impact locations, resulting in a strongly biased
estimation of subjective impact in a direction ipsilateral to the hemispatial
origin of the trajectory. We demonstrate that predicting the future position
of a moving target 1) implies access to a multiple reference frames system
and 2) the spatial alignment of the reference frames among them and with
respect to the origin of the trajectory is critical for a correct prediction.
Furthermore, normal subjects' performance was not homogeneous, some individuals
relying mostly on retinal cues, others on head-centered cues and yet others
on mixed spatial cues. These results are consistent with current concepts
of space representation in the parietal cortex, which stress the simultaneous
coding of visual information in multiple reference frames.
Action generation in schizophrenia:
two experimental studies
A. Posada, N. Franck, N. Georgieff & M. Jeannerod.
Institute for Cognitive Science - CNRS, 67 Bd Pinel, 69675 Bron, France.
We can define a ‘mind level’ where words, or symbols, acquire mean,
are consciously managed, can be communicated to others through a language
and can generated voluntary actions. We call this level ‘explicit‘ by its
strong relation with language. Our intentions, which will guide our actions
to their completion are always generated at this level and one of their
characteristics is they are ahead of execution. We used this characteristic
to design the first experiment and evaluate the ability of schizophrenics
to generate precise actions based on explicit information. In this
task subjects explicitly learned a repetitive color sequence and then received
the instruction to give an anticipatory motor response by pushing corresponding
color keys before the next element in the sequence was displayed on a computer
screen. Different types of sequences (temporal and spatial) and experimental
conditions were tested in both a group of normal subjects and a group of
schizophrenic patients. Schizophrenia is a condition known to alter conscious
executive function. Our results showed a strong deficit in performing the
anticipation task in schizophrenic patients. Although they were found to
be able to acquire a conscious knowledge of the sequences almost normally,
their anticipation ability was reduced by comparison to normal subjects
in all the tested conditions. After this positive result, we realized a
second experiment to evaluate more precisely the deficit of schizophrenics
to perform these particular actions. In this task, we measured reaction
time in three different conditions: simple reaction to a color displayed
on screen, categorization of the color displayed on screen (push corresponding
color button) and implementation of an explicit rule to modify the color-button
relation of the second condition. As control subjects did, schizophrenic
patients showed similar pattern response in the first and the second condition;
the reaction times of patients were a few slower. But in the third condition
when subjects used the explicit rule, the reaction times of patients grew
considerably. And, unlike the controls which showed a progressive diminution
of reaction times through time (reflecting an automation of rule), schizophrenics
showed an enhancement of the reactions times (reflecting fatigue and the
absence of any automation). These results are interpreted as a deficit
of schizophrenic patients in consciously using the explicit information
to consequently generate actions and create motor programs. They
also expand the notion of a working memory deficit in schizophrenia and
bear strong implications for understanding executive disorders observed
in such patients.
Reaching and grasping in
the monkey: effect of perturbation of object size and location.
A.C. ROY, Y.PAULIGNAN, D.BOUSSAOUD.
Institut des Sciences Cognitives, BRON, FRANCE
Human prehension movements are classically described as consisting of
two components: a reaching component guided by spatial location of objects,
and a grasping one guided by their size and shape. To understand the anatomical
and physiological correlates of these visuomotor channels, we sought to
describe the kinematics of reaching and grasping in macaque monkey trained
to reach for and grasp 3-D objects. In this study, we tested the degree
of independence of reaching and grasping by perturbing either object location
or its size. Hand path and grip aperture were measured in three dimensions
using an optotrak system.
Two animals are included in this study, and the analysis was based
on 8 000 movements. Both monkeys were tested on perturbation of object
location and size. Perturbation of object location affected both reaching
and grasping components. In particular, grip aperture displayed a double
peak pattern in 20% to 80% of trials, depending on the direction of perturbation.
By contrast, perturbation of object size lead to a single peak of grip
aperture, corresponding to the last object size. This observation indicates
that the new object size has been taken into account earlier. As it has
been reported in human subjects, size perturbation induces a smooth reorganization
of movement and "recoordination" of reaching and grasping. These
findings suggest that reaching and grasping are not strictly independent
visuo-motor channels. This view contrasts with observations of segregated
anatomical pathways linking parietal cortex and premotor areas.
Convergent and divergent effects of neck proprioceptive and visual motion stimulation on ego- and allocentric space in neglect
Igor Schindler1, Georg Kerkhoff²
1INSERM U 371"Cerveau et Vision", 18 avenue Doyen Lépine, F-69675
Bron, France
²EKN - Clinical Neuropsychology Research Group, Dept. of Neuropsychology,
Bogenhausen Hospital, Dachauerstr. 164, D-80992 München, Germany
Spatial neglect can affect egocentric as well as allocentric-based reference
frames. The egocentric aspect of the disorder is manifested as an exploration
bias corresponding to the ipsilesional deviation of the subjective visual
straight ahead estimation (SSA). The allocentric deficits can be observed
in the size estimation task as well as distance-judgment independently
of the relative position of the stimulus configuration to the individuals
head and body sagittal plane, thus attesting these patients a size- and
space-distortion beyond their ipsilesionally deviated SSA. Whereas optokinetic
or slow visual motion stimulation (SVS) as well as neck-muscle vibration
have been shown to modulate essentially the egocentric aspect, effects
on allocentric spatial deficits were only reported for optokinetic stimulation.
To this purpose we compared the modulatory effects of SVS and neck
vibration on two types of visual-spatial tasks in neglect: One egocentric,
measuring the visual subjective straight ahead (SSA) and three allocentric
spatial tasks in the horizontal dimension: Line bisection, size and distance
judgment. Five patients with left-sided spatial neglect were tested across
four conditions: A baseline condition without any manipulation was compared
with the experimental condition of SVS to the left and left neck vibration.
As control condition we used vibration of the left hand and righward SVS.
In the baseline condition all patients not only showed a right-sided
deviation of the visual straight ahead but also a left-sided size distortion
in line bisection and size estimation as well as a space distortion for
left-sided distance estimation. SVS to the left improved significantly
the SSA as well as the size and distance estimation. In contrast during
neck muscle vibration only the SSA was significantly ameliorated. In the
control conditions no influence on the allocentric tasks was obvserved.
The present data suggests that SVS not only affects the representation
of the subjective midline but also allocentric space representations. In
contrast the efficacy of neck muscle vibration seems to be specific for
egocentric tasks. One explanation is, that leftward background motion facilitates
the direction of attention towards neglected regions of space. Another,
compatible hypothesis, attributes the motion effect to a facilitating generation
of egocentric space representation by providing a directional visual input
to this disturbed representation. We propose that the largely intact visual
motion system is capable of modulating the neuronal activity in the lesioned
parieto-temporal cortex of neglect patients. These findings offer therapeutic
applications for neglect rehabilitation since motion stimulation obviously
exerts a powerful influence on perception.
The Role of Medial Prefrontal Cortex in the Representation of Task Specific Meaning
Sophie K. Scott, Alex Leff, Richard R. J. Wise.
Institute of Cognitive Neuroscience UCL.
Two positron emission tomography (PET) studies are presented which address the neural correlates of representations in different central executive tasks. A role is identified for left medial prefrontal cortex in the generation, representation and choice between alternative response mappings for individual stimuli. This system is involved in central processing of sound structure, or read words, when the task requires that the stimuli be assessed in a novel and ambiguous manner, in which multiple candidate responses are considered. Activity in this region correlates with longer reaction times in such choice tasks, but not in other tasks where the response class is not ambiguous.
Spatial attention and memory
vs motor preparation. The involvement of the premotor cortices as assessed
by fMRI.
Stéphane Simon*, Martine Meunierº, Anna Berardi*, Loÿs
Piettre*, Christoph Segebarth*, Driss Boussaoudº
* INSERM U438, Grenoble, France. º Institut des Sciences Cognitives,
Lyon, France
Introduction. Single cell recordings in awake behaving monkeys have
shown that, in visually guided movements, the majority of neurons in the
premotor cortex are preferentially active in relation to motor preparation,
rather than to spatial attention or memory1. To examine whether these findings
apply to the human brain, we used fMRI (ten healthy volunteers) combined
with experimental paradigms very much inspired by those applied in neurophysiological
studies in monkeys.
Material and methods. Paradigms. Block paradigms (8 min overall) were
applied, comprising three control and three activation epochs of equal
duration. Two paradigms were designed to dissociate activation related
to motor preparation (MP) from that reflecting spatial attention and/or
working memory (SAM). Subjects were asked to press one of two keys depending
on the color of visual cues presented on a screen. Visual stimuli and motor
responses were identical in the two conditions, but subjects performed
different tasks. In the SAM paradigm, control and activation epochs differed
in terms of attentional/mnemonic requirements, whereas in the MP paradigm
they were different in terms of motor preparation. An example of trials
performed during the control and activation epochs of the SAM paradigm
is shown below (Fig 1). The central cross represents the fixation cross.
White squares were presented successively at pseudo-random locations within
the visual field (presentation time 500 ms, delay between successive presentations
of white squares was pseudo-randomized, ranging between 1.2 and 2.0 s).
Following presentation of a variable number of white squares, two adjacent
red and green squares were presented for 1.5 s, and the fixation cross
turned either red or green. One of these colored squares was at the same
location as the last white square. During the activation epochs, subjects
were instructed to attend to and memorize the successive positions of the
white squares, and press one of the two keys depending on the color of
the square positioned at the location of the white square. During the control
epochs, the subject’s response was determined by the color of the fixation
cross. Thus, during the control epochs, the performance of the task did
not require spatial attention and/or memory.
MR acquisition and data processing. Examinations were performed at
1.5 T (Philips NT). A GRE EPI sequence was applied (TR=3.7s, TE=45ms, ?=90°).
The volume of interest was composed of 25 adjacent slices. Resolution was
(4 mm)3. Functional maps were generated by means of SPM 96. Statistical
significance threshold for clusters obtained in the group analysis was
established at p=.05.
Figure 1. SAM paradigm.
Results. SAM paradigm. A large activation was found, bilaterally, within the posterior superior parietal cortices. The right dorsolateral prefrontal cortex and, bilaterally, the lateral and mesial premotor cortices were also activated (Fig 2A). A faint activation was also found within the caudate nuclei.
Figure 2. Activations in
premotor areas for the SAM (A) and MP (B) paradigms. Talairach y coordinates
extend from -19 (top) to +9 (botttom). Right hemisphere is on right side.
MP paradigm. No activation was detected within the parietal cortices.
Frontal activations were found, bilaterally, within the latero-dorsal and
mesial premotor cortices, with extension into the left precentral gyrus
(Fig 2B). Subcortically, activations were also obtained within the caudate
nuclei, the right thalamus, the left substantia nigra and the red nucleus,
as well as within the left amygdala.
Discussion and conclusion. An interesting finding is the different
involvement of the premotor cortices for spatial attention/memory vs motor
preparation, as illustrated in Figure 2. The posterior part of the mesial
as well as lateral left premotor cortices are devoted to motor preparation.
In contrast, the anterior part of the lateral premotor cortex seems specifically
involved in spatial attention and memory. These observations are in line
with recent studies and support the notion of functional distinction between
anterior
and posterior regions of the premotor cortex2-4.
References
1. Boussaoud D et al., Behav Brain Res (1996), 72: 1-15.
2. Tanné J et al., NeuroReport (1995), 7: 267-272.
3. Johnson P et al, Cereb Cortex (1996), 6: 102-119.
4. Toni I et al, Cereb Cortex (1999), 9: 35-49.
Sensorimotor integration of retinal and gaze signals during smooth pursuit eye movements
Robert J. van Beers1, Daniel M. Wolpert2, Patrick Haggard1
1Institute of Cognitive Neuroscience, UCL (University College London),
UK.
2Sobell Department of Neurophysiology, Institute of Neurology, UCL,
UK.
To localise a seen object, the central nervous system has to integrate
the object's retinal location with the direction of gaze. We investigated
this process by examining the localisation of static objects during smooth
pursuit eye movements. We performed a series of experiments in which subjects
performed localisation tasks. From the localisation errors found in these
experiments we could derive the errors in the retinal and in the gaze signals
and also how these signals are combined.
During an eye movement, a static object's image moves across the retina.
Objects that produce retinal slip are known to be mislocalised [1, 2]:
objects moving toward the fovea are seen too far on in their trajectory,
whereas errors are much smaller for objects moving away from the fovea.
These effects are usually studied in experiments in which subjects localise
the moving object relative to a briefly flashed one during fixation. In
this situation the moving objects are mislocalised, but flashes are not.
In our first experiment we found that a similar differential mislocalisation
occurs for static objects relative to flashes in the retinally equivalent
case during pursuit, in accordance with earlier results [1, 3]. In addition,
we found that this effect is not specific for horizontal pursuit, but it
was also found in other directions. In a second experiment we examined
how this effect generalises to positions outside the line of eye movement.
We found that large localisation errors were found in the entire hemifield
ahead of the pursuit target, and these were predominantly aligned with
the direction of eye movement. In a third experiment, we determined whether
it is the flash or the static object which is mislocalised ahead of the
pursuit target. Subjects tracked a target and compared the positions of
two sequentially presented stimuli, the first one ahead of the pursuit
target and the second one behind it. We found that when the first stimulus
was a flash, it was mislocalised, but when it was a static object, it was
not. The nature of the second stimulus did not have an effect. These results
demonstrate that during pursuit it is the flash, not the static object,
which is mislocalised. Note that this is the reversed pattern of that during
fixation. In a fourth experiment, subjects pointed at flashed and static
stimuli which were presented in complete darkness during pursuit. The results
were similar to those of the previous experiment. This shows that the mislocalisation
of flashes during smooth pursuit reflects absolute rather than relative
localisation, and is therefore based on integrated retinal and gaze signals.
We conclude that the central nervous system compensates for the retinal
localisation errors (for objects which produce retinal slip) during smooth
pursuit eye movements in order to maintain position constancy for static
objects. This compensation is achieved in the process of sensorimotor integration
of retinal and gaze signals: different retinal areas are integrated with
different gaze signals to guarantee the stability of the visual world.
1. Mateeff, S. and Hohnsbein, J. (1988) Vision Research 28: 711-719.
2. Mateeff, S., Yakimoff, N., Hohnsbein, J., Ehrenstein, W. H., Bohdanecky,
Z., and Radil, T. (1991) Vision Research 31: 131-138.
3. Mitrani, L. and Dimitrov, G. (1982) Vision Research 22: 1047-1051.
State estimation under stochastic and deterministic time-varying contexts
Philipp Vetter & Daniel M. Wolpert
Sobell Department of Neurophysiology, Institute of Neurology, University College London
The consequences of a motor command depend on the prevailing movement context, that is the state of our own body and the environment. This context is not static but changes over time. For instance the weight of a bottle changes as we pour from it, and muscles fatigue with use. We explored how the central nervous system's estimate of finger position evolves when visual feedback is removed by introducing a discrepancy between the true and displayed finger position. In the first experiment the context changed stochastically between two discrete discrepancies, and in the second experiment the discrepancy changed continuously in a deterministic fashion.
In the stochastic experiment, subjects made sequential pointing movements, during which their visual feedback was offset from the finger by two possible fixed amounts. When visual feedback was removed, subjects' pointing tended toward that appropriate for the average of the two contexts. This behavior is optimal in the sense that it minimizes the mean squared pointing error.
In the deterministic experiment, the displayed finger height changed with respect to the true height in a sinusoidal fashion over time. Subjects were unaware of this perturbation, and continued to compensate for the time-varying sinusoidal discrepancy, even in the absence of visual feedback. Their performance demonstrates that subjects can estimate the rate of change of the context. Taken together, these results suggest that the central nervous system actively models both stochastic and deterministic components of the time-varying context without requiring awareness.
Beware and be aware:
Capture of attention by emotional stimuli in patients with hemispatial
neglect
Patrik Vuilleumier (1,2) & Sophie Schwartz (1,3)
(1) Institute of Cognitive Neuroscience, University College, London,
UK
(2) Department of Neurology, University of California, Davis, CA, USA
(3) Institute of Cognitive Sciences, University of California, Berkeley,
CA, USA
Adaptive-evolutionary arguments and empirical evidence suggest that
emotional significance of stimuli might prioritize the allocation of spatial
attention to potentially relevant stimuli. We asked whether such effects
might still occur in patients with unilateral neglect and visual extinction
who usually remain unaware of contralesional stimuli when these compete
with concurrent ipsilesional stimuli. In a first experiment, shapes or
faces with either neutral, happy, or angry expressions were presented in
right, left, or both visual fields. On bilateral trials, three parietal
patients extinguished faces on the contralesional side much less often
than shapes, and faces with happy or angry facial expressions much less
than neutral faces. In a second experiment, pictures of spiders or flowers
made of similar low-level features were presented in right, left, or both
fields. Again, parietal patients were more likely to perceive emotional
stimuli (spiders) than other similar but neutral pictures (flowers). We
suggest that in patients with neglect in whom mechanisms of spatial attention
are impaired after parietal damage, intact visual pathways to the ventral
temporal lobe and amygdala could still mediate mechanisms of "emotional
attention".
Impaired visual search
and extinction following reversible inactivation of monkey lateral intraparietal
area (LIP)
Claire Wardak 1 , Etienne Olivier 2 and Jean-René Duhamel
1
1 Institut des Sciences Cognitives, CNRS, Bron, France
2 Université catholique de Louvain, Bruxelles, Belgique
Behavioral expressions of unilateral sensory and/or motor neglect have been studied experimentally in monkeys following lesions restricted to the parietal cortex, to the frontal cortex and to certain subcortical structures. Severe and lasting deficits have rarely been observed in monkeys, as compared those described in humans following right hemispheric damage. This discrepancy has been interpreted as reflecting the absence of functional hemispheric lateralization in the monkey, which permits greater compensation of the deficits through neuronal plasticity mechanisms. Reversible inactivation methods overcome this problem by allowing the study of reproducible, acute behavioral effects in the same animal. We inactivated the lateral intraparietal area (LIP), a region of the parietal cortex implicated in visual attention and saccadic eye movements, by placing discrete injections of muscimol, a GABA agonist, at sites identified through single unit recordings, in a macaque performing visual search and saccades to simple and double simultaneous stimuli. No change in the latency or accuracy were observed during visually-guided saccades to single contralesional targets. However, the frequency of saccades to briefly flashed contralesional targets is reduced if a simultaneous ipsilesional target is presented. Also, the contralesional target was often undetected when presented alone. In the visual search task, feature search of a colored target among distractors of a different color is unaffected by LIP inactivation, but conjunction search of a target defined by both color and shape shows a marked contralesional impairment and a concurrent ipsilateral facilitation. These deficits show on saccade latency and on the proportion of trials where the initial saccade went to a distractor. Inactivation of the adjacent ventral intraparietal area (VIP) had no effects on any of these tasks. These results indicate that LIP inactivation produces a selective deficit in attentional target selection, but no saccade programming or execution impairment. Selective reversible inactivation of the parietal cortex can thus produce behavioral impairments analogous to those observed in human neglect, and could be used to gain a better understanding of how the functional subdivisions of the parietal lobe contribute to the different aspects of this syndrome.
Planning and action knowledge in schizophrenic patients: Comparisons with patients with frontal lobe damage.
Tiziana Zalla, Andres Posada, Nicolas Franck, Nicolas Georgieff, Angela Sirigu
Institut des Sciences Cognitives, BRON, France
Schizophrenia is characterized by disorders in a large variety of cognitive
domains including executive function, attention, memory, learning and language.
A line of researches has drawn attention to a putative frontal dysfunction
or a frontal dementia and documented that schizophrenics are impaired on
executive deficits. However, the issue whether executive disturbances in
schizophrenic patients correspond to selective deficits qualitatively comparable
to the performance of frontal lobe patients is still controversial, mostly
because standard neuropsychological assessments, such as the Wisconsin
Card Sorting Test, do not allow to disentangle the component processes
required for executing the task.
By using a set of multicomponent tasks which segregated various information
processes used in action planning, we assessed the ability to generated
knowledge and elaborate a plan of action in a group of seventeen schizophrenic
patients. Their performance was compared with that of a group of nine patients
with frontal lobe lesions and a group of sixteen normal control subjects.
Schizophrenics were found to perform significantly worse than normal subjects
and frontal lobe patients in retrieving action knowledge data base related
to non-routine activities. Moreover, similarly to frontal lobe patients,
they were impaired in processing the sequential organization of actions
in both the generation and the ordering tasks, in setting priorities among
the script actions, in script rule violation, and in estimating the importance
of the individual actions in relation to the script's goals. These results
indicated that schizophrenics are impaired in processing critical features
of action knowledge necessary to accomplish a plan of action and that this
pattern of deficits is similar to that observed in patients with damage
to frontal cortex.
Institut
des Sciences Cognitives UMR 5015 CNRS
- UCB LYON 1
67, boulevard
Pinel 69675 BRON cedex
33
(0)4 37 91 12 12 
33 (0)4 37 91 12 10 
berger@isc.cnrs.fr