Vous êtes sur la page 1sur 16

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/8978610

Why and how are posture and movement coordinated?

Article  in  Progress in brain research · February 2004


DOI: 10.1016/S0079-6123(03)43002-1 · Source: PubMed

CITATIONS READS
120 3,347

3 authors:

Jean Massion Alexey V Alexandrov


French National Centre for Scientific Research: emeritus Russian Academy of Sciences
238 PUBLICATIONS   6,330 CITATIONS    12 PUBLICATIONS   315 CITATIONS   

SEE PROFILE SEE PROFILE

Alexander Frolov
Russian Academy of Sciences
232 PUBLICATIONS   1,988 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Informational characteristics of NN Associtive Memories View project

All content following this page was uploaded by Jean Massion on 03 November 2017.

The user has requested enhancement of the downloaded file.


Progress in Brain Research, Vol. 143
ISSN 0079-6123
Copyright ß 2004 Elsevier BV. All rights reserved

CHAPTER 2

Why and how are posture and movement


coordinated?

Jean Massion1,*, Alexei Alexandrov2 and Alexander Frolov2

1
Laboratoire Parole et Langage, Universite´ de Provence, 13621 Aix-en-Provence, France
2
Institute of Higher Nervous Activity and Neurophysiology, Russian Academy of Science, Moscow 117865, Russia

Abstract: In most motor acts, posture and movement must be coordinated in order to achieve the goal of the task. The
focus of this chapter is on why and how this coordination takes place. First, the nature of posture is discussed. Two of
its general functions are recognized; an antigravity role, and a role in interfacing the body with its environment such
that perception and action can ensue. Next addressed is how posture is controlled centrally. Two models are presented
and evaluated; a genetic and a hierarchical one. The latter has two levels; internal representation and execution. Finally,
we consider how central control processes might achieve an effective coordination between posture and movement. Is a
single central control process responsible for both movement and its associated posture? Alternatively, is there a dual
coordinated control system: one for movement, and the other for posture? We provide evidence for the latter, in the
form of a biomechanical analysis that features the use of eigenmovement approach.

Introduction understand why and how this coordination occurs, it


is important to precisely define posture and explore
During the execution of voluntary movements, many how it differs from movement.
aspects of the same task are coordinated and A simple, coarse-grain definition is that posture
performed simultaneously. Gaze orientation (toward (also called ‘attitude’) is the body segments’ config-
a target) results from the parallel control of eye and uration at any given time (Thomas, 1940). This
head movements (Bizzi et al., 1977). Coordination definition is purely descriptive and it does not refer
between posture and limb movement is another to the many functions subserved by posture. To this
example seen in daily life. Often, two goals must be end, several approaches have been proposed in the
achieved at the same time. There is need for an literature. One refers to stance as a genetically
accurate performance of a goal-directed movement determined reference posture, with characteristics
on the one hand, and the maintenance of equilibrium which are species dependent. This view was cham-
and an appropriate posture, or set of postures, on the pioned in the classical works of Sherrington (1906),
other (for review, Massion, 1992, 1994, 1998; Horak Magnus (1924) and Rademaker (1931). They all
and Macpherson, 1996). Similarly, the elaboration of emphasized that the body’s orientation with respect to
a locomotor gait requires coordination between a gravity was determined by a group of reflexes that sets
postural function (support of the body segments the body segments’ collective orientation and stabi-
against gravity; forward acceleration of the center of lizes this orientation against external disturbances.
mass, CM) and the rhythmic movements of loco- A second approach is related to the concept of
motion (Mori, 1989; Mori et al., 1999). In order to support. Kuypers (1981) proposed it on the basis of
anatomical and functional considerations. In this
*Corresponding author: Tel.: þ 33-4-4257-1228; concept, axial and proximal body segments, and their
Fax: þ 33-4-4257-1228; E-mail: jean.massion@lpl.univ-aix.fr corresponding musculature, serve as a support for the

13
DOI: 10.1016/S0079-6123(03)43002-1
14

distal musculature, such as the hands for reaching chain from feet to head should be able to support the
and grasping. This view was actually proposed body’s weight against gravity forces and resulting
previously by Hess (1943). He made a distinction ground reaction forces. In addition, the kinematic
between the ‘eiresmatic’ (supporting function) aspect chain should be able to provide a dynamic support to
of a motor act and the ‘teleokinetic’ (goal-directed the moving segments when they are involved in a
movement) aspect. Hess proposed that missing the goal-directed action.
goal resulted from the lack of appropriate support for The second process is equilibrium control. It
the movement (see Stuart et al., Chapter 1). More ensures that under static conditions, the center of
recently, Bouisset et al. (1992) proposed a modified pressure and the projection of the center of gravity
version of the support concept, ‘posturo-kinetic (CG) should remain inside the support surface (i.e.,
capacity’. It involves the capacity of the supporting the support’s contact with the environment).
segments to assist the ongoing movement in terms of
their speed and forcefulness.
A third approach is to distinguish relatively sharply Interface between perception and action
between posture and movement. Its proponents argue
that the maintenance of a posture means that the Each body segment position can be defined with
position of one or several body segments are fixed with respect to the other body segments. An egocentric
respect to other segments (also termed ‘joint fixation’) reference frame is used to calculate each body
or with respect to space (i.e., the external environ- segment’s position with respect to the axis of a
ment). This position is stabilized against external given reference segment, such as the trunk or head.
disturbances. In contrast, ‘movement’ to these work- When a target is moving in external space, its
ers means a change of posture (or position), i.e., the position can be coded into a coordinate system
initial posture is destabilized to achieve a new external to the body (allocentric reference frame). The
position, which is reached through a trajectory (see latter usually uses the vertical gravity axis as a
also S. Mori et al., Chapter 33). Feldman’s (1966) reference (Berthoz, 1991; Paillard, 1991). In order to
equilibrium point hypothesis (for an introductory match the egocentric reference frame to the external
review: Latash, 1993) is in line with this view. world, the orientation of the trunk or head (the
referent in the egocentric system) is calculated with
respect to the vertical gravity axis (the referent for the
Posture has two primary functions
external world). This calculation is based primarily
on a set of sensors: the otoliths, and body gravi-
It is clear that the current definitions of posture are
ceptors (in the soles of the feet and pelvis, possibly
multiple, each resulting in different experimental and
also muscle Golgi tendon organs), which monitor the
theoretical approaches. One way to identify a
gravity axis (Horak and Macpherson, 1996; Massion,
unifying model is to ask the question ‘‘what are the
1998). The known orientation of the head and/or
underlying postural functions common to these
trunk with respect to the gravity axis can then be
different views?’’. From this perspective, two primary
utilized both for the calculation of the other body
functions emerge: antigravity control, which requires
segments’ orientation with respect to the external
the buildup against gravity of the body segments’
world, and the external target’s position with respect
collective, overall configuration, and the interface
to the body (Assaiante and Amblard, 1993;
between perception and action (i.e., controlling the
Mouchnino et al., 1993; Pozzo et al., 1995). The
relationship between the external world and the
reference is for the perception of body’s movements
body).
with respect to the environment and the control of
balance. Body segment orientation with respect to the
Antigravity control vertical axis also serves as a reference frame for
calculation of a target’s position in space and
Two related processes accomplish antigravity con- the trajectories required to reach this target (Wise
trol. First, the various segments of the kinematic et al., 1991).
15

Synthesis of the two roles These three functions are closely related to the
primary ones of antigravity control and interface
Antigravity control and interface between perception between perception and action.
and action are closely related to the execution of
movement. For example, when raising a leg laterally, Body orientation with respect to gravity
equilibrium is maintained by shifting the CG toward
the supporting leg (Mouchnino et al., 1992). This is a In mammals, with a body which is segmented into
form of antigravity control. During the same task, it head, trunk and leg sections, the otoliths and vision
can be shown that the orientation of body segments provide sensory input for controlling the head’s
with respect to space, which serves as an interface orientation with respect to gravity. The investigations
between perception and action, is also preserved. In on righting reflexes described by Rademaker (1931)
untrained subjects, the head axis remains vertical illustrate this claim. Note that a cat falling from an
throughout the movement, whereas the trunk is upside-down initial posture first reorients the head
inclined toward the supporting leg. In expert dancers, along the horizontal plane, then the body along the
however, the trunk axis also remains vertical. head plane, and finally, the legs adopt a vertical
Mouchnino et al. (1992, 1993) have shown that the position that makes the animal ready for landing.
trunk axis serves as a reference frame for the There are also a series of reflexes aimed at stabilizing
calculation of leg angle (egocentric reference frame). the segmental orientation with respect to gravity. For
While the trunk axis remains vertical throughout the example, Vestibulo-collic reflexes stabilize head
movements of an elite dancer, the calculation of leg orientation in all manners of movements. Other
angle relative to the trunk (egocentric coordinates) reflexes help orient the foot with respect to the support
becomes equivalent to the calculation of leg angle (e.g., placing reactions) or the leg with respect to the
relative to the vertical gravity axis. The latter is the gravity axis (hopping reactions).
reference axis for the external space coordinating
system. Thus, for the expert dancer, keeping the trunk
Antigravity function
axis vertical throughout the task clearly simplifies the
calculation of leg position in space.
There is need to support the body segments against
the contact forces exerted by gravity forces acting
Central organization and control of posture
upon individual segments and the body as a whole.
Postural muscle tone (i.e., the level of EMG activity
Each species has its own genetic patrimony, which
in antigravity muscles) is a key means of meeting this
provides throughout fetal development, birth and
need. It depends in large part on operation of the
ontogeny a central and reflex organization support-
myotatic (stretch reflex) loop, predominantly in
ing basic behaviors, which are critical for survival.
antigravity extensor muscles. Such muscle activity is
There is consensus that posture and locomotion are
exaggerated in decerebrate animals, which exhibit the
examples of such basic, genetically determined behav-
well known, albeit mistermed, ‘decerebrate rigidity’
iors. There has been much controversy these past
(i.e., it is really a form of spasticity). These postural
three decades, however, concerning the relative contri-
reactions, which occur in the presence of an external
bution of genetic versus learned networks to the central
perturbation of stance, are determined, in part, by
organization and control of posture (Prochazka
both genetic endowment and learning throughout
et al., 2000; Grillner and Wallén, Chapter 1).
ontogeny (Forssberg, 1999).
Genetic model of posture
Adaptation of the body’s segments to ongoing
Three main functions are identified in the genetic movement
model of posture: the body segments’ orientation
with respect to gravity, their support against gravity Genetic patrimony is also fundamentally concerned
and the adaptation of posture to ongoing movement. with the coordination between posture and
16

movement. The distribution of tone throughout the Anticipatory postural adjustments


body’s musculature depends on the orientation of the
head in space and in relation to the trunk, as achieved Analysis of the coordination between posture and
in part by labyrinthine and neck reflexes. Note also movement argues against a purely genetically based
the role of lumbar reflexes, which are concerned with reflex organization of posture. During the perfor-
the orientation of the trunk with respect to the pelvis mance of most voluntary movements, the primary
(Tokizane et al., 1951). These latter reflexes adapt functions of posture must be preserved in order to
tone in muscles supplying the upper and lower accomplish the goal in an accurate fashion. Preserving
segments to the ongoing movement. equilibrium is a necessary condition for an accurate
In summary, the genetic model of posture proposes performance and preserving the orientation of the
that its purpose is to support the body against body’s segments with respect to gravity are essential
gravity, preserve balance, orient the body with because they provide a reference value for planning the
respect to gravity and also adapt the body’s segments trajectory of the movement toward its goal. In
to the ongoing movement. addition, the supporting function of the kinematic
chain from the ground to the moving body segments
has to be preserved during the dynamic conditions of
Limitations of the genetic model
movement execution.
The coordination between posture and movement
There are several challenges to the idea that the
must accommodate a major problem. The perfor-
genetic model of posture can be the sole basis for the
mance of a movement is a source of disturbance of
central organization and control of posture. Two key
posture for two reasons. First, it changes the body’s
issues concern the flexibility of postural reactions,
geometry and, as a consequence, the CG’s position
and anticipatory postural adjustments, of which the
with respect to the support surface (often the ground).
latter ones are involved in the coordination between
Second, the movement is initiated by internal muscle
posture and movement (see also F. Mori et al.,
forces, which are associated with reaction forces
S. Mori et al., and Takakusaki et al., Chapters 19, 33
acting on the supporting segments and the ground.
and 23 of this volume).
The resulting dynamic interactions between body
segments must be accommodated. Otherwise, there
Flexibility of postural reactions would be a deviation from the planned trajectory and
misreaching would occur. This is prevented by
Postural reactions induced by external disturbances anticipatory postural adjustments. These were first
are now known to be remarkably flexible (for review, described by Belenkiy et al. (1967) for an arm-raising
Macpherson, 1991; Horak and Macpherson, 1996). task and they have since been observed in a wide
For example, when a balance disturbance occurs variety of voluntary movements (for review, Massion,
while standing, the leg muscles are the main ones 1992; Horak and Macpherson, 1996; see also
involved in the correction. If the subject simulta- Bouisset and Zattara, 1987). Figure 1 models how
neously grasps a support with the hands, however, anticipatory adjustments correct in advance for
the postural reaction will mainly involve the arm perturbations of posture and equilibrium that are
muscles (Nashner and McCollum, 1985). Thus, the associated with the execution of movement.
postural reactions to the same stimulus depend, in Anticipation involves prediction of a forthcoming
terms of their spatial distribution, on external perturbation. It implies that due to the effect of
constraints, such as the conditions of support. This repeated experience and learning on adaptive central
marked flexibility is, however, a feature of proprio- neural networks, memorized models are developed in
ceptive reflex organization (Forssberg et al., 1975; the central nervous system (CNS) and used during the
Stuart, 2002). Segmental reflexes are both task and performance of a task. Such models accommodate the
context dependent, and under a higher level of external world, the body’s biomechanical character-
control for selecting appropriate reflex actions (gate istics and their interactions. This idea of a memorized
and gain control). representation (or model) was first proposed for the
17

movements, are much better accounted for by a


hierarchical model of posture. This approach is
derived from an analogous model of movement
organization proposed initially by Bernstein. (The
evolution of this idea, too, is presented in
Bernstein’s 1967 volume.) It includes both inborn
genetically endowed reactions (Gurfinkel and Shik,
1973) and those built up by learning (Clément et al.,
1984; Gurfinkel et al., 1988; Gurfinkel and Levik,
1991). This model proposes that two types of CNS
processing are required for the control of posture.
A higher-order one accomplishes an internal repre-
Fig. 1. Schematic representation of the integrated control of sentation of body posture (the so called ‘body
posture and movement. Note the need for a postural control schema’). In parallel, lower-order control brings
system that integrates interactions between postural reactions
about the kinematics and force (kinetics), which are
and anticipatory adjustments. The execution of a movement
provokes a perturbation of posture (internal perturbation), required for implementing postural functions as a
which is compensated for by a postural reaction. This occurs necessary aspect of the task. In reality, this proposal
with a delay, however. It is comprised of the time for the is not basically different from the body schema
perturbation to be detected plus that for transmission through model proposed by Head (1920) on neurological
the corrective pathway(s). In contrast, anticipatory postural grounds, or the concept of internal representations
adjustments correct in advance for the forthcoming perturba-
tion. Adaptive neural networks, whose development requires (or models) required for the organization of
learning, control this anticipatory process. movements (Wise et al., 1991; Wolpert et al., 1995,
1998). For example, in an arm target-reaching task,
whole somatic system by Bernstein on the basis of his one may identify a set of control processes, which
analysis of motor learning (for the evolution of his depend on representational mechanisms, such as the
ideas, see his posthumous 1967 volume). The idea was coding of target position in head and trunk
further extended into the postural domain by Clément coordinates, the coding of hand position in trunk
et al. (1984), Gurfinkel et al. (1988) and Gurfinkel and coordinates and planning the desired trajectory. All
Levik (1991). It has also appeared in the area of three processes require internal models. For the
oculomotor control, where it was initially termed execution of an arm-reach trajectory, both static
‘internal models’ (see Robinson, 1986). and dynamic models, and direct and inverse models,
The genetic model of posture cannot account for are used for implementing the command in terms
the important role of anticipation in helping achieve of muscle force (Wise et al., 1991; Wolpert et al.,
an appropriate coordination between posture and 1998).
movement. The key reason is that such anticipation is
based on learning.
It thus appears that in addition to the genetic model Higher-order processing of body representation
of posture, which nonetheless remains valid for many
basic postural controls, a higher level of postural Higher-order processing has been shown in experi-
organization is present. It makes use of experience and ments that made use of artificial or biased sensory
learning, and provides a finely tuned adaptation of inputs. These have included a visual-moving scene
postural functions to the movement needs of daily life. (Dichgans et al., 1972; Lestienne et al., 1977),
galvanic stimulation of the labyrinth (Lund and
Broberg, 1983; Gurfinkel et al., 1988; Hlavacka et al.,
Hierarchical model of posture 1995) and tendon vibration (the latter stimulus
activates largely muscle sensory Ia axons; Roll and
The flexibility of postural reactions, and anticipa- Roll, 1988; Roll et al., 1993, 1998). These studies
tory postural adjustments associated with voluntary showed that the internal CNS representation of the
18

body’s posture (or body schema) includes at least Lower-order implementation of posture
three components. First is body kinematics, i.e., the
kinematic chain from feet to head. These make use of The higher-order component of the hierarchical
muscle proprioceptive Ia afferent input and/or an model of posture is rather stable even when the
efferent copy of the motor command. The latter constraints are drastically changed. This occurs, for
mainly concerns the position of the eyes in their example, when the subject is subjected to micro-
respective orbits (Roll et al., 1998). Second are the gravity conditions (Gurfinkel et al., 1993a,b; Massion
body segments’ mass and inertia, which are derived et al., 1998). In contrast, the lower-order control of
from sensory inputs not yet clearly identified. Finally, implementation (kinematics and kinetics) is quite
there are the positions of the body’s individual flexible and adaptable to the constraints.
segments and their weight, both with respect to the Experiments performed in microgravity have
external world. The vertical axis related to the gravity illustrated this initially surprising claim (Vernazza-
vector is assessed by sensory inputs from multiple Martin et al., 2000; Baroni et al., 2001). For example,
sources, including vestibular graviceptors, body the performance of trunk bending in normogravity is
graviceptors (pelvis—Mittelstaedt, 1998; plantar characterized by opposite movements of upper and
sole—Kavounoudias et al., 1998, 1999) and Golgi lower body segments. These minimize a shift of the
tendon organ receptors in extensor muscles (Dietz CM’s horizontal position with respect to the feet, and
et al., 1989, 1992). In addition, visual input provides thereby preserve equilibrium during the movement.
information about the vertical axis (Dichgans et al., In a short-term exposure to microgravity, such as a
1972; Lestienne et al., 1977; Dijkstra et al., 1994). parabolic flight, the kinematics of axial movements
Finally, haptic (tactile) contacts of the hands on during trunk bending is preserved just as well as in
external supports, like a wall or a cane in contact normogravity. Similarly, the close covariation
with the ground, serve to adjust the body’s posture between ankle, knee and hip joint movements is
relative to the external support (Jeka and Lackner, retained (Vernazza-Martin et al., 2000). Also, the
1995). CM’s displacement during movement is minimized in
According to Mergner and Rosemeier (1998), the microgravity to the same extent as in normogravity
representation of the body segments’ positions with (Massion et al., 1997; Baroni et al., 2001). All these
respect to space is accomplished in two ways. A top- findings suggest that in normogravity, a movement is
down one uses the head as a reference frame with planned kinematically on the basis of the internal
respect to space. The position of the various body representation of body kinematics (higher-order
segments can then be calculated with respect to the processing). This representation is well preserved in
head. A bottom-up representation uses the basis microgravity. In contrast, the EMG patterns of the
of support as a reference frame. This serves mainly limbs’ musculature, which correspond to the imple-
to calculate the CG position with respect to the mentation of kinematic synergies, are markedly
ground. changed in microgravity. This shows that muscle
Evidence for this internal representation of the forces adapt to new constraints in order to maintain
proprioceptive kinematic chain is illustrated by an invariant kinematic synergy.
experiments which noted the effects of vibration of
the biceps’ tendon when a human subject was holding
the nose or skull with a hand (recall that biceps Summary on models of posture
vibration stimulates Ia biceps afferents, thereby
mimicking his muscle’s stretch during arm extension). Both the genetic and hierarchical models of posture
Surprisingly, the resulting illusory perception is an make use of several identical mechanisms for
elongation of either a finger or the skull (Lackner, achieving postural control. These include: (1) joint
1988). This finding shows that Ia afferent input is stiffness, which restricts the body’s deviation from the
processed by the CNS in such manner as to provide desired position while standing; (2) postural reac-
the most likely interpretation according to the tions, synergies and strategies, which reduce the
subject’s previous experience. effect of postural and balance disturbances; and
19

(3) anticipatory postural adjustments, which are movements, such as arm-reaching ones. The unique
associated with movement performance in order to feature of posture in the execution of any movement
preserve the body segments’ orientation and equili- involves the need to control balance and/or the
brium, and provide appropriate postural support for body’s segmental orientation during the motor act.
an accurate movement performance. Despite these Thus, two primary problems of postural control in a
similarities, the genetic model alone cannot explain given movement task require resolution. The first is
the flexibility of postural reactions and anticipatory how to simplify control by reducing the number of
postural adjustments. The hierarchical model of degrees of freedom (Bernstein, 1967). The second
posture, with its higher-order processing of the body problem is how to counteract the disturbing effects of
schema and lower-order processing of implementa- movement performance on posture in terms of body
tion, is far more appropriate for accommodating the segment orientation and equilibrium. It is interesting
experimental findings in the domain of posture and to see how far the various hypotheses proposed for
posturo-kinetic coordination. It is not exclusive of movement control can accommodate the above two
the genetic model, however, because it uses many requirements. Three such concepts are presented and
genetically prewired circuits for accomplishing evaluated below, using the guidelines that (1) internal
postural tasks. representations of body kinematics and dynamics
provide a basis for planning movement trajectory
(internal ‘direct’ model), and (2) during the execution
Integrated CNS control of posture and movement of movement, dynamic interactions between seg-
ments perturb the movement’s trajectory and the
In daily life, postural control is most intimately body’s equilibrium.
associated with the execution of movements (reaching
and grasping tasks, locomotion, etc.). As such, it is
one key aspect of the overall control of motor tasks, ‘Referent posture’
which usually involve both postures and movements.
A major difficulty in postural control, especially in An early 20th century concept in motor control was
humans, concerns the ‘multijoint kinematic chain’ that EMG patterns were sufficient to reveal the
from the feet to the head. It includes body segments characteristics of a CNS-operated central pattern
with quite different mass and inertia, which are linked generator (Brown, 1911; see also Pearson’s present
by muscles with their own idiosyncratic viscoelastic Chapter 12). This concept was first put into question
characteristics. Their totality (for the whole body) is by Wachholder (1928). He introduced the idea that
responsible for the production of force and kine- the neural pattern also depended on the biomecha-
matics. For example, Eng et al. (1992) showed that nical characteristics of the segment to be moved.
for an arm-reaching task, each single joint movement These include inertia, weight and the dynamic
involves dynamic interactions with the other seg- characteristics of the task, such as velocity.
ments of the kinematic chain, and this becomes a Similarly, Weiss (1961) insisted on the fact that the
source of postural disturbance. In the human, the EMG pattern of individual muscles in a given goal-
relatively high level of the CG’s position, and the oriented task varied from trial to trial without
narrow support surface for the feet, are other sources changes in the characteristics of the overall perfor-
of instability, which further confound postural mance. He introduced the concept that the perfor-
control during the execution of movement. mance resulted from the action of the whole system
It bears emphasis that the multijoint chain from involved in the task (for review, Paillard, 1960;
the feet to the head is not specific to posture. Wiesendanger, 1997; Sternad and Corcos, 2001). The
Movement of segments of this chain accompanies all lamba model of motor control was proposed by
manner of bodily movements during daily life, e.g., Feldman (for its history, see Feldman and Levin,
trunk bending forwards and backwards. As such, the 1995). This idea is in keeping with Bernstein’s (1967)
control of axial movement involves issues not views on the need for a systemic approach. Feldman
basically different from those for other multijoint has argued that EMG patterns are not directly
20

programmed by a central controller, rather, they direct dynamic model is used to assess on-line the
emerge from the interaction between the central dynamic state of the body’s biomechanical system.
command and pattern-generating signals, reflex The inverse dynamic model represents the central
signals and the biomechanical system’s components. command which takes into account all of the
Furthermore, central control must also accommodate dynamic properties of the system required to perform
the biomechanical system’s interactions with the a given task. These concepts have been used by
environment (see also Hasan and Enoka, 1985; several authors (Kawato et al., 1987; Atkeson, 1989;
Latash, 1993). Miall et al., 1993; Wolpert et al., 1995).
The concept of a ‘referent posture’ (Feldman et al., Gomi and Kawato (1996, 1997) claim that
1999) is related to Feldman’s lamba model. (The Feldman’s equilibrium point model cannot account
equilibrium point hypothesis or alpha model, which for dynamical interactions between segments. Thus, it
is based on other premises, also proposes that the fails to describe the accurate performance of a goal-
final posture is programmed by the CNS; see Bizzi directed task. They have opined that a direct dynamic
et al., 1992.) Feldman’s referent posture is based on model and/or an inverse dynamic control is more
the idea that the CNS controls the joints’ positions by likely, i.e., one which emphasizes that the CNS has
adjusting the spring-like properties of the muscles. a control system for dynamic interactions between
These are determined, in part, by the muscles’ body segments during the accurate performance of
‘threshold’ for their own myotatic reflex. This spring- a task. In keeping with Ito (1984), they have proposed
like adjustment results in an equilibrium position of that the cerebellum plays a critical role in building
the joints being reached, thereby defining a final and storing the direct and/or inverse models
position of the limb. A movement results from a appropriate for each motor task (cf. Wolpert et al.,
change in the equilibrium position of the joints, 1998).
which also produces a new referent postural In summary, direct and inverse dynamic models
configuration. The actual body configuration differs account for both balance control and movement
from that of the referent one, however, due to the performance by the use of a single central control
action of external forces, such as gravity. system.
Feldman’s concept is attractive due to its relative
simplicity. A single control process results in
achieving both the movement trajectory and pre- Concept of dual, coordinated control
serving postural functions. There is thus a reduction
of the controlled degrees of freedom. The main The above two concepts feature motor actions being
controversy about this concept concerns the control controlled as a whole, i.e., no specific separate control
of dynamic interactions between the body’s segments, processes for posture and movement. The third
which disturb movement trajectory, posture and concept in vogue is that two independent, yet
equilibrium. For example, is muscle stiffness along coordinated, control processes are used for the
the kinematic chain able on its own to efficiently execution of motor acts. One is for the prime
minimize the disturbing effect on the joints’ move- (focal) movement, and the other for the associated
ments on the dynamic interactions between body postural movement (in terms of equilibrium pre-
segments? This question is central to evaluation of servation and appropriate orientations of the body
Feldman’s ideas (Gomi and Kawato, 1996, 1997). segments). The suggestion that any movement
comprises these two components was first claimed
in the works of Hess (1943) and Belenkiy et al. (1967).
Direct and inverse dynamics More recently, Cordo and Nashner (1982) proposed
that the same repertoire of synergies could be utilized
Direct and inverse dynamic models of the control of for both reactive and anticipatory postural adjust-
posture and movement are based on the assumption ments associated with a voluntary movement. They
that the CNS has used learning to create an internal demonstrated that the postural component often
model of the dynamics of the motor system. The started earlier than the prime one in order to
21

minimize the perturbation of equilibrium caused by The eigenmovements represent movements along the
the expected prime movement component (Cordo eigenvectors of the motion equation. Each eigenvec-
and Nashner, 1982). Bouisset and Zattara (1987) tor defines a set of ratios between the relevant joint
have argued that these anticipatory postural adjust- angles and joint torques. The number of eigenvectors
ments are preprogrammed and integral parts of the is equal to the number of degrees of freedom in the
central motor program for the movement in question. multijoint system. For example, for the three joints
This concept of a dual-coordinated control for primarily involved in a trunk-bending task (hip, knee,
posture and movement has been questioned in several ankle), there are three degrees of freedom. Three
studies. Some of them have involved an analysis of eigenvectors define for each joint a given set of ratios
trunk bending. In this task, the movement involves between the three joints’ angles and their correspond-
motion of the upper trunk. Such motion imposes a ing torques. An eigenmovement is a unique movement
large dynamic perturbation between body segments in which the ratios between the changes in joint angle
along the kinematic chain from the ground to the are fixed and, simultaneously, the changes in joint
head. The task requires a precise coordination torque are fixed. Since in each eigenmovement, all
between movement and posture. The size of the joint angular changes are synchronized, all angular
base of support is limited by the relatively small changes in each eigenmovement can be described by
length of the feet, and the CM is located at a the time course of a single eigenmovement’s ‘kine-
relatively large distance above the support. On the matic amplitude’. Similarly, all joint torques changes
other hand, decomposition of trunk bending into its can be described by the time course of a single
prime movement and postural components is not so eigenmovement’s ‘dynamic amplitude’. The relation-
evident as in the case of an arm-reaching task. Trunk ship between kinematic and dynamic amplitudes of
bending is accompanied by opposite movements of an individual eigenmovement is described by the
the lower segments (Babinski, 1899; Oddsson and simple motion equation of an inverted pendulum
Thorstensson, 1986; Crenna et al., 1987; Pedotti et al., with its individual inertia. Any given multijoint
1989). The identification of a postural versus move- movement can be represented as a superimposition of
ment component cannot be made on the basis of the eigenmovements of the relevant joints. Due to the
movement kinematics. The motions of the various eigenmovements’ dynamical independence from each
joints along the kinematic chain (hip, knee, ankle) are other, and due to the simplicity of their motion
too highly synchronized. This suggests that a single equations, the eigenmovement approach is a standard
central control system is responsible for all of the and useful method for the analysis of the dynamics
kinematics (Alexandrov et al., 1998). As a result, a of any multijoint system, be it living or inanimate.
different approach becomes necessary, like the one Recently, we applied this method to the investigation
presented in the next section. of human trunk bending in the sagittal plane
(Alexandrov et al., 2001a,b). For this, a three-joint
(hip, knee, ankle) biomechanical model of the human
The eigenmovement approach body was used.
The fixed ratios between joint angles and their
The main difficulty in the control of a multijoint corresponding joint torques, and the inertia in each
mechanical system, such as the human body, is the eigenmovement, are determined by the structure of
dynamical interaction between its different segments. the multijoint chain (from feet to head in the present
The movement of any one segment influences the case) and by the inertia of its segments. Figure 2
movement of all the other segments of the kinematic represents the ratios between joint angles (stick
chain. As a result, the control of a given single diagrams) and joint torques (sizes of circles) for each
segment cannot be performed without simultaneous eigenmovement of our three-joint human body
control of the neighboring ones. For such multijoint model, using standard anthropometric parameters
systems, however, a set of dynamically independent (Winter, 1990).
multijoint movements (multijoint motion units) can In Fig. 2, the eigenmovements are termed
be described by using the eigenmovement approach. according to their predominant joint. For our
22

hip flexion, the H-eigenmovement was the main


reason for a backward shift of the CM (to about
4.5 cm), which perturbed equilibrium. The A-eigen-
movement (ankle dorsiflexion with knee extension
and hip flexion) was responsible for only 5% of total
angular variance at the hip, knee and ankle. Its
contribution to the total CM displacement was
comparable to that of the H-eigenmovement, how-
ever. The changes in ankle angle produced a forward
CM shift of about 5 cm, which compensated for the
backward CM shift caused by the H-eigenmovement.
Therefore, the A-eigenmovement was concerned
Fig. 2. Application of the eigenmovement approach to a mainly with equilibrium control during the task.
human trunk-bending task. Kinematic (stick diagrams) and The above analysis suggests that in a trunk-
kinetic (circles at joints) patterns of ankle (A-), knee (K-), and bending task, the CNS uses a dual-control process;
hip (H-) eigenmovements during voluntary trunk bending. The one for bending, per se (the H-eigenmovement), and
term, ‘eigenmovement’ is defined in the text. In each
the other for equilibrium maintenance (the A-eigen-
eigenmovement, the sticks’ inclinations and the sizes of circles
correspond to the ratios between the changes in joint angles and movement). The arguments in favor of the idea that
joint torques, respectively. The values at the bottom correspond the A- and H-eigenmovements are controlled in the
to the normalized ratios between the three eigenmovement’s task as entire (involving all three joints) motion
inertias (i.e., relative to a value of 1 for the K-eigenmovement). units were developed previously (Alexandrov et al.,
2001a,b).
An important aspect of the control of H- and
trunk-bending task, ankle (A-) eigenmovement A-eigenmovements during trunk bending is their
involved ankle dorsiflexion combined with knee relative timing. Due to the large inertia of the
extension and hip flexion. Hip (H-) eigenmovement A-eigenmovement (five times greater than that of
consisted of hip flexion combined with knee and H-eigenmovement), it must start earlier and last
ankle extension. Finally, knee (K-) eigenmovement longer than the H-eigenmovement. Our previous
was composed of knee and hip flexion, and ankle modeling experiments on forward bending on an
dorsiflexion. The inertia of the A-eigenmovement was extremely narrow support (Alexandrov et al., 2001a)
five times greater than that of the H-eigenmovement. showed that when both eigenmovements started
In turn, the inertia of the H-eigenmovement was 20 simultaneously, the A-eigenmovement, which pro-
times greater than that of the K-eigenmovement. In duces forward acceleration of the CM, could not
our three-joint model, any arbitrary movement in the compensate for the backward CM movement
sagittal plane could be represented, in terms of its associated with the H-eigenmovement. As a result,
kinematics and kinetics, as a superimposition of these the body experienced a fall backward.
three eigenmovements. Fig. 3 shows that during our present trunk-
We found previously that the K-eigenmovement bending task, the A-eigenmovement actually started
(again, knee and hip flexion with ankle dorsiflexion) 70  20 ms earlier and lasted longer (for 820  200
during human forward upper trunk bending was ms) than did the H-eigenmovement (510  80 ms).
negligibly small (Alexandrov et al., 2001b). Therefore, The initial forward CM acceleration in the A-eigen-
we could decompose the task into only two (H- and movement was presumably produced by activ-
A-) of our three eigenmovements. We found that the ation of ankle dorsiflexors. This interpretation
H-eigenmovement (hip flexion with knee and ankle clarifies the functional meaning of the early EMG
extension) was responsible for the main part of the (activation) pattern of tibialis anterior, which is
movement kinematics (i.e., it described more than usually observed in trunk-bending tasks (Oddsson
95% of the total angular variance at the hip, knee and Thorstensson, 1986; Crenna et al., 1987; Pedotti
and ankle). Due to the ankle extension accompanying et al., 1989).
23

the CM and minimizing its perturbation by the prime


mover (i.e., the H-eigenmovement). Third, recall the
experiments of Horak and Nashner (1986), in which
they studied postural reactions to an external
disturbance of the erect posture. They proposed
that two strategies, hip and ankle ones, are the main
mechanisms for restoring balance control in response
to sudden perturbations. They showed that an ankle
strategy (rotation of the whole body around the ankle
joint) was of much higher inertia than a hip strategy
(rotation of the trunk around the hip). They argued
Fig. 3. Calculated changes in the CG’s velocity for A- and that the two strategies were task dependent, with the
H-eigenmovements during voluntary forward trunk bending by ankle strategy used to compensate for small and slow
a human subject. Note that a positive (upward) velocity
corresponds to forward CG displacement, and vice versa. The perturbations, whereas the hip strategy accommo-
thin vertical lines show the onset and offset of the CG velocity dated large and fast ones.
due to the H-eigenmovement. The vertical broken lines show the The ankle and hip strategies of Horak and
onset and offset of the CG velocity due to the A-eigenmove- Nashner (1986) can be identified in our A- and H-
ment (threshold effect at 5% of the maximal velocity). Note eigenmovements. For example, in the A-eigenmove-
that the onset of the CG velocity changes due to the A-
eigenmovement started earlier and ended later than did that due ment (ankle dorsiflexion with knee extension and
to the H-eigenmovement. Note also that the velocity changes hip flexion), the ankle dominates, just as it does in
because these two eigenmovements are in an opposite direction. their ankle strategy. Similarly, the kinematic
Abbreviation: CG, center of gravity. See text for further pattern of H-eigenmovement (hip flexion with knee
information.
and ankle extension) is close to that of their hip
strategy (see also Kuo and Zajac, 1993). Note further
The results of our eigenmovement approach that the high inertial A-eigenmovement (again like
suggest that the CNS needs an internal model of the ankle strategy in the experiments of Horak and
the inertia of the mentioned multijoint motion units. Nashner) is efficient in slow CM displacements,
It can use this model to adapt the timing of its neural whereas the five times less inertial H-eigenmovement
control mechanisms to the task in question, i.e., (as in their hip strategy) is efficient when the
movement performance (trunk bending in our perturbation is fast.
experiments) and equilibrium maintenance. In summary, our biomechanical analysis of human
trunk bending suggests that hip and ankle strate-
gies (best quantified as H- and A-eigenmovements)
Summary on eigenmovements represent elementary multijoint motion units of the
biomechanical system. They can be independently
We believe that the eigenmovements identified by our controlled by the CNS to achieve different behavioral
analysis correspond not only to purely mathematical goals within the same task (trunk bending using
formalisms, but also to real components of the CNS’s H-eigenmovement control, and equilibrium main-
control system. First, in each eigenmovement, joint tenance using A-eigenmovement control). They act
angles and joint torques change synchronously in all conjointly to prevent a potentially forthcoming
three (hip, knee, ankle) joints. In our opinion, this is disturbance of posture and equilibrium provoked by
the simplest scaling procedure to be controlled by the the movement performance. The same motion units
CNS. Second, the two goal components of the task are also presumed by us to operate either separately
(i.e., movement and appropriate postures) can be or in conjunction with other tasks, e.g., when
explained by each one having its own single eigen- initiating gait or rising on tiptoe (Crenna and Frigo,
movemment; the H-eigenmovement for the prime 1991); or, during low or high amplitude body
movement of bending per se, and the A-eigenmove- oscillations during fixating on an oscillating visual
ment for postural control by forward acceleration of target (Bardy et al., 1999).
24

How these two motion units are controlled by the (projects 02-04-48410a and 01-04-48924a) and the
CNS, including their relative timing, is still an open Russian Foundation of Humanity (project 00-06-
question. Is an internal model of the biomechanical 00242a).
characteristics (mass, inertia) of each of these
multijoint motion units stored at the cerebellar
level? This would be in keeping with the direct or Abbreviations
inverse dynamic model of Gomi and Kawato (1996),
Gomi and Kawato (1997) and Wolpert et al. (1998). A ankle
Is the control of these multijoint motion units CM center of mass
compatible with the equilibrium point hypothesis CNS central nervous system
and, more specifically, with the concept of referent CG center of gravity
posture as proposed by Feldman and Levin (1995) H hip
and Feldman et al. (1999)? The answering of these K knee
questions remains open in the absence of special
analytical techniques, like those of Domen et al.
References
(1999), or a model of the neuromuscular apparatus
similar to those developed for the human arm Alexandrov, A.V., Frolov, A.A. and Massion, J. (1998). Axial
(Flanagan et al., 1993; Gribble et al., 1998; Frolov synergies during human upper trunk bending. Exp. Brain
et al., 2000). Res., 118: 210–220.
Alexandrov, A.V., Frolov, A.A. and Massion, J. (2001a).
Biomechanical analysis of movement strategies in human
Summary and conclusions forward bending. I. Modeling. Biol. Cybern., 84: 425–434.
Alexandrov, A.V., Frolov, A.A. and Massion, J. (2001b).
The coordination between posture and movement Biomechanical analysis of movement strategies in human
results from two parallel controls. They operate on forward bending. II. Experimental study. Biol. Cybern., 84:
multijoint motion units (eigenmovements) of the 435–443.
Assaiante, C. and Amblard, B. (1993). Ontogenesis of head
overall biomechanical system responsible for both the
stabilization in space during locomotion in children: influence
movement and its associated postural adjustment(s). of visual cues. Exp. Brain Res., 93: 499–515.
An example is ankle and hip eigenmovements during Atkeson, C.G. (1989). Learning arm kinematics and dynamics.
voluntary forward bending of the trunk. In this Annu. Rev. Neurosci., 12: 157–183.
case, a critical aspect for the coordination of the two Babinski, J. (1899). De l’asynergie cérébelleuse. Rev. Neurol., 7:
eigenmovements is their timing, which is a function 806–816.
of their respective inertias. It is suggested that this Bardy, B.G., Marin, L., Stoffregen, T.A. and Bootsma, R.J.
(1999). Postural coordination modes considered as emergent
type of control might be a general rule for co-
phenomena. J. Exp. Psychol. Hum. Percept. Perform., 25:
ordinating posture and movement. It is likely that the 1284–1301.
cerebellum plays a critical role in this coordination, Baroni, G., Pedrocchi, A., Ferrigno, G., Massion, J. and
particularly in terms of the timing of eigen- Pedotti, A. (2001). Static and dynamic postural control in
movements. Recall, for example, that it has long long-term microgravity: evidence of dual adaptation. Am. J.
been known that the coordination between move- Physiol., 90: 205–215.
ment and posture is impaired in cerebellar patients Belenkiy, V.E., Gurfinkel, V.S. and Paltsev, E.I. (1967). On
elements of control of voluntary movements. Biofizica, 12:
(Babinski, 1899).
135–141.
Bernstein, N.A. (1967). The Coordination and Regulation of
Acknowledgments Movements. Pergamon Press, New York.
Berthoz, A. (1991). Reference frames for the perception and
control of movement. In: Paillard J. (Ed.), Brain and Space.
We thank Dr. Becky Farley, The University of Oxford University Press, Oxford, pp. 81–111.
Arizona, for her very effective help in reviewing an Bizzi, E., Kalil, R.E. and Tagliasco, V. (1977). Eye-head
early draft of this chapter. The work was supported co-ordination in monkeys: evidence for centrally patterned
by the Russian Foundation For Basic Research organization. Science, 1: 452–454.
25

Bizzi, E., Hogan, N., Mussa-Ivaldi, F.A. and Giszter, S. (1992). Feldman, A.G. and Levin, M.F. (1995). The origin and use of
Does the nervous system use equlibrium-point control to positional frames of reference in motor control. Behav. Brain
guide single and multiple joint movements?. Behav. Brain Sci., 18: 723–806.
Sci., 15: 603–613. Feldman, A.G., Archambault, P. and Lestienne, F.G. (1999).
Bouisset, S. and Zattara, M. (1987). Biomechanical study Multimuscle control is based on the specification of
of the programming of anticipatory postural adjustments referent body image. In: Gantchev G.S., Mori S. and
associated with voluntary movement. J. Biomech., 20: Massion J. (Eds.), Motor Control Today and Tomorrow.
735–742. Academic Publishing House ‘Prof. Marin Drinov’, Sofia,
Bouisset, S., Do, M.C. and Zattara, M. (1992). Posturo-kinetic pp. 163–179.
capacity assessed in paraplegics and parkinsonians. In: Flanagan, A.G., Ostry, D.J. and Feldman, A.G. (1993).
Woollacott M. and Horak F. (Eds.), Posture and Gait: Control of trajectory modification in target-directed reach-
Control Mechanisms, Vol. II, University of Oregon Press, ing. J. Mot. Behav., 25: 140–152.
Eugene, pp. 19–22. Forssberg, H. (1999). Neural control of human development.
Brown, T.G. (1911). The intrinsic factors in the act of pro- Curr. Opin. Neurobiol., 9: 676–682.
gression in the mammal. Proc. R. Soc. Lond. B. Biol. Sci., 84: Forssberg, H., Grillner, S. and Rossignol, S. (1975). Phase
308–319. dependant reflex reversal during walking in chronic spinal
Clément, G., Gurfinkel, V.S., Lestienne, F., Lipshits, M.I. and cats. Brain Res., 85: 103–107.
Popov, K.E. (1984). Adaptation of postural control to Frolov, A.A., Dufosse, M., Rizek, S. and Kaladjian, A. (2000).
weightlessness. Exp. Brain Res., 57: 61–72. On the possibility of linear modelling the human arm
Cordo, P.J. and Nashner, L.M. (1982). Properties of postural neuromuscular apparatus. Biol. Cybern., 82: 499–515.
adjustments associated with rapid arm movements. Gomi, H. and Kawato, M. (1996). Equilibrium-point control
J. Neurophysiol., 47: 287–302. hypothesis examined by measured arm stiffness during
Crenna, P. and Frigo, C. (1991). A motor program for the multijoint movement. Science, 272: 117–120.
initiation of forward oriented movement in man. J. Physiol. Gomi, H. and Kawato, M. (1997). Human arm stiffness and
(Lond.), 437: 635–653. equilibrium-point trajectory during multi-joint movement.
Crenna, P., Frigo, C., Massion, J. and Pedotti, A. (1987). Biol. Cybern., 76: 163–171.
Forward and backward axial synergies in man. Exp. Brain Gribble, P.L., Ostry, D.J., Sanguineti, V. and Laboissiere, R.
Res., 65: 538–548. (1998). Are complex control signals required for human arm
Dichgans, J., Held, R., Young, L. and Brandt, T. (1972). movement?. J. Neurophysiol., 79: 1409–1424.
Moving visual scenes influence the apparent direction of Gurfinkel, V.S. and Levik, Y.S. (1991). Perceptual and
gravity. Science, 178: 1217–1219. automatic aspects of the postural body scheme. In:
Dietz, V., Horstmann, G.A., Trippel, M. and Gollhofer, A. Paillard J. (Ed.), Brain and Space. Oxford University Press,
(1989). Human postural reflexes and gravity—an under water Oxford, pp. 147–162.
simulation. Neurosci. Lett., 106: 350–355. Gurfinkel, V.S. and Shik, M.L. (1973). The control of posture
Dietz, V., Gollhofer, A., Kleiber, M. and Trippel, M. (1992). and locomotion. In: Gydikov A.A., Tankov N.T. and
Regulation of bipedal stance: dependency on ‘load’ receptors. Kosarov D.S. (Eds.), Motor Control. Plenum Press, New
Exp. Brain Res., 89: 229–231. York, pp. 217–234.
Dijkstra, T.M.H., Schöner, G. and Gielen, C.C.A.M. (1994). Gurfinkel, V.S., Levik, Yu.S., Popov, K.E., Smetanin, B.N. and
Temporal stability of the action-perception cycle for postural Shlikov, Y. (1988). Body scheme in the control of postural
control in a moving visual environment. Exp. Brain Res., 97: activity. In: Gurfinkel V.S., Ioffé M.E., Massion J. and
477–486. Roll J.-P. (Eds.), Stance and Motion: Facts and Concepts.
Domen, K., Latash, M.L. and Zatsiorsky, V.M. (1999). Plenum Press, New York, pp. 185–193.
Reconstruction of equilibrium trajectories during whole- Gurfinkel, V.S., Lestienne, F., Levik, Yu.S. and Popov, K.E.
body movements. Biol. Cybern., 80: 195–204. (1993a). Egocentric references and human spatial orientation
Eng, J.J., Winter, D.A., Colum, D., MacKinnon, D. and in microgravity. I. Perception of spatial tactile stimuli. Exp.
Patla, A.E. (1992). Interaction of the reactive moments and Brain Res., 95: 339–342.
center of mass displacement for postural control during Gurfinkel, V.S., Lestienne, F., Levik, Yu.S., Popov, K.E. and
voluntary arm movements. Neurosci. Res. Commun., 11: Lefort, L. (1993b). Egocentric references and human spatial
73–80. orientation in microgravity. II. Body-centred coordinates in
Feldman, A.G. (1966). Functional tuning of the nervous system the task of drawing ellipses with prescribed orientation. Exp.
with control of movement or maintenance of a steady Brain Res., 95: 343–348.
posture. II. Controllable paramets of the muscle. Biophysics, Hasan, Z. and Enoka, R.M. (1985). Isometric torque-ankle
11: 565–578. relationship and movement-related activity of human elbow
26

flexors: implications for the equilibrium-point hypothesis. Macpherson, J.M. (1991). How flexible are muscle synergies?.
Exp. Brain Res., 59: 441–450. In: Humphrey D.R. and Freund H.J. (Eds.), Motor Control:
Head, H. (1920). Studies in Neurology. Hodder and Stoughton, Concepts and Issues. John Wiley, Chichester, pp. 33–37.
London. Magnus, R. (1924). Der Korperstellung. Springer Verlag,
Hess, W.R. (1943). Teleokinetisches und erreismatisches Berlin.
Kräftesystem in der Biomotorik?. Helv. Physiol. Pharmac. Massion, J. (1992). Movement, posture and equilibrium:
Acta, 1: C62–C63. interaction and co-ordination. Prog. Neurobiol., 38: 35–56.
Hlavacka, F., Kriskova, M. and Horak, F.B. (1995). Massion, J. (1994). Postural control system. Curr. Opin.
Modification of human postural response to leg muscle Neurobiol., 4: 877–887.
vibration by electrical vestibular stimulation. Neurosci. Lett., Massion, J. (1998). Postural control systems in developmental
189: 9–12. perspective. Neurosci. Biobehav. Rev., 22: 465–472.
Horak, F.B. and Macpherson, J.M. (1996). Postural orientation Massion, J., Popov, K., Fabre, J.C., Rage, P. and Gurfinkel, V.
and equilibrium. In: Towell L.B. and Shepherd J.T. (Eds.), (1997). Is the erect posture in microgravity based on the
Handbook on Integration of Motor Circulatory, Respiratory, control of trunk orientation or center of mass position?. Exp.
and Metabolic Control During Exercise. American Brain Res., 114: 384–389.
Massion, J., Amblard, B., Assaiante, C., Mouchnino, L. and
Physiological Society, Bethesda, pp. 255–292.
Vernazza, S. (1998). Body orientation and control of
Horak, F.B. and Nashner, L.M. (1986). Central programming
coordinated movements in microgravity. Brain Res. Rev.,
of postural movements: adaptation to altered support-surface
28: 83–91.
configurations. J. Neurophysiol., 55: 1369–1381.
Mergner, T. and Rosemeier, T. (1998). Interaction of vestibu-
Ito, M. (1984). The Cerebellum and Neural Control. Raven
lar, somatosensory and visual signals for postural control
Press, New York, pp. 580.
and motion perception under terrestrial and microgravity
Jeka, J.J. and Lackner, J.R. (1995). The role of haptic cues from
conditions—a conceptual model. Brain Res. Rev., 28:
rough and slippery surfaces in human postural control. Exp.
118–135.
Brain Res., 103: 267–276.
Miall, R.C., Weir, D.J., Wolpert, D.M. and Stein, J.F. (1993).
Kavounoudias, A., Roll, R. and Roll, J.P. (1998). The plantar
Is the cerebellum a Smith predictor?. J. Mot. Behav., 25:
sole is a ‘dynamometric map’ for human balance control.
203–216.
Neuroreport, 9: 3247–3252.
Mittelstaedt, H. (1998). Origin and processing of postural
Kavounoudias, A., Gilhodes, J.C., Roll, R. and Roll, J.P.
information. Neurosci. Biobehav. Rev., 22: 473–478.
(1999). From balance regulation to body orientation: two
Mori, S. (1989). Contribution of postural muscle tone to full
goals for muscle proprioceptive information processing?. expression of posture and locomotion movements: multi-
Exp. Brain Res., 124: 80–88. faceted analyses of its setting brainstem-spinal cord mechan-
Kawato, M., Furukawa, K. and Suzuki, R. (1987). A
isms in the cat. Jpn. J. Physiol., 39: 785–809.
hierarchial neural-network model for control and learning Mori, S., Matsuyama, K., Kuze, B. and Mori, F. (1999).
of voluntary movement. Biol. Cybern., 69: 169–185. Features of the fastigio-reticulo-spinal system involved in the
Kuo, A.D. and Zajac, F.E. (1993). A biomechanical analysis of control of posture and locomotion in the cat. In:
muscle strength as a limiting factor in standing posture. Gantchev G.N., Mori S. and Massion J. (Eds.), Motor
J. Biomech., 26 Suppl. 1: 137–150. Control Today and Tomorrow. Academic Publishing House
Kuypers, H.G.J.M. (1981). Anatomy of the descending path- ‘Prof. M. Drinov’, Sofia, pp. 31–43.
ways. In: Brooks V.B. (Ed.), Handbook of Physiology, Sec. Mouchnino, L., Aurenty, R., Massion, J. and Pedotti, A.
1, The Nervous System, Vol. II, Motor Control,. American (1992). Coordination between equilibrium and head-trunk
Physiology Society, Bethesda, pp. 597–666 Part 1. orientation during leg movement: a new strategy built up by
Lackner, J.R. (1988). Some proprioceptive influences on the training. J. Neurophysiol., 67: 1587–1598.
perceptual representation and orientation. Brain, 111: 281– Mouchnino, L., Aurenty, R., Massion, J. and Pedotti, A.
297. (1993). Is the trunk a reference frame for calculating leg
Latash, M. (1993) Control of Human Movement. Human position?. Neuroreport, 4: 125–127.
Kinetics, Champaign, IL, USA. Nashner, L.M. and McCollum, G. (1985). The organization of
Lestienne, F., Soechting, J. and Berthoz, A. (1977). Postural human postural movements: a formal basis and experimental
readjustment induced by linear motion of visual scenes. Exp. synthesis. Behav. Brain Sci., 8: 135–172.
Brain Res., 28: 363–384. Oddsson, L. and Thorstensson, A. (1986). Fast voluntary trunk
Lund, S. and Broberg, C. (1983). Effects of different head flexion movements in standing: primary movements and
positions on postural sway in man induced by a reproducible associated postural adjsutments. Acta Physiol. Scand., 128:
vestibular error signal. Acta Physiol. Scand., 117: 307–309. 341–349.
27

Paillard, J. (1960). The patterning of skilled movements. In: Sternad, D. and Corcos, D. (2001). Effect of task and
Field J., Magoun H.W. and Hall V.E. (Eds.), Handbook of instruction on patterns of muscle activation: wacholder and
Physiology, Neurophysiology, Vol. III, American beyond. Motor Control, 5: 307–336.
Physiological Society, Bethesda, pp. 1679–1708. Stuart, D.G. (2002). Reflections on spinal reflexes. Adv. Exp.
Paillard, J. (1991). Motor and representational framing of Biol. Med., 508: 249–257.
space. In: Paillard J. (Ed.), Brain and Space. Oxford Thomas, A. (1940). Equilibre et Equilibration. Masson, Paris.
University Press, Oxford, pp. 163–184. Tokizane, T., Murao, M., Ogata, T. and Kondo, T. (1951).
Pedotti, A., Crenna, P., Frigo, C. and Massion, J. (1989). Electromyographic studies on tonic neck, lumbar and
Postural synergies in axial movements: short and long-term labyrinthine reflexes in normal persons. Jpn. J. Physiol., 2:
adaptation. Exp. Brain Res., 74: 3–11. 130–146.
Pozzo, T., Levik, Y. and Berthoz, A. (1995). Head and Vernazza-Martin, S., Martin, N. and Massion, J. (2000).
trunk movements in the frontal plane during complex Kinematic synergy adaptation to microgravity during for-
dynamic equilibrium tasks in humans. Exp. Brain Res., ward trunk movement. J. Neurophysiol., 83: 453–464.
106: 327–338. Wachholder, K. (1928). Willkürliche Haltung und Bewegung.
Prochazka, A., Clarac, F., Loeb, G., Rothwell, J.C. and Ergeb. Physiol., 26: 568–775.
Wolpaw, J.R. (2000). What do reflex and voluntary mean? Weiss, P. (1961). Self differentiation of basic patterns of
Exp. Brain Res., 130: 417–432.
coordination. Comp. Psychol. Monogr., 17: 1–96.
Rademaker, G.G.J. (1931). Das Stehen: Statische Reactionen,
Wiesendanger, M. (1997). Paths of discovery in human motor
Gleichwichtsreaktionen und Muskeltonus unter Besondere
control. In: Hepp-Reymond M.-C. and Marini G. (Eds.),
Berucksichtung Ihres Verhaltens Bei Kleinhirnlosen Tieren.
Perspectives of Motor Behavior and Its Neural Basis. Karger,
Springer Verlag, Berlin.
Basel, pp. 103–134.
Robinson, D.A. (1986). The systems approach to the oculo-
Winter, DA. (1990). Biomechanics and Motor Control in
motor system. Vision Res., 1: 91–99.
Human Movement, 2nd ed., John Wiley and Sons, New
Roll, J.P. and Roll, R. (1988). From eye to foot: a
proprioceptive chain involved in postural control. In: York.
Amblard B., Berthoz A. and Clarac F. (Eds.), Posture and Wise, S.P., Alexander, G.E., Altman, J.S., Brooks, V.B.,
Gait: Development Adaptation and Modulation. Elsevier, Freund, H.-J., Fromm, C.J., Humphrey, D.R., Sasaki, K.,
Amsterdam, pp. 155–164. Strick, P.L., Tanji, J., Vogel, S., Wiesendanger, M. (1991).
Roll, J.-P., Popov, K., Gurfinkel, V., Lipshits, M., André- What are the specific functions of the different motor areas?
Deshays, C., Gilhodes, J.C. and Quoniam, C. (1993). Humphrey D.R. and Freund H.-J. (Eds.), Motor Control:
Sensorimotor and perceptual function of muscle propriocep- Concepts and Issues. John Wiley & Sons, New York,
tion in microgravity. J. Vestibul. Res., 3: 259–273. pp. 463–485.
Roll, R., Gilhodes, J.C., Roll, J.P., Popov, K., Charade, O. and Wolpert, D.M., Ghahramani, Z. and Jordan, M.A. (1995). An
Gurfinkel, V. (1998). Proprioceptive information processing internal model for sensori-motor integration. Science, 269:
in weightlessness. Exp. Brain Res., 122: 393–402. 1880–1882.
Sherrington, C.S. (1906). The Integrative Action of the Nervous Wolpert, D.M., Miall, R.C. and Kawato, M. (1998). Internal
System. Constable, London. models in the cerebellum. Trends Cogn. Sci., 2: 338–347.

View publication stats

Vous aimerez peut-être aussi