You are on page 1of 52

Physiol Rev 94: 655–706, 2014

doi:10.1152/physrev.00009.2013

CORTICAL MECHANISMS UNDERLYING THE


ORGANIZATION OF GOAL-DIRECTED ACTIONS AND
MIRROR NEURON-BASED ACTION
UNDERSTANDING
Giacomo Rizzolatti, Luigi Cattaneo, Maddalena Fabbri-Destro, and Stefano Rozzi

Department of Neuroscience, University of Parma, Parma, Italy; Center for Mind/Brain Sciences, University
of Trento, Trento, Italy; and Brain Center for Motor and Social Cognition, Italian Institute of Technology,
Parma, Italy

Rizzolatti G, Cattaneo L, Fabbri-Destro M, Rozzi S. Cortical Mechanisms Underly-

L
ing the Organization of Goal-Directed Actions and Mirror Neuron-Based Action Under-
standing. Physiol Rev 94: 655–706, 2014; doi:10.1152/physrev.00009.2013.—
Our understanding of the functions of motor system evolved remarkably in the last 20
years. This is the consequence not only of an increase in the amount of data on this
system but especially of a paradigm shift in our conceptualization of it. Motor system is not
considered anymore just a “producer” of movements, as it was in the past, but a system crucially
involved in cognitive functions. In the present study we review the data on the cortical organization
underlying goal-directed actions and action understanding. Our review is subdivided into two major
parts. In the first part, we review the anatomical and functional organization of the premotor and
parietal areas of monkeys and humans. We show that the parietal and frontal areas form circuits
devoted to specific motor functions. We discuss, in particular, the visuo-motor transformation
necessary for reaching and for grasping. In the second part we show how a specific neural
mechanism, the mirror mechanism, is involved in understanding the action and intention of others.
This mechanism is located in the same parieto-frontal circuits that mediate goal-directed actions.
We conclude by indicating future directions for studies on the mirror mechanism and suggest
some major topics for forthcoming research.

I. INTRODUCTION 655 uli that may satisfy their immediate or future needs. When
II. VISUOMOTOR BEHAVIOR: GENERAL... 656 they find something appropriate for their need, they might
III. THE ANATOMICAL ORGANIZATION... 657 interact with it. In voluntary behavior stimuli, however, do
IV. FUNCTIONAL ORGANIZATION OF... 662 not determine the response, they only set the occasion for it.
V. THE MIRROR NEURON... 671 Therefore, according to their needs, animals may respond
VI. MIRROR NEURON SYSTEM IN... 672 to the same stimulus in different ways. They may approach
VII. THE MIRROR NEURON SYSTEM IN... 678 it, avoid it, or even ignore it.

In this review we address two action-related issues. The first


I. INTRODUCTION
is how animals are able to organize object-directed actions.
The distinct property that characterizes animals’ behavior is We will first briefly discuss this issue examining the behav-
their capacity to generate movements. Some movements are ior of species with relatively poor corticalization (frogs and
determined by internal or external stimuli. They are called then rodents). Next, we examine how primates, a species
reflexes. Others are manifestations of a centrally generated endowed with a complex brain, solve the same problems.
intention to act. They are called voluntary actions. Inten- Our aim is not to give a complete picture of the neurophys-
tion to act might originate in several ways. It might origi- iological process underlying the visuomotor transformation
nate from bodily needs or from higher order deliberations. that leads from object to action, but rather to outline the
In voluntary actions what counts is that the individual has general principles underlying this behavior.
the intention to achieve an overarching goal, and this over-
arching goal determines a series of movements leading to its In the second part of this review we examine the role that
achievement. the cortical motor system plays in understanding the behav-
ior of others. Although the notion that there is a strict link
Because of their capacity to generate movements volun- between action execution and action perception is not new
tarily, animals move in their environment looking for stim- (356), the notion that there is a mechanism, the mirror

0031-9333/14 Copyright © 2014 the American Physiological Society 655


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

mechanism, located in the motor system that plays a fun- the chapter on Motor System with the following sentence:
damental role in action understanding has represented a “The motor systems of the brain exist to translate thought,
paradigmatic conceptual shift in cognitive neuroscience and sensation and emotion into movement. At present the initial
psychology. As nicely summarized by Jeannerod: “. . . the steps of this process lie beyond analysis. We do not know
observation of actions performed by other agents generates how voluntary movements are engendered nor where the
in the brain of the observer representations similar to those ‘orders’ come from.” This sentence expresses with clarity
of the agents. This circular process, from the self to action and elegance the difficulty to understand how the motor
and from action to other selves, has as a consequence that system could organize motor behavior. In fact, accepting
action representations can be shared by two or more people. the standard view on cortical organization of that time, the
These new findings have radically changed the traditional visual and motor systems were two incommunicado worlds
view of the motor system as an executive system that merely devoid of any clear functional link. What was the standard
follows instructions elaborated somewhere else. Motor sys- view about vision? The view was that vision provides the
tem now stands as a system that allows understanding the observer with a global perceptual representation of the ex-
behavior of others and even as a probe that explores the ternal word. This representation is unitary, precedes motor
external world for interacting with other people and gath- activity, and is used for organizing all types of actions from
ering new knowledge” (184). In TABLE 1, we specify the gazing to walking, from escaping to interacting with ob-
precise “technical” meaning of the terms that we use in this jects. This view of the visual system was radically wrong. By
review to describe the different elements forming motor accepting it, it is indeed impossible to understand how ac-
actions and their comprehension (goal and intention). tions are organized. In the next paragraphs we will show
where the error stood.
II. VISUOMOTOR BEHAVIOR:
GENERAL PRINCIPLES In 1973 Ingle (177) published a series of fundamental stud-
ies on frogs’ vision. Ingle studied two frogs’ basic visuomo-
In one of the most authoritative textbooks of physiology of tor behaviors: 1) turning and snapping at small prey-like
the end of the last century, The Medical Physiology, edited objects, and 2) jumping away from large looming discs. The
by Vernon Mountcastle, Elwood Henneman (166) started experiments were carried out before and after removal of
the optic tectum, the most important visuomotor integration
center in amphibians. Following unilateral removal of the op-
tic tectum, frogs totally ignored stimuli presented in the field
Table 1. Terms used to describe motor actions and contralateral to the ablated tectum. However, over the next
comprehension
month, vision recovered and the frog started to turn and snap
Terms and Meanings in response to prey-like objects presented in the contralateral
formerly blind visual field and to react in response to threat-
Execution ening stimuli shown in that field. However, the observed re-
Movement: Displacement of a body-part regardless of the sponses were quite bizarre. When presented with prey-stimuli
intended goal (e.g., movement evoked by electrical
stimulation of the motor cortex).
in the formerly blind contralateral field, the frogs, instead of
Motor act: A series of body-part movements that allow an
turning and snapping towards them, directed their re-
individual to reach a goal (e.g., grasping an object). sponses toward locations in the ipsilateral field. Similarly,
Motor action: A series of motor acts (e.g., reaching, when a large threatening disc was suddenly introduced into
grasping, bringing to the mouth) that allow an individual the field opposite to the lesion, the frogs jumped towards it,
to fulfill their intention (e.g., eating).
instead of avoiding it (FIGURE 1).
Observation
Understanding the goal of a motor act done by another
individual implies the understanding of “what” the other is What had happened? Unlike in mammals, in frogs the pro-
doing (e.g., grasping an object). jections from the retina to the brain regrow after transec-
Understanding a goal of a motor action done by another tion. In agreement with this well-known fact, histological
individual implies the understanding of “why” the other is analysis revealed that the transected optic tract regenerated.
doing it, that is the action overarching intention (e.g.,
grasping for eating). However, its sprouting axons, finding no tectum on their
A further level, possibly not related to the mirror brain side, crossed the midline and innervated the opposite
mechanism, is the understanding of the reasons tectum. The behavior of the “rewired” frogs is now easy to
underlying the observed motor action (e.g., eating explain. When stimuli were shown to the left eye, they elic-
because hungry or for other reasons).
Mental rehearsal
ited the behavior proper of the left tectum, rather than of
An attempt to replicate internally a motor act or a motor
right tectum as normally occurs, hence the paradoxical re-
action either spontaneously or under instruction. sponses.
Table includes the “technical” meaning of the terms used in this
review to describe the differenct elements forming motor actions The interpretation that following surgery the entire percep-
and their comprehension (goal and intention). tual world of the rewired frog was inverted with the percep-

656 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

C Geniculate

A
Pretectum
B
C
Tectum

FIGURE 1. Responses to visual stimuli and organization of the retino-collicular pathways in a “rewired” frog.
Left: circles represent the location of prey/objects; crosses the mirror position frog’s snaps in the opposite
visual field to the corresponding stimuli. Right: diagram of the “rewired” projections (interrupted lines) from the
retina of the left eye to the ipsilateral tectum. The right eye and the contralateral tectum were ablated. Inputs
from loci A, B, and C on the retina cross the contralateral tectum, their target in a normal frog, and project
to sites in the ipsilateral tectum that would normally receive input from corresponding points in the retina of the
right eye. When the left tectal sites are activated by the stimulation of the left eye, they direct responses to
contralateral locations, as if activated by the right eye. [Modified from Ingle (177). Reprinted with permission
from AAAS.]

tion remaining unitary was rued out by a subsequent exper- (267). These authors showed a striking homology in neural
iment. A barrier was placed in front of a rewired frog. The architecture between amphibians and rodents. They sug-
frog was touched on the rear and escaped by jumping away. gested that a series of independent parallel visuomotor cir-
In contrast to their paradoxical behavior shown with prey- cuits is an evolutionary ancient characteristic of all verte-
stimuli and looming discs, the frogs jumped correctly, as brate brains.
normal frog, and avoided to hit the barrier placed in front of
them. Such behavior was observed immediately after the The encephalon of primates, and humans in particular, is
unilateral tectum ablation, when the animals still neglected characterized by the enormous development of the cerebral
the contralateral stimuli. cortex. This development correlates with an extraordinary
increase of sensory, motor, and cognitive capacities. A large
The dissociation between the behaviors just described is due part of the evolutionary new primate cerebral cortex is
to the different neural substrates that mediate them. In the formed by areas traditionally called “association areas.”
case of prey-catching and looming disk avoidance, the be- The classical view of the functional role of the association
havior is mediated by the tectum, while in the case of loco- areas (posterior parietal lobe and the inferotemporal cor-
motion around barrier, it is the pretectal system that medi- tex) was that they “put together” (associate) information
ates it. These findings show that what counts in the organi- coming from different sensory modalities. The results of
zation of the visuomotor responses is the neural pathway these associations are percepts. As will be shown in the next
linking the visual input to the motor centers and not the paragraphs, these conclusions are simplistic. In the case of
“perception” of spatial location of the stimulus. There is no parietal lobe, they misrepresent the neurophysiological or-
such thing as a unified representation of the external word. ganization of the posterior parietal cortex, while in the case
The frogs possess a parallel set of independent visuomotor of inferotemporal lobe, they jump to conclusions far be-
circuits that are activated by specific stimuli present in the yond the established facts.
environment and determine the appropriate responses.
What counts for behavior is not space “perception” but the
activity of these motor centers. III. THE ANATOMICAL ORGANIZATION OF
THE MOTOR CORTEX OF PRIMATES
The organization of the visual system in parallel visuomotor
circuits is not a feature unique to amphibians. Other verte- The first detailed cytoarchitectonic map of human cerebral
brate classes show a similar organization. For many years, cortex was published by Campbell in 1905 (59). In his map
however, the evidence was rather limited in mammals be- he distinguished the cortex located immediately in front of
cause studies on vision were typically formulated in these the central sulcus (“precentral” cortex) from the cortex lo-
species in terms of perception and cognition, ignoring the cated in front of it and caudal to the prefrontal lobe (the
visual control of motor output. A notable exception was the “intermediate precentral cortex”). According to Campbell,
work of Goodale and co-workers on rats (151) and gerbils the precentral cortex is directly involved in motor control,

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 657


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

while the intermediate sector is involved in higher order 6 (mesial, dorsal, and ventral) each results to be formed by
motor functions. Subsequently, Brodmann in 1909 (44) a caudal and a rostral subdivision.
identified two motor areas (area 4 and area 6), both char-
acterized by an agranular structure. The idea that this ar-
chitectonic organization reflects functional differences was B. Classification of Frontal Motor Areas and
advanced by Fulton (126) who showed that the ablation of Their Cortico-cortical and Descending
area 6 (defined as “premotor cortex”) in primates produces Connections
specific deficits of skilled movements.
The brain regions, with which the frontal motor cortex is
The existence of a higher order motor area located rostral to mostly connected, are as follows: the parietal cortex, the
area 4 was challenged by electrophysiological studies em- prefrontal lobe, and cingulate cortex (334). The output
ploying surface electrical stimulation. Woolsey et al. (406) from the parietal cortex is the major source of input to the
identified a somatotopic map on the lateral surface of the primary motor area and the caudal premotor areas (F3, F4,
agranular frontal cortex and stated that area 4 and the F5c, and F5p). The posterior parietal cortex, similarly to the
caudal part of area 6 form a functional entity: the primary motor cortex, is formed by a mosaic of areas (FIGURE 3),
motor cortex (M1). The same authors identified a second each of which deals with specific aspects of sensory infor-
complete somatotopic motor representation located on the mation and with the control of specific effectors (mouth,
mesial sector of area 6: the “supplementary motor area” hand, arm, and eyes). The areas of the inferior parietal
(SMA). Finally, the rostral part of area 6, whose electrical lobule (IPL) and the posterior areas of the superior parietal
stimulation did not produce body movements, was consid- lobule (SPL) process visual and somatosensory informa-
ered not to be part of the motor cortex. Woolsey et al. (406) tion, whereas the rostral sectors of both lobules, SPL in
wrote “no single cytoarchitectural area of any worker co- particular, deal mostly with somatosensory information
incides with the extent of our precentral field.” (58, 173, 332, 344, 403).

More recently, neuroanatomical research has significantly The parietal and motor areas are strictly linked one with
boosted the study of the organization of motor cortex. This another (see Ref. 329). It is, therefore, possible to identify
led to the notion that in both monkeys and humans the the existence of a large number of cortical circuits formed
motor cortex is formed by several distinct areas, each of by parietal and motor areas. Typically, these circuits link
them connected in a specific manner with other cortical SPL with dorsal and mesial premotor areas, and IPL with
areas. In the next sections, we will present these data. ventral premotor areas. Functional evidence, when avail-
able, indicates that parietal and frontal areas forming each
of these circuits have common properties. Thus each of
A. The Agranular Frontal Cortex in the these circuits is specifically involved in particular aspects of
Monkey sensory-motor transformation and represents the func-
tional units of the cortical motor system.
The monkey agranular frontal cortex (motor cortex) is lo-
cated between the fundus of the central sulcus and that of As far as the visual input is concerned, the parieto-frontal
the arcuate sulcus. According to Brodmann (44), it consists circuits can be grouped into two major pathways: a dorsal
of two areas: area 4 and area 6. This oversimplified subdi- one, linking SPL with the dorsal premotor cortex, and a
vision of motor cortex was criticized by several authors, ventral one, linking IPL with the ventral premotor cortex
e.g., Vogt and Vogt in 1919 (396) and von Bonin and Bailey (335). The input to SPL and IPL originates mostly from two
in 1947 (397), who all found a much more complex map of distinct extrastriate visual areas: area V6 and MT, respec-
the motor cortex (FIGURE 2). Modern studies, which com- tively (12, 81, 131, 229, 343). On this basis, it has been
bined the classic architectonic methods with new histolog- proposed that the classical dorsal stream should be subdi-
ical techniques (e.g., neurochemistry, immunohistochemis- vided into two separate streams: the dorso-dorsal involved
try, receptor autoradiography, etc.), confirmed that in the in visuomotor transformation for online control of move-
agranular frontal cortex there is multiplicity of anatomical ments, and the ventro-dorsal involved in the organization of
distinct areas. goal-directed motor act and in various cognitive functions
(335). In this respect, note that, IPL, but not SPL, is con-
A modern map of the monkey motor cortex based on the nected with polymodal areas of the superior temporal sul-
works of Matelli and co-workers (31, 259, 260) is shown in cus (STS) and with the inferotemporal cortex where seman-
FIGURE 3. These authors, in analogy with von Economo and tic information on objects is encoded (37, 229, 282, 343,
Koskinas (398), named the different motor and premotor 358).
areas using the letter F (“frontal”) and consecutive Arabic
numbers (F1-F7). In this map, F1 indicates the primary Prefrontal projections to the agranular frontal cortex are
motor cortex (or M1, roughly corresponding to area 4), primarily directed to the rostral premotor areas: F6, F7 and
while the three gross anatomical sectors of Brodmann’s area F5a (23, 145, 146, 236, 237, 315, 334, 349). Prefrontal

658 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

Brodmann 1903 Vogt and Vogt 1919

4 4a
6 6aa
Cg PO 6ab Cg
PO
C
IP 4a
6 6aa
4 C
6ab
AS L Lu 4b
IP
P AS Lu
AI ST 4c
6 P
IO AI 6aa
L

6a bb
b
6
ST

von Bonin and Bailey 1947 Barbas and Pandya 1987

6DC
FA MII
FB
FC Cg
Cg
PO
PO
C
4 C

6DC
FC FB 6DR
FA IP AS IP Lu
4C
FB P 6Va
P A L AI L
FC 6Vb
Bm
AI Lu
ST IO
IO ST

FIGURE 2. Lateral and mesial views of the monkey brains showing classical maps of the agranular frontal
cortex. Dashed lines indicate borders between different areas. Dotted lines mark the transitions between
different areas according to von Bonin and Bailey. AI, inferior arcuate sulcus; AS, superior arcuate sulcus; C,
central sulcus; Cg, cingulated sulcus; IO, inferior occipital sulcus; IP, intraparietal sulcus; L, lateral fissure; Lu,
lunate sulcus; P, principal sulcus; PO, parieto-occipital sulcus; ST, superior temporal sulcus. [Modified from
Matelli et al. (259), with kind permission from Springer Science and Business Media.]

input to F7, except for the supplementary eye field (SEF), premotor areas have access to the spinal cord and can be
originates only from the dorsal part of the prefrontal cortex thus involved in the generation and control of movements,
(DLPF), while afferences to F6 and the SEF originate from both through F1, or in parallel with it.
both DLPF and the ventral (VLPF) part of the prefrontal
cortex; finally, input to F5a originates only from VLPF. In The descending output of the rostral premotor areas termi-
addition, F6 and the SEF are the targets of strong afferents nates in various parts of the brain stem, but not in the spinal
originating from the rostral cingulate cortex (area 24c). cord (196). These areas are not directly connected with F1,
but are richly connected with other motor areas. The only
The organization of the corticospinal and corticobulbar exception is the dorso-rostral part of F7 (SEF) (352). This
projections is in agreement with the subdivision of the pre- oculomotor area is strongly linked with the frontal eye field
motor areas into rostral and caudal areas. Strick and co- (FEF) (351). These data indicate that rostral premotor areas
workers (103, 162, 163) showed that the corticospinal pro- could play a role in the generation of motor behavior only
jections originate almost exclusively from the primary mo- indirectly, through their subcortical projections or their
tor area and caudal premotor areas. It is also of interest to connections with the caudal premotor areas.
note that each caudal premotor area displays somatotopi-
cally organized connections with each other and with F1. Taken together, these data indicate a different functional
Although F1 is the only sector directly connected to the role for caudal and rostral premotor areas. The former
bulbar and spinal motor neuron pools (312), all caudal transform sensory information into potential motor acts;

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 659


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

PE
F1 PEc
F3
F6 PEcg
Cg 24d PGm
PO

PEc
F1 PE V6A Ca FIGURE 3. Lateral and mesial view of the monkey
F2 MIP brain showing a modern detailed anatomical and func-
F7 LIP
PEip Opt tional parcellation of the agranular frontal cortex and
AS the posterior parietal cortex. Intraparietal, arcuate, and
VIP AIP PG
DLPF IP cingulated sulci are shown unfolded. For the nomencla-
P F4 PFG ture and definition of the areas of agranular frontal and
C PF posterior parietal cortices, see text. Ca, calcarine fis-
VLPF F5c
Lu sure; DLPF, dorsolateral prefrontal cortex; VLPF, ven-
F5p
trolateral prefrontal cortex; other nomenclature are as
F5a in FIGURE 2.
AI

L
IO
ST

the latter are involved in the control of movements at the large pyramids in layer Vb. F5c is characterized by a poorly
higher level. More specifically, the rostral areas receive in- laminated appearance, resulting from an overall homoge-
formation concerning agent’s goal, motivation and spatial, neity in cell population. Finally, F5a is characterized by the
and non-spatial information from the prefrontal, cingulate, presence of relatively large pyramids in lowest layer III and
and opercular frontal cortex. This information is then con- a prominent and homogeneous layer V (FIGURE 4). Area F6,
veyed to the caudal premotor areas determining when and unlike areas F1 and F3, is clearly laminated and has a dark
which potential motor acts, coded in these areas, might and very prominent layer V, well distinct from the less dense
become actual actions. layers III and VI. Area F7 is clearly laminated and differs
from F2 in two aspects: the presence of a prominent layer V
C. Architectonics of Macaque and Human and the subdivision of layer VI into two parts. Subsequent
cytoarchitectonic studies confirmed and extended the
Motor Cortex
knowledge on the anatomical structure of this cortical re-
The cytoarchitetonic characteristics of the areas forming gion.
the monkey motor cortex have been extensively studied.
Here are their main features. Area F1 has a low cell density, Cytoarchitectonic features of the human motor cortex are
poor lamination, absence of layer IV, and a very prominent similar to that of the macaque monkey. Low cell density,
layer V with giant pyramidal cells arranged in multiple rows poor lamination, absence of layer IV, and giant pyramidal
(259). Area F2 is poorly laminated but differently from F1, cells (Betz cells) in layer V characterize the precentral motor
contains only few scattered giant pyramidal cells. Lower cortex (Brodmann area 4) also in human brain, while scat-
layer III is characterized by medium-sized pyramids distrib- tered or absent giant pyramidal cells and densely packed
uted in a thin row. Layer V is denser than in F1. Area F3 is pyramids in lower layer III characterize Brodmann area 6.
also poorly laminated, but shows an increase in cellular Close to the midline, the rostral border of area 4 lies on the
density in the lower part of layer III and in layer Va. Giant cortical convexity. Moving ventrally and laterally, this bor-
pyramidal cells are almost absent. The ventral premotor der moves in caudal direction and eventually ends in the
cortex consists of two areas: area F4 and area F5. In area F4, depth of the central sulcus.
the overall cell density is low, and pyramid size tends to
increase from superficial to deep layer III. In layer V, large The mesial cortical surface is cytoarchitectonically similar
pyramidal cells are scattered especially in its dorsal part, in humans and macaques. In particular, Zilles and co-work-
near the spur of the arcuate sulcus. Area F5 is clearly lami- ers (412, 413) found that areas F3 and F6 and human areas
nated and shows a cell density higher than F4. Area F5 has 6a␣ and 6a␤ of Vogt and Vogt are strikingly similar (396).
been recently subdivided into three sectors: F5p, F5a, and Other similarities have been found on neurochemical bases
F5c (31). F5p is characterized by a relatively homogeneous (147, 412, 413). Altogether, these data strongly suggest that
layer III, a cell dense layer Va, and the presence of relatively human mesial areas 6a␣ and 6a␤ are homologous to ma-

660 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

FIGURE 4. Architectonic features of area F5. A: a low-power photomicrograph from a Nissl-stained para-
sagittal section, centered on the precentral gyrus. Arrows mark the borders between architectonic areas.
Scale bar, shown in A, applies also to B. B: a low-power photomicrograph of an SMI-32-immunostained
parasagittal section centered on the precentral gyrus and the whole extent of the inferior arcuate sulcus.
C: dorsal view of a monkey brain. Horizontal lines indicate the approximate location of the sections shown in A and
B. D: higher magnification views of representative fields of F5p (top row) and F5a (bottom row) showing the
cytoarchitecture (right) and the laminar distribution pattern of SMI-32 immunoreactivity (left). In SMI-32-
stained sections, the borders between layers are reported from an adjacent Nissl-stained section. Scale bar,
shown in the top left photomicrographs, applies to all fields. Nomenclature is as in FIGURES 2 AND 3.
[Modified from Belmalih et al. (31), with permission from Wiley.]

caque areas F3 (SMA proper) and F6 (pre-SMA), respec- connections (353, 382). PMd is connected with superior
tively. The border between the primary motor cortex and parietal cortex, dorsal prefrontal cortex, and cingulate cor-
area SMA proper (F3) coincides in humans approximately tex; PMv with inferior parietal cortex and ventral prefron-
with the vertical plane through the posterior commissure tal cortex.
(VCP line), while the border between area pre-SMA and
SMA proper coincides with the vertical plane through the Taken together, architectonic studies of the human agranu-
anterior commissure (VCA line) (412). The border between lar frontal cortex show that this cortical region is formed by
SMA proper and pre-SMA as well as its position at the a mosaic of architectonic areas. Spatial congruence between
antero-posterior level of the VCA line has been confirmed architectonic borders and macroscopic landmarks in the
by means of diffusion magnetic resonance imaging studies human cortex is less clear and less consistent than in the
(85, 189, 219). macaque. This variability is a factor that has to be taken
into consideration in the interpretation of activation foci in
The dorsolateral part of the human premotor cortex is functional imaging studies. Classical cytoarchitectonic
formed by two major sectors: the dorsal (PMd) and ventral maps, based on observations of one or few brains, do not
(PMv) premotor cortex, each showing a specific pattern of provide information on interindividual variability. Recent

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 661


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

probabilistic maps solve in part the problem, allowing a produces a population vector that corresponds closely to
more reliable comparison between the observed activations the actual movement direction (143, 192).
and their architectonic localization (see Ref. 11).
One of the oldest ways to investigate the functional role of
an anatomical area is to evaluate the deficits occurring in
IV. FUNCTIONAL ORGANIZATION OF specific aspects of behavior following brain damage. In
ACTION EXECUTION IN PRIMATES monkeys, lesions involving the internal capsule determine
spastic paralysis. In contrast lesions confined to the pyrami-
dal tract or to the primary motor area produce flaccid pa-
A. Movement Execution: Areas F1 and F3
resis (218, 383; for a review, see Refs. 223, 224). Most
deficits tend to compensate over time, even if skilled hand
1. Area F1 functions never completely recover.

The primary motor cortex of monkey (area F1) is a nodal 2. Area F3


point of convergence of inputs from the caudal premotor
and parietal areas, and plays a major role in movement Similarly to area F1, area F3 (or SMA proper) is electrically
execution. In particular, F1 is fundamental for the control excitable with low-intensity currents and contains a com-
of independent finger movements (see Ref. 312), a function plete motor map of the contralateral body. Some important
that relies on direct cortico-motoneuronal projections, orig- differences distinguish, however, the two areas. In F1, prox-
inating from F1 (223, 224, 325). The electrical stimulation imal and distal movements are anatomically segregated,
of F1 shows a somatotopic organization: a leg field is lo- while the activity of most F3 neurons is less strictly coupled
cated on the medial cortical surface and extends on the to specific body parts and instead appears to be associated
dorsolateral convexity, an arm field lies in an intermediate with coordinated movements of the hand, arm, head, or
position, and a face field occupies a ventrolateral position. trunk (238). Furthermore, electrical stimulation typically
Within the arm field, proximal movements tend to be rep- evokes single joint movements in F1, while two or more
resented rostrally and distal movements caudally, in the joints are most often observed after stimulation of F3 (238,
anterior bank of the central sulcus. 265). Finally, in F3, the proximal movements are much
more represented than the distal ones (238, 303, 406). In-
Evarts (108) found that single neurons in this area typically terestingly, a subpopulation of F3 neurons code specific
became active before the onset of muscle electric activity. sequences of movements (375), suggesting a role of this area
Asanuma and Rosén (16) proposed that the primary motor in the sequential organization of learned multiple move-
cortex is organized in terms of cortical columns, each of ments (377).
which related to a single muscle. A different organization
was described by Kwan and co-workers (216, 277). They The functional role of area F3 appears to be that of control-
found a “core and surround” (horseshoe-shaped) move- ling motor activity in a global way. Furthermore, this area
ment representation in M1. Digit movements were repre- plays a role in predictive postural adjustments. With the use
sented in the “core,” located mostly within the bank of the of information about the body-parts position in a given
central sulcus; wrist, elbow, and shoulder movements were moment, it prepares the posture necessary for the impend-
located in the “surround” where they formed parallel bands ing movements (256). F3 is also involved in learning motor
on the cortical convexity. Other studies described a “frac- sequences (168).
tured” somatotopy. According to them, different but over-
lapping neuronal populations encode arm, hand, and digit
movements. There is now general agreement that the pri- B. Planning and Execution of Arm
mary motor cortical organization is in term of movements Movements: Area F2
(and motor acts) and not in term of muscles and joints
(347). Planning an action implies the necessity to decide which
movements to execute to achieve a specific goal. Numerous
A population of neurons recorded in the motor cortex dis- studies, in particular of the dorsal premotor cortex, have
charges during reaching arm movements (143). The activity attempted to identify the neural substrate underlying move-
of each neuron is strongest for a specific direction (preferred ments planning and execution. This was generally done by
direction) and weakest for the opposite one. Furthermore, inserting a delay between the instruction about what move-
its activity pattern describes a broad directional tuning ment (usually arm movements) to perform and the cue to
curve, decreasing gradually from the preferred to the oppo- actually execute it (84, 403). The results showed that none
site direction. Different cells have different preferred direc- of the studied cortical areas (including area F1 and area F2)
tions. Numerous neurons with different preferred direc- hosts a homogeneous population of neurons dedicated only
tions discharge at different intensities during each move- to planning or execution. Neurons, devoted to this func-
ment, and the vectorial addition of all neurons contribution tion, are present in both areas and in particular in F2.

662 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

Area F2 is an electrically excitable area showing a rough particular effector. The sense of space arises from our motor
somatotopic organization. The leg is represented dorsal to interactions with the world (see Refs. 242, 311).
the superior precentral dimple and the arm ventral to it
(103, 162, 214, 322). Numerous F2 neurons show somato- Lesion studies and single neuron recordings are consistent
sensory responses, mainly to proprioceptive input. Neurons with this last view. They indicate the existence of two main
in its ventro-rostral sector (F2vr) also respond to visual types of space: the space within our arm reach (“periper-
stimuli (322, 323). One of the most interesting finding sonal space”) and a space outside it (“extrapersonal
about F2 neurons concerns the feature of their discharge. space”). Ablation of PMv in the monkey determines a syn-
Wise and co-workers (403) distinguished three types of neu- drome characterized by the lack of awareness of the con-
rons: 1) signal-related neurons, discharging at the appear- tralateral side of the body and of the space around it
ance of the instruction stimulus; 2) set-related neurons, dis- (“peripersonal neglect”) (336). Far stimuli were perceived
charging during the delay period before the go signal; and 3) normally as in nonlesioned monkeys. The premotor area
movement-related neurons, discharging in association with crucially involved in encoding peripersonal space is area F4
movement onset. The activity of single neurons and popu- (123).
lations during the delay period of reach-to-grasp tasks con-
veys information necessary for correctly reach an object, Electrical stimulation and single neuron studies showed
controlling the direction of the arm movement (FIGURE 5) that area F4 contains a representation of arm, neck, face,
(77, 84). Furthermore, F2 neurons also integrate informa- and mouth movements (123, 142, 150). Prolonged electri-
tion about target location and arm to be used is to plan cal stimulation elicits a complex protective movement pat-
actions (171). F2vr sector is also involved in the control of tern similar to that performed by monkeys when presented
the configuration of the hand for reaching to grasp move- with actual threat (153). Most F4 neurons discharge during
ments (323). the execution of goal-directed motor acts, such as reaching
for food and bringing it to the mouth (142). F4 neurons also
It is important to stress that the dorsal (F2d) and ventro-rostral respond to sensory stimuli. According to their responses,
(F2vr) sectors of area F2 differ in the pattern of cortical con- they can be subdivided into two main classes: unimodal
nections: F2d is mainly connected with parietal areas PEc and (tactile) and bimodal (visual and tactile) neurons (123,
PEip; F2vr has strong connections with parietal areas MIP and 142). The tactile receptive fields (RFs) are large and pre-
V6A (22, 132, 258). Furthermore, F2vr, but not F2d, is target dominantly located on the face, arm, and the upper part of
of a minor, but consistent projection from the dorsal part of the body. The visual RFs are three-dimensional fields lo-
prefrontal area 46 (46d, as defined in Ref. 240). Several pari- cated around different body parts in register with the tactile
etal areas, especially those connected with area F2vr, share fields. The visual RFs are independent of the eye position
with this latter some functional properties. In particular, they and are anchored to the tactile one (FIGURE 6). Some F4
contain neurons active during the execution of reaching move- neurons have also an auditory RF in register with the visual
ments and responds not only to somatosensory, but also to and tactile ones (123, 142, 154).
visual stimuli (43, 81, 130, 180, 193, 250).
It has been suggested that the visual responses in many F4
bimodal neurons represent a potential motor act within the
C. The Pragmatic Coding of Peripersonal peripersonal space. Indeed, the RFs of many F4 bimodal
Space: Area F4 neurons expand in depth when tested with approaching
stimuli moving at increasing velocities (123). This indicates
Our intuitive idea of space is that of a single continuous that these neurons do not encode space in geometric terms,
entity that extends in all directions and within which ob- but encode it dynamically to allow the preparation of ap-
jects have locations relative to one another and the ob- propriate motor acts.
server. According to classical neurology, the neural coun-
terpart of the subjective space perception is a single map Area F4 is strongly connected with parietal areas PF, PFG, and
located in the parietal lobe constructed by the association of VIP (12, 24, 73, 74, 229, 257, 280, 343). In particular, area PF
inputs from different sensory modalities (see Ref. 86). This is the source of somatosensory input related to the face and the
unified, multimodal representation of the world is assumed mouth, while PFG and VIP send somatosensory input related
to provide all the information necessary for acting on an to the forelimb and visual information that they receive from
object and is shared by the different motor circuits that areas involved in the visual motion analysis (MT, MST, FST)
control eyes, arms, hands, and other effectors. An alterna- (41, 81, 262, 389). The main parietal node for representation
tive more recent view is that there are many maps each of peripersonal space is area VIP. The electrical stimulation of
encoding space in terms of different effector movements. this evokes head and arm movements (82, 379). Most neurons
These spatial representations emerge as a result of interac- in VIP respond to the presentation of visual moving stimuli
tions of individuals with their environment and define a (79, 80, 102). Among them there are bimodal neurons, whose
series of motor relations determined by the properties of a properties are for many aspects similar to those of F4. Their

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 663


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

A Cue on Go
RT:

DD:

B
RT Trials DD Trials
GO Cue GO

FIGURE 5. Response properties of neurons


a in dorsal premotor cortex. A: cartoons of the
target panel illustrating the sequence of stimu-
lus events in reaction-time (RT) trials and direct-
delay (DD) tasks. In RT, the monkey has to
reach for a target as soon as it appears on a
screen; in DD task, the monkey is instructed on
where to reach (Cue) and to withhold the move-
ment till a second signal (Go). B: rasters and
histograms illustrating the responses of three
PCA222
PMd cells (a– c) in the RT and DD. In RT trials,
only the activity immediately preceding and after
b the Go signal is shown. In DD trials, single-trial
and histogram data are shown separately ori-
ented to both the Cue and Go signals (vertical
dashed lines). [Modified from Crammond and
Kalaska (84).]

PCB107

PCA167

visual RFs are located in space sectors corresponding to the D. Goal Coding and the “Vocabulary of
tactile fields (FIGURE 6) (102). Recently, bimodal spatial neu- Motor Acts”: Area F5
rons have also been found in area PFG, where also arm-hand
motor acts are represented (344). This area appears therefore To execute an action (e.g., grasping and eating a peace of
to represent a second parietal node for peripersonal space cod- food), one must have a prior intention (to eat the food), select
ing. a sequence of motor acts (reaching, grasping, bringing to the

664 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

A B C
C

AS
P F4 VIP

AI Lu
L

ST

FIGURE 6. Response properties of bimodal neurons in area F4 and VIP. A, top: examples of different types of
tactile and visual RFs of F4 bimodal neurons. Solids around different body parts indicate the 3D visual RFs. Bottom:
example of an F4 bimodal neuron the visual RF of which is independent of eye position. The monkey had to fixate an
LED (fixation position indicated by the asterisks) and released a bar upon its dimming. During fixation a 3D stimulus
was moved toward and away from the monkey face at constant velocity inside (left) and outside (right) of the visual
receptive field. Each panel shows horizontal and vertical eye position, rasters, and histograms representing the
neuronal response and the variation in time between the stimulus and the monkey. Large dots in the rasters
indicate dimming of the fixation point. The neuron tactile RF was located on the lower face. The stimulus direction
reversed when the stimulus was a few centimeters from the tactile RF. Note that the discharge is present when the
visual stimulus is moved toward the tactile RF independent of eye position, and consequently is not related to a
retinocentric frame of reference. Abscissae, time; ordinates, spikes per bin; bin width, 20 ms. [Modified from
Fogassi et al. (123).] B: lateral view of the monkey brain showing the location of area F4 and VIP. C, top: schematic
representation of somatosensory and visual RFs of VIP neurons. The shaded areas on the monkey and screen
represent tactile and visual RF surfaces, the arrows the preferred direction of motion. Bottom: neuron with
direction selective responses to tactile and visual stimulation. Single trial rasters and histograms below the
cartoons are synchronized on stimulus movement onset (vertical line). The neuron responds best to a tactile
stimulus moving front to back on the side of the head and to a large stimulus moving in the same direction in the
far contralateral visual periphery. Vertical calibration bar (left) corresponds to 100 impulses/s; tick marks on
horizontal axis are 200 ms apart. [Modified from Duhamel et al. (102).]

mouth, biting), and finally execute this sequence. This view on ments that form them (328). Neurons discharging for a
action organization implies the organization of different ele- specific motor act typically do not discharge during the
ments at different hierarchical levels: movements, motor acts, execution of similar movements aimed at a different goal,
actions (see TABLE 1). For achieving the action intention, single for example, a neuron that discharges during finger move-
elements must be linked one to the other according to a precise ments for grasping an object does not discharge during
temporal structure, so to generate the “kinetic melody” (241) similar movements aimed at scratching. Furthermore, some
that characterizes normal behavior. of F5 neurons discharge when the same goal is achieved by
using different effectors (e.g., the right hand, the left hand,
While F1 is crucial for movement execution (see above), or the mouth), and thus requires a completely different set
area F5 plays a fundamental role in encoding hand and of movements (FIGURE 7A). Many neurons code specific
mouth motor acts (116, 142, 167, 215, 328). This state- types of grip, such as precision grip, whole hand prehen-
ment is based on single neuron studies that showed that sion, or finger prehension. Finally, concerning the timing of
most F5 neurons encode specific motor acts such as grasp- grasping, some neurons discharge during the whole motor
ing, breaking, holding, rather than the individual move- act (e.g., opening and closure of the hand), while others

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 665


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

only during a certain part of it (e.g., shaping of the hand or temporal aspects of the motor act to be executed (186). A
tacking possession of the object). further demonstration that F5 neurons encode motor acts
has been recently provided by a study in which the same
On these bases, it has been proposed that F5 contains a motor goal (i.e., taking possession of food) was achieved by
“vocabulary” of motor acts (328). This motor vocabulary is means of opposite movements (387). Monkeys grasped ob-
constituted of “words,” each of which is represented by a jects using normal pliers, which require hand closure to take
population of F5 neurons. Some of them encode the general possession of the object, and reverse pliers, that instead
goal of a motor act, others encode how a specific goal- require hand opening to achieve the same goal. Most of F5
directed motor act must be executed, and others specify the neurons encode goal achievement (e.g., taking possession of

A B

666 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

the target) independent of the specific fingers movement spond to somatosensory stimuli, but a consistent percent-
(flexion or extension) required to achieve it (FIGURE 7B). age (⬃25%) is responsive to the presentation of visual stim-
uli (328). Among these latter, a set discharges in response to
Most of the data discussed in this section derive from the the presentation of 3D objects (275, 324). These neurons
study of the neuronal properties of F5p and F5c. Both these are called canonical neurons (FIGURE 8A). They are mostly
sectors are densely anatomically connected with the parietal located in area F5p.
cortex and in particular with areas AIP, PF, PFG, and SII
(37, 74, 145, 239, 257, 308, 343). Interestingly, F5 and Canonical neurons are activated by objects of a specific size,
PFG share numerous functional properties (34, 173, 222, shape, and orientation. Visual and motor object specificities
344). In particular, both areas contain motor neurons en- are reciprocally congruent. There is evidence that the re-
coding motor acts. sponse of canonical neurons to the presentation of visual
stimuli cannot be accounted for in terms of motor prepara-
Recently, the issue of how the intention of an action, i.e., tion. In fact, these neurons also respond when the monkey is
the overarching goal of a series of motor acts, is encoded not required to grasp the presented object, but simply to
in the nervous system has been studied in PFG and in F5, observe it (275). The interpretation of the discharge of ca-
using an identical paradigms. Grasping neurons were re- nonical neurons is that these neurons encode the stimulus
corded in two conditions: in one the monkey grasped a motorically: when an object is seen, the discharge of canon-
piece of food and brought it to the mouth for eating; in ical neurons encode a potential motor act congruent with
the other, it grasped an object or a piece of food to place the properties of the presented object, independently of
it into a container (35, 121). Some neurons showed a whether the act will be executed or not.
stronger discharge when the monkey grasped food to
bring it to its mouth, and weaker or absent when it F5p is strongly connected with the anterior intraparietal
grasped food to put it into a container. Others had an area (AIP) (37, 144, 239). The functional properties of this
opposite behavior (FIGURE 7C). Note that the difference area have been extensively studied by Sakata and co-work-
in response in the two conditions was observed during ers (276, 347, 373). AIP neurons can be subdivided into
grasping that was executed in the same way. Further- three main classes: “motor-dominant,” “visual and mo-
more, factors such as grasping force, kinematics of reach- tor,” and “visual-dominant” neurons. Motor-dominant
ing movements, type of stimulus, and motivation could neurons discharge during grasping and holding both when
not account for the context-specific activation of the neu- the grasping action is performed in light and when it is
rons (34, 121). The activation of these neurons, together executed in complete darkness; they do not discharge dur-
with that of the other neurons within the same sequence ing object fixation. Visual-dominant neurons become active
of motor acts, represents the neural correlate of its over- when grasping is performed in light, but not in dark. They
arching goal, which is the intention of the acting individ- respond to simple object fixation. Visual and motor neu-
ual. This concept will be developed later in the section on
rons discharge during object fixation and during grasping in
action and intention understanding.
light and in dark (FIGURE 8B).

E. Visuomotor Transformations for Grasping Objects can be grasped with various grips depending, not
only on their affordances, but also according to different
In addition to purely motor neurons, area F5 contains neu- behavioral context. In recent studies, Scherberger and co-
rons also responding to sensory stimuli. Many of them re- workers (30, 120) trained monkeys to grasp a handle with

FIGURE 7. Goal and intention encoding in areas F5 and PFG. A, top left: lateral view of the monkey brain showing the location of area F5.
Bottom right: discharge of an F5 neuron active during grasping with the mouth, the right hand, and the left hand. Conventions are as in FIGURE
6A. [Modified from Rizzolatti et al. (328), with kind permission from Springer Science and Business Media.] B: example of an F5 neuron
discharging during grasping with normal and reverse pliers. Top: pliers and hand movements necessary for grasping with the two types of pliers.
Bottom: rasters and histograms of the neurons’ discharge during grasping with pliers. The alignments are with the end of the grasping closure
phase (asterisks). The traces below each histogram indicate the hand position, recorded with a potentiometer, expressed as function of the
distance between the pliers handles. When the trace goes down, the hand closes; when it goes up, it opens. The values on the vertical axes
indicate the voltage change measured with the potentiometer. Other conventions are in FIGURE 6A. [Modified from Umilta et al. (387).
Copyright 2008 National Academy of Sciences, U.S.A.] C: example of motor neuron in PFG modulated by action intention. Top left: lateral view
of the monkey brain showing area PFG. Top right: paradigm used for the motor task. The monkey, starting from a fixed position, reaches and
grasps a piece of food or an object, then it brings the food to the mouth and eats it (I, grasp-to-eat), or places it into a container (II/III,
grasp-to-place). Bottom left: activity of three IPL neurons during grasping in the two actions. Rasters and histograms are aligned with the
moment when the monkey touched the object to be grasped. Red bars, monkey releases the hand from the starting position; green bars, monkey
touches the container. Conventions are as in FIGURE 6A. Bottom right: responses of the population of neurons selective for grasping to eat and
grasping to place. The vertical lines in the two panels indicate the moment when the monkey touched the object and the moment in which the
grasping was completed, respectively. The y-axes are in normalized units. [Modified from Fogassi et al. (121). Reprinted with permission from
AAAS.]

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 667


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

A Canonical neuron of area F5 B Visuo-motor neuron of area VIP


Manipulation Manipulation Object
in light in dark fixation

ab c d e f g ab c de f g

plate ring Motor


dominant
50/s

0
Fix Hold Fix Hold Fix

Visuomotor
cube cylinder

50/s

0
Fix Hold Fix Hold Fix

cone sphere Visual


10 sp/s dominant

1s
50/s

0
Fix Hold Fix Hold Fix

FIGURE 8. Visuomotor transformation for grasping. A: example of a selective F5 canonical neuron. Panels
show neural activity recorded during the grasping in light task with 6 objects. Rasters and histograms are
aligned (vertical bar) with key press (onset of object presentation). Small gray bars in each raster indicate onset
of fixation (red LED, a), object presentation (key press, b), go signal (green LED, c), start of movement (key
release, d), onset of object pulling (e), onset of second green LED (f), and object release (g), respectively. Other
conventions are as in FIGURE 6A. [Modified from Murata et al. (275), with kind permission from Springer
Science and Business Media.] B: different types of hand-manipulation-related neurons under the 3 different
conditions: manipulation in light, manipulation in dark, and object fixation. Raster and histograms are aligned with
the moment at which the monkey released the key in the manipulation task and when the monkey pressed the key
starting fixation in the object-fixation task. In the manipulation task, the line below the histogram shows the mean
duration of the “fixation” period (FIX) and “hold” period (HOLD). In the fixation task, the line below the histogram
shows the mean duration of the “fixation” period (FIX). Other conventions are as in FIGURE 6A. [Modified from
Murata et al. (276), with kind permission from Wiley-Blackwell.]

two different grip types on the bases of distinct contextual visuomotor neurons and neurons in F5 that coding the most
cues (lights of different colors). They found that in both AIP appropriate grip. The information relative to the chosen grip is
and F5, a set of neurons were active after cue presentation, then sent from F5 to F1, where the different movements nec-
showing context-dependent grasp planning activity. essary to grasp the object are selected and the final command
for its execution is generated. Recent physiological data from
The evidence that AIP and F5 are two nodes of a circuit for Lemon and co-workers (75, 313, 361) confirmed this point
visuomotor transformations for grasping received a strong showing that the electrical stimulation of F5 facilitates F1 mo-
support from inactivation studies. Transient inactivation ob- tor output to hand muscles. In addition to the output from F5
tained by injection of muscimol, in either monkey AIP (129) or to F1, descending pathways directly linking F5p with the pro-
area F5 (122), produced a dramatic deficit in the capacity to priospinal system (38) could also play a role in grip generation.
shape the hand according to the object physical features. Once The connections with the cervical propriospinal neurons ap-
touched, the object was correctly grasped, thus showing the pear to play also a role in the control of dexterous fingers
lack of motor deficits. movements, as shown by functional recovery after brain le-
sions (178, 204).
How do the visuomotor transformations for grasping occur?
There are various models that try to explain the role that AIP
F. Possible Functions of the Rostral
and F5 play in this process (112, 120, 186, 334, 373). A com-
mon idea underlying these models is that when an object is
Premotor Areas: Decision of When to
observed, AIP neurons extract specific aspects of its physical Act, the Organization of Motor
properties (affordances as defined by Gibson, Ref. 149) and Sequences and Motor Learning: Area F6
provide F5 with the description of the possible ways in which
the object could be to grasped. On the basis of the intention of The presence of potential motor acts, coded in the posterior
the individual and the context, the prefrontal lobe selects AIP motor areas, implies that there should be a system control-

668 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

ling these motor acts, either permitting their transforma- presented with a panel hosting 16 LEDs, 2 of which could
tions into actions, or preventing their implementation. This be turned on, and had to press them in a specific order.
control system, which comprises many prefrontal and cin- When the first sequence (“set”) was learned, another se-
gulate areas, also includes the rostral premotor areas. quence was shown. Finally, a “hyperset” formed by five
different sets was presented. The results showed that F6
Evidence that a sector of the premotor region plays a con- neurons discharged when the monkeys were learning a
trol on motor execution has been provided by electroen- new sequence and at the beginning of a correct sequence
cephalograpic (EEG) studies in humans (95). These studies performance. In contrast, they showed almost no activity
showed that, about 1 s before movement onset, a slowly during task execution when the monkeys had mastered
increasing negative potential, the “Bereitschaftspotential” the hyperset (FIGURE 9B).
(or “readiness potential”) occurs in the motor region with
its maximum close to the cortical mesial surface. On this Hikosaka and co-workers (346) also examined the activity
basis, Kornhuber (207) suggested that the mesial motor of F6 (pre-SMA) in humans. In an fMRI experiment, three
cortex specifies when a movement has to be done. Interest- learning paradigms similar to that employed in monkey
ingly, the amplitude of this potential is wider when it pre- experiment were used. In one of them, the subjects were
cedes self-paced movements than when it precedes those required to respond on the basis of learned color associa-
externally driven (181). These data suggest that the mesial tions, and not on the basis of learned sequence. It was found
motor cortex plays a role in selecting and allowing volun- that area F6 was activated also in this condition. It appears
tary actions and, in particular, self-initiated movements. therefore that area F6 plays a fundamental role in control-
However, the low spatial resolution of EEG technique does ling action execution in general and not only during sequen-
not allow one to identify precisely the source of this poten- tial actions.
tial, whose precise location remains controversial.
A similar role for F6 was also proposed by Rizzolatti et al.
Single neuron recordings in the monkey suggest that the (333). These authors studied the activity of single neurons
mesial region controlling action selection and initiation in monkey F6 by employing a naturalistic setting. Their
mainly corresponds to area F6 (pre-SMA). This area is
results showed that F6 neurons do not specifically control
weakly excitable to electrical stimulation. The movements
distal or proximal movements, but are active when mon-
observed using currents of high intensities are slow, com-
keys were presented with objects possible targets of motor
plex, and concern mainly the forelimbs (238). Single neuron
acts. Typically, F6 neurons increased their discharge when
recording studies showed that visual responses are common
an object entered into the monkey reaching space. These
in this area, while somatosensory responses are rare. Fur-
neurons, however, did not code the peripersonal space, be-
thermore, differently form F1 and F3 neurons, whose dis-
cause the discharge was present only if the approaching
charge slightly precedes movement onset, the vast majority
stimulus was target of an actual monkey’s movement and
of F6 neurons discharge well in advance of movement ini-
no RF could be mapped. What they appear to code is, in
tiation (7, 261).
contrast, the possibility to act on an object. Indeed, their
Tanji and co-workers (375–377) studied the role of the activity strongly depended on the contingencies of stimulus
medial premotor areas during the performance of move- presentation (near or far, presence or absence of obstacles
ment sequences. They trained monkeys to execute three to action) and on motivational factors (FIGURE 9C).
different acts (push, pull, and turn of a manipulandum) in
different orders, separated by a waiting time. F6 neurons F6 is strongly connected with dorsal and ventral sectors of
typically discharge during the interval between one partic- the prefrontal area 46, with the rostral cingulate area 24c
ular act and another (e.g., during the preparation of a push and with the caudal premotor areas (236, 237). It is there-
following a pull), or having a particular rank order inside fore a site of convergence of information about objects and
the sequence (e.g., third act in the sequence, irrespective of locations in space memory, temporal planning of actions,
the type of movement executed, FIGURE 9A). These results and motivation, which are then used for action selection
prompted the authors to conclude that F6 neurons contrib- and initiation.
ute to the organization of complex motor sequences. In
addition to this role, F6 neurons seem also to code time
interval per se, when crucial for behavior (264, 278).
G. Functions of the Other Rostral Premotor
Areas
While Tanji and co-workers studied the activity of F5
neurons during motor sequences consisting of three dis- Unlike the other premotor areas, very little is known on the
crete motor acts, Nakamura et al. (279a) investigated the functional proprieties of the two other rostral premotor
role of F6 in the sequential execution of multiple reach- areas: areas F7 and F5a. Here we summarize their connec-
ing, directed to different spatial positions. Monkeys were tions and some preliminary functional data.

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 669


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

1. Area F7 sector of F7 (F7-non SEF) is mostly involved in the control of


body movements. The discharge of some neurons in this area is
Area F7 is formed by two sectors. The dorsal sector, the supple- related to both arm and eye movements (125). Part of these neu-
mentary eye field (SEF) is an oculomotor area. Its functional rons show visual responses when the location of the stimulus is
properties will not be dealt with in the present review. The ventral the same of the target of an arm movement (392).

670 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

F7 is strongly connected with the dorsal part of the dorso- geting this area, but probably involving it. These studies
lateral prefrontal cortex, mainly with area 46d and area 8b show that F5a plays a role in the evaluation and comparison
(240). There are also weak connections with the parietal of current and remembered sensory information for percep-
areas PGm and V6A (132, 258). Lesions involving the sec- tual decisions (225, 340) and in the evaluation of the out-
tor of F7-non SEF impair the performance of monkeys in comes of these decisions for future behavior (295, 296).
previously learned motor association tasks and prevent new
learning (see Ref. 297). It is very likely that this area, to-
gether with area F6, plays a role in motor learning. V. THE MIRROR NEURON MECHANISM

2. Area F5a As discussed at the beginning of this review, the ideas on the
functional role of the motor system radically changed in
Area F5 is a caudal premotor area. Recent anatomical these last 20 years. One of the major reasons for this change
data showed, however, that its rostral sector F5a is part was the discovery of the “mirror mechanism.” The essence
of the “family” of the rostral premotor areas (31). F5a of this mechanism is the following. There are neurons, lo-
displays transitional architectonic features between those cated in various part of the nervous system, that encode a
of the agranular-premotor cortex and those of the gran- specific motor behavior and which, in addition, are acti-
ular-prefrontal one (32). More specifically, it is charac- vated by the observation of motor behavior similar to that
terized by the presence of some rare granular cells, a they encode motorically. Thus an observed action pro-
homogeneous and prominent layer V, a low SMI-32 im- duces, in the observer’s brain, a motor activation, as if the
munoreactivity with medium-sized pyramids, and a high observer was actually programming its execution (see Ref.
density of calbindin-immunopositive nonpyramidal neu- 337).
rons over the whole layer III (FIGURE 4). The classifica-
tion of F5a as a rostral premotor area is in line with the The mirror mechanism has been found in monkeys, humans
proposed homology of this area with human area 6r1 of (see below), and, more recently, in birds (197, 314). In
Amunts and Zilles (11), which is part of Broca’s region. monkeys and humans, neurons endowed with the mirror
mechanism have been discovered in cortical center control-
As far as its cortical connections are concerned, F5a is char- ling actions devoid of emotional content (“cold actions”),
acterized by strong connections with the ventral prefrontal but also in centers involved in emotions, like the insula and
areas 12 and 46v, and the rostral frontal operculum (144, the cingulate cortex (128, 367). In birds, mirror neurons
145). Inside the premotor region, in addition to strong con- have been described in motor centers involved in song pro-
nections with the other F5 sectors, F5a is also densely con- ductions (197, 314). These findings indicate that the mirror
nected with area F6. Altogether this connection pattern in- mechanisms have radically different functions that depend
dicates that F5a could be the gate through which higher- on its anatomical location.
order information coming from the ventral prefrontal
cortex access the motor cortex. In this review we discuss only the mirror functions related
to “cold” action. Our review of mirror functions is subdi-
A role of area F5a in cognitive-motor functions is also dem- vided into two parts. In the first we review the functional
onstrated by some physiological studies, not specifically tar- role of the mirror mechanism in monkeys, and in the second

FIGURE 9. Functional properties of area F6. A: discharges of a F6 (pre-SMA) neuron whose activity increased when the monkey was preparing
to initiate the third movement, irrespective of the movement type (push, pull, or turn). In the raster displays, each row represents a trial; dots,
individual discharges of the neuron; filled squares, the occurrence of signals that triggered each movement; ⫹, the onset of movement. The
histograms show sum discharges over 9 trials. The bin width is 40 ms. [Modified from Shima and Tanji (359).] B: activity of an F6 (pre-SMA)
neuron for new hypersets and learned hypersets (see text). Spike activity, shown by rasters and histograms, are aligned at the time when the
monkey pressed the first button for each set. Only activity for correctly executed trials is shown; the first trial at top, the last at bottom. Inverted
triangles in the raster indicates other task-related events (stimulus onset and second button press). For the new hypersets, the cell showed
phasic activity for every set before the first button press; for the learned hypersets, it was nearly silent except for the 1st trial. Note that the
activity for the new hyperset decreased as learning proceeded. Top 3 rows: data when the monkey used the hand contralateral to the recording
site. Bottom row: data when the ipsilateral hand was used. Correct orders of button presses for the hypersets used are shown below. [Modified
from Nakamura et al. (279a).] C, top: cartoons of the sequence of reaching-grasping task during testing of neurons in F6. Top sequence:
reaching-grasping in vision. An object was presented and moved by means of a mechanical carrier toward the monkey who gazed on it, and, when
reachable, reached and grasped it. White lines, monkey gaze; yellow curve, approximate extension of reaching distance. Bottom sequence:
reaching-grasping without visual control. The vision of the target is prevented, and the movement is executed without visual guidance. Bottom:
F6 neuron active during premovement and movement periods. a and b: reaching and grasping in response to a visual object; c: same action
toward a hidden object. In the left panel, the activity is aligned (vertical line) with the onset of the saccades elicited by the object approaching the
monkey. In the middle and right panels, the activity is aligned with the time when the monkey touched the object; the dashed line in the histograms
represents the average time of movement onset. Triangles on the individual trial lines indicate the saccade on the target in the visual condition.
[Modified from Rizzolatti et al. (333).]

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 671


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

we examine its role in humans. We refer to the ensemble of The first experiments on mirror neurons investigated
cortical motor centers endowed with the mirror mechanism mostly the upper sector of F5, i.e., the sector in which hand
as the mirror neuron system (or mirror system). actions are represented. Subsequently, the ventral part of
F5, related to mouth movements, was also explored (116).
The results showed that ⬃25% of neurons in this part of F5
VI. MIRROR NEURON SYSTEM
have mirror properties. Mouth mirror neurons were subdi-
IN MONKEYS
vided into two classes: ingestive and communicative neu-
rons. Ingestive mirror neurons (⬃80% of the total amount
A. Mirror Neurons of Area F5 of the recorded mouth mirror neurons) discharge when the
monkey observes actions related to ingestive functions (bit-
1. Basic properties ing, grasping food with the mouth, or sucking). Communi-
cative mirror neurons responded to the observation of com-
Mirror neurons are a distinct class of motor neurons that dis- municative gesture such as, for example, lip smacking.
charge both when a monkey performs a specific motor act and However, as the ingestive mirror neurons, they strongly
when it observes the same or a similar motor act done by discharged during the performance of ingestive acts. The
another individual (monkey or human) (FIGURE 10). The mo- reason for the discrepancy between the effective visual input
tor properties of F5 mirror neurons are similar to those of (communicative) and the effective active motor act (inges-
other neurons of this area. Unlike canonical neurons, they tive) is not clear. Hypotheses on language evolution suggest
do not respond to presentation of objects, including “inter- that communicative gestures derived from ingestive actions
esting” objects like food. Observation of intransitive ac- (243, 394). On this basis, one may argue that the commu-
tions is ineffective in triggering mirror neurons (98, 127, nicative mouth mirror neurons reflect a stage of corticaliza-
330). tion of communication not yet freed from the original in-
gestive functions of this area.
A fundamental property of mirror neurons is that they re-
spond to the observation of motor acts having the same goal 2. Mirror neurons of area F5 code the goal of the
as that they encode motorically. According to the first de- motor acts
scriptions of F5 mirror neurons, the observation of actions
done with tools was ineffective in triggering these neurons A crucial issue concerning the mirror neurons is their role in
(127, 330). Further investigations reported, however, that a cognition. Before addressing this issue one must address,
set of mirror neurons discharged in response to tool actions however, a preliminary problem: what do the parieto-fron-
(327). More recently, these findings were confirmed by tal mirror neurons encode when they discharge in response
Rochat et al. (338) (see also Ref. 301). to the observation of others’ actions? The response is
straightforward. In the case of extracellular recording (like
in all studies described up to now), what is studied is the
A neurons’ output. This is true both for action execution and
for action observation. Thus, if one knows what a neuron
encodes during the execution of a motor act, one also
knows what it encodes during the observation of motor act
done by another individual. The output of a neuron does
not change in relation to how the neuron is triggered. Thus
since there is compelling evidence that, during active behav-
ior, the neurons of area F5 encode the goal of motor acts,
also the observation of motor acts done by others must
generate the representation of the motor act goal.
B
An interesting issue is whether to activate F5 neurons is
necessary the vision of the motor act done by another indi-
vidual or is enough to have information sufficient for un-
derstanding the goal of others’ motor act even in the ab-
sence of its vision. This point is crucial for establishing the
generalization capacity of the mirror neurons. If mirror
neurons mediate goal understanding, regardless the stimu-
lus contingencies, it is reasonable to conclude that their
500 msec
activity reflects the meaning of the observed motor act. Two
FIGURE 10. Example of an F5 mirror neuron. A: grasping obser-
series of experiments addressed this problem. The first
vation. B: grasping execution. [Modified from Di Pellegrino et al. (98) tested whether F5 mirror neurons were able to recognize the
with kind permission from Springer Science and Business Media.] motor acts from their sound (206), and the second tested

672 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

whether the understanding of a motor act, based on some of its presence. The results showed that neurons responding
memory clues, can trigger their activity (388). to the observation of grasping in the full vision condition
also discharge in the hidden condition. It appears therefore
Kohler et al. (206) studied mirror neurons while the mon- that is the meaning of the observed actions, and not the
key was observing a motor act characterized by a typical vision of it, that trigger mirror neurons.
sound for an action (e.g., ripping a piece of paper), and
while this sound was presented without the vision of the Although some authors interpreted the results of this exper-
corresponding motor act. The results showed that ⬃15% of iment in terms of action prediction, this interpretation goes
mirror neurons responding to the presentation of motor act beyond the data. When a mirror neuron fires, its discharge
accompanied by sounds also responded to the presentation describes the goal of a given motor act. It “says” what
of the sound alone. Control experiment showed that the happened here and now, i.e., grasping. The same is valid for
response to the sound of motor acts did not depend on the “hidden” condition of the study by Umilta et al. (388).
unspecific factors such as arousal or the emotional content Because there is sufficient information in the context for
of the stimuli (FIGURE 11). understanding the motor act, the neuron signals the goal of
the observed motor act, exactly as when the grasping is
The logic of the second experiment (388) was the following: visible. Prediction is an arbitrary interpretation.
if mirror neurons respond to the goal of a motor act, they
should also discharge when monkey has sufficient informa- Taken together, the two experiments described above show
tion to produce a mental representation of what another that mirror neurons respond to the goal of the observed
individual is doing, even in the absence of visual informa- motor act also in the absence of visual cues. These cues
tion. Neurons were tested in two conditions. In one, the trigger mirror neurons only insomuch as they allow the
monkey saw an object-directed action (“full vision” condi- understanding the goal of motor act. When goal compre-
tion); in the other, the same action was presented but with hension is possible on other bases, mirror neurons are able
its final part hidden (“hidden” condition). A piece of food to signal it, even in the absence of visual stimuli.
was placed behind the screen with the monkey being aware
3. Factors influencing the discharge of F5 mirror
neurons: space, valence, view and type of the stimuli
Population
Already in the early studies of mirror neurons, it was noted
n=33
0.6 that other factors than mere action observation could influ-
ence the discharge of mirror neurons. Among them, the
hand the experimenter used for interacting with objects and
Vision + Sound 0
the spatial (left or right) location of the experimenter’s mo-
tor act (see Ref. 127).
n=33
0.6 A recent study addressed a specific aspect of this issue (54)
net normalized mean activity

and, namely, whether the location of the stimuli near the


Vision 0
monkey (peripersonal space) or far from it (extrapersonal
space) was instrumental in triggering F5 mirror neurons.
The results showed that about half of the recorded neurons
n=33 0.6 were modulated by the spatial location of the observed
motor act. Of these, ⬃50% preferred stimuli presented in
the peripersonal space and 50% stimuli presented in the
Sound 0 extrapersonal space. In a subsequent experiment, the nature
of this representation was investigated. Specifically, it was
tested whether space was represented in F5 in terms of the
n=28
0.6 distance from the monkey’s body (metric representation) or
in terms of working space (operational representation). An
Motor experiment was, therefore, carried out in which a transpar-
0
ent panel was introduced between the monkey and the site
1s of the action. This prevented the monkey, which obviously
FIGURE 11. Responses of the population of tested neurons selec- saw the action, from the possibility to interact with objects
tively activated by the vision-and-sound, vision-only, sound-only, and in its peripersonal space. In a purely metric representation,
motor conditions. Responses to the most effective stimuli are shown this manipulation would leave the peripersonal and extrap-
in blue, and those to the poorly effective stimuli are in red. Vertical
lines indicate the auditory response onset in the population. The
ersonal regions unchanged, while in an operational repre-
y-axes are in normalized units for the population. [Modified from sentation it would lead to a remapping of the peripersonal
Kohler et al. (206). Reprinted with permission from AAAS.] into extrapersonal space. By using this paradigm it was

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 673


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

found that ⬃50% of tested mirror neurons, all selective for eas, thus binding the goal of the observed motor act (visu-
peripersonal space, changed their tuning properties. That is, ally unspecified grasping) with its pictorial description.
after positioning of the frontal panel, these mirror neurons
lost their tuning. These results indicate that area F5 neurons From the data presented above it is clear that spatial aspects
encode space in two different ways. The first is purely met- of the observed actions may modulate the responses of mir-
ric and depends only on the distance between the monkey’s ror neurons. The recent anatomical discovery that informa-
body and observed actions, while the second one is action- tion on object semantics reaches F5 (see anatomical section)
centered and depends on the possibility of the monkey to raised the question of whether the value of the object acted
execute a motor act in a given space region close to its body. upon by others may influence the activity of mirror neurons.
This problem was recently addressed in three experiments
Most experiments on mirror neurons have been performed (52). The results showed that a large subset of mirror neu-
using what is called “naturalistic testing.” This testing con- rons in F5 is influenced by the subjective value of the object
sists of recreating, during the experiment, a quasi-real life on which the observed action is performed. These data im-
condition with the experimenter placed in front of the mon- ply that a subset of F5 mirror neurons have access to key
key and manipulating food, feeding the monkey, and per- information that may be useful to shape the behavior of the
forming a variety of other actions (98, 127, 330). Mirror observer, according to the value of the object grasped by
neurons were discovered thanks to the naturalistic testing. another individual.
It is obvious, however, that the precision of this method is
limited by the variability of the experimenter’s movements. Recently an experiment investigated the output of area F5
To test mirror neurons in a quantitative way, one must use by recording from identified pyramidal tract neurons
a more controlled setup and, especially, standardized stim- (PTNs) and by testing them for possible mirror properties
uli like filmed actions. But, are these stimuli effective? Be- (211). It was found that ⬃50% of the tested PTNs were
cause there were some negative, although rather anecdotic modulated by action observation. This finding indicates
results, on this point, the problem of the efficacy of filmed that mirror neuron activity can be transmitted directly to
actions was formally addressed by Caggiano et al. (53). the spinal cord via PTNs. It was also found that several
They trained monkeys to fixate a central spot during film
PTNs (17/64) showed suppression of discharge during ac-
presentation and compared the neuron responses to those
tion observation, while they fired during active grasping.
obtained during naturalistic testing. The results revealed
The authors suggested that inhibitory effect might play a
that 43% of the tested neurons showed no significant pref-
role in preventing movement generation during action ob-
erence for either type of action stimulus. The remaining
servation.
neurons exhibited (in almost all cases) a stronger response
for naturalistic motor acts. These findings opened the pos-
In a following study, the same authors recorded from PTNs
sibility to examine in a more quantitative way the property
originating from primary motor cortex (F1) (395). Many of
of mirror neurons.
them were modulated during action observation. Most
PTNs increased their discharge during observation (“facil-
One of these studies investigated the view dependence of the
itation-type” mirror neurons), while others showed re-
visual responses of mirror neurons (53). To this purpose
movies were presented showing the same motor act (grasp- duced discharge (“suppression-type” mirror neurons). By
ing) seen from three different viewpoints: from the mon- comparing the properties of PTNs F1 mirror neurons with
key’s perspective (subjective point of view) (0°), from a side those of PTNs F5 mirror neurons they found that the visual
view (90°), and from the frontal view (180°). The results response in F1 was much weaker than in F5. Thus, although
showed that the large majority of tested neurons (74%) many F1 PTNs discharge during action observation, their
were view-dependent, i.e., they showed a significant dis- input to spinal cord circuitry might be insufficient to pro-
charge preference for at least one view. In the remaining duce overt muscle activity. Taken together with previous
neurons (26%) the responses did not vary in the three view data on F5 PTNs, these findings indicate that motor goal
conditions. There are two possible interpretations of these understandong is not function of F5 mirror neurons, but
data. The first is similar to that concerning the spatial prop- rather of the activation of complex motor schema that in-
erties of mirror neurons. When a grasping neuron fires, it volves corticospinal tract neurons.
can “say” only an undifferentiated “grasping.” However, if
different sets of view-dependent mirror neurons have differ-
ent anatomical connections, their firing determine or B. Mirror Neurons and Motor Intention
prompt the activation of other neurons allowing in this way Understanding: Areas PFG and Area F5
an adequate preparation to the observed stimulus. Another
possible interpretation of the role of view dependent neu- The observation of a motor act done by another individual
rons is that their discharge is conveyed through descending allows the observer to understand the goal of the observed
(top-down) selective connections to higher-level visual ar- motor act, but often also the intention behind it. Intention

674 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

understanding can be based on how the agent interacts with ing neurons discharge with a different intensity according to
the object. For example, a glass of beer grasped by its body the overarching goal (intention) of the action in which
indicates the desire to drink the beer, while the same glass grasping motor act was embedded (action-constrained neu-
grasped by the top suggests the intention to move the glass. rons). Examples of action-constrained grasping neurons are
Even when two grips are identical, one may guess the inten- shown in FIGURE 7C.
tion of the agent from the object the agent is grasping or
from the context. In this section we present experiments This organization of motor acts represent, most likely, the
showing that mirror neurons are involved, besides in encod- mechanism on the basis of a fluid action execution. Neurons
ing the goal of the motor acts also in motor intention un- that encode a specific motor act within a given action are
derstanding. linked with neurons that encode the following motor act.
Thus motor chains that represent the entire action are
The first experiments trying to assess a possible relation formed. These motor chains represent the neural substrate
between motor intention and mirror activity were carried for implementing the agent’s motor intention.
out in the inferior parietal lobule, mostly in area PFG. As
described in previous sections, the motor organization of The aim of the second part of the experiment (visual para-
area AIP and PFG is based on goal-directed motor acts digm) was to find out whether the visual responses of ac-
(121, 276). In both these areas, there are neurons encoding tion-constrained neurons were modulated by the overarch-
specific motor acts and specific types of grip. Furthermore, ing goal of the actions in which the motor acts were embed-
both these areas contain mirror neurons whose general ded. The results showed that the intensity of the discharge
properties are similar to those of area F5. The experiments of the majority of IPL mirror neurons was modulated by the
consisted of two parts (121). In the first part, monkeys were agent’s intention. Examples are shown in FIGURE 12.
trained to grasp objects with two different intentions: eating
or placing using the same grip (motor paradigm). In the The discharge of action-constrained grasping neurons is
second part, monkeys observed the experimenter grasping longer in their specific action than in other actions. It is
object with two different intentions using the same grip likely that this prolonged discharge facilitates the next mo-
(visual paradigm). Neurons were recorded from IPL tor act in the executed action. Data on the receptive field
(mostly PFG), and their discharge during grasping was stud- properties of parietal neurons are in accord with this inter-
ied in the motor and visual paradigm. The results obtained pretation. They show that neurons that encode the next
during active grasping showed that two-third of IPL grasp- motor act in an action sequence (408) are activated by

Visual response of mirror neurons

Unit 87 Unit 39 Unit 80

100 100 100


Spk/s

Grasp to eat

Grasp to place

1s

FIGURE 12. Example of motor neuron in PFG modulated by observation of action intention. Activity of three
IPL neurons during grasping observation in two actions. Rasters and histograms are aligned with the moment
when the experimenter touched the object to be grasped. Red bars: experimenter grasps the object. Unit 87
is selective for observation of grasping to eat, while unit 39 is selective for observation of grasping to place. Unit
80 is not selective. Conventions are as in FIGURE 7C. [Modified from Fogassi et al. (121). Reprinted with
permission from AAAS.]

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 675


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

movements performed actively or passively. Thus, when an clips showing a hand grasping an object activate the premo-
action-constrained grasping neuron became active, it trig- tor areas F5a and F5c and the prefrontal areas 45B. When
gers the whole motor chain in the observer that in this way instead of a hand the video clip showed an individual grasp-
may grasp the agent intention. ing an object, activation was also present in F5c. Finally,
area 45B, but not the various sectors of premotor F5, re-
Bonini et al. (34) recorded neurons from F5 applying the sponded to shape presentation. Area 45B appears therefore
same two-intention paradigms. The results showed that to be a higher order area where converges both object and
also area F5 contains “action-constrained” grasping neu- action information.
rons active both during action execution and action obser-
vation. The authors concluded that the similarities between In a subsequent fMRI study, the same authors examined the
mirror neuron properties of the two areas indicate that they activations in the monkey STS region and posterior parietal
constitute a functional circuit underlying others’ intention lobe during the observation of grasping actions (282). As in
understanding. the previous study, videos were presented where either only
a grasping hand or an entire person was shown grasping an
C. The Mirror Neuron Circuit for Hand object with his hand. Both types of videos activated areas in
Action and Some Considerations on Its the depths of lower and upper banks of STS, and areas PFG
Functional Role and AIP (FIGURE 13). A further ROI analysis showed that,
relative to static control and scrambled stimuli, two areas
1. The mirror neuron circuit for hand actions were active in IPL: PFG on the convexity and AIP within the
intraparietal sulcus. Areas consistently more active during
Single neuron recording is undoubtedly the best technique action observation than during control conditions in STS
for understanding brain mechanisms. However, to have a were MT/V5, FST, LST, LB2, and STPm.
global picture of areas involved in a given function, brain-
imaging techniques are of invaluable help. The first study The large number of STS areas active during action obser-
that employed fMRI for defining the areas active during the vation raise the question of which of them sends informa-
observation of motor acts in monkeys was carried out by tion to the two parietal areas involved in grasping coding.
Nelissen et al. (283). The focus of this study was the frontal To solve this problem, retrograde tracers were injected into
lobe areas. By using a ROI analysis, they found that video areas AIP and PFG. Injections into area AIP yielded a wide-

FIGURE 13. STS-IPL-F5 grasping observation network.


A: lateral view of a macaque brain showing locations of
B three regions involved in action observation. B: flattened
representation of STS, IPS/IPL, and IAS (inferior arcuate
sulcus). FEF (frontal eye fields); F5c, F5p, F5a. Visual infor-
mation on observed actions are sent forward from STS
through parietal cortex to area F5 along two functional
routes: a STPm-PFG-F5c pathway and a LB2-AIP-F5a/p
pathway indicated with red and blue arrows, respectively.
Area 45B receives parietal input from LIPa and also has
direct connections with the lower bank of STS (green ar-
rows). The colored arrows specify the functional routes.
For abbreviations, see text. [Modified from Nelissen et al.
(282).]

676 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

spread ipsilateral STS labeling. In particular, three sectors This hypothesis that the activation of mirror neurons plays
of the STS were strongly labeled: MSTd, a sector in the a fundamental role in action understanding is referred to as
middle portion of the upper bank near the fundus, and one “direct matching hypothesis” (329). This hypothesis main-
in LB2 (FIGURE 13). The consistency analysis (the presence tains that, given that neurons in the ventral premotor cortex
of labeling in all studied monkeys) indicated, however, that encode the goal of a motor act, when they became active
consistent labeling was only observed in LB2 and laterally during the observation of a motor act done by others, the
in the inferior temporal cortex. Injections into area PFG goal of that motor act is represented in the observer’s motor
produced labeling concentrated in three sectors of the ipsi- system and hence understood. Note that, as discussed
lateral upper bank of the STS (FIGURE 13 area MSTd, above, it is not the activity of mirror neurons per se that
STPm, and UB1). These data were confirmed by the consis- allows others’ action understanding, but the fact that the
tency analysis. It is important to note that of among these activation of mirror neurons “ignites” a large number of
labeled sectors, only STPm showed a consistent activation neurons located in other motor centers, among which the
during grasping action observation. basal ganglia (6, 248) and cortico-spinal tract (395). The
congruence between the exogenously generate motor
In conclusion, these data indicate that action information
schema and that endogenously generated during voluntary
from STS is sent to the ventral premotor cortex (F5) along
actions allows the understanding of the goal of the observed
two distinct functional pathways. One pathway links the
motor act (see Ref. 337).
upper bank of the STS with area PFG that, in turn, projects
to area F5c. The other connects the lower bank of the STS
An alternative interpretation has been proposed by Kil-
with area AIP that in turns projects to areas F5a and F5p.
ner et al. (199, 200). According to these authors the role
Both pathways provide information necessary for under-
of mirror neurons in understanding the goals of the ob-
standing the observed motor act. However, they possibly
served motor acts is better accounted for within a predic-
provide different information for understanding the inten-
tion underlying the motor act. The first pathway is con- tive coding framework. In the predictive coding frame-
cerned with the agent of the motor act, while the second work, each level of cortical hierarchy employs a genera-
route is more concerned with the details of hand grip and tive model to predict representation in the level below.
object semantics and may aid in understanding motor acts Backward connections convey the prediction to the lower
with respect to these factors. level where it is compared with the representation in the
subordinate level to produce a prediction error. This pre-
diction error is then sent back to the higher level, via
2. The functional role of visual and motor areas in
forward connections, to adjust the neuronal representa-
action understanding
tion of sensory cause.
The circuit activated by action observation includes parietal The strong point of “direct matching hypothesis” is the
and premotor areas with motor properties and a series of
view that excitation of the mirror neurons (and the conse-
higher-order purely visual areas. Why should the areas with
quent activation of specific motor schemata) allows the un-
motor properties be involved in the understanding of motor
derstanding of the goal of an observed motor act regardless
acts? While should the activation of visual areas be not
of how it is executed (generalization). The weakness is that
sufficient? Indeed, as shown by Perrett and colleagues (188,
it is a strictly forward model that does not take into account
306), STS neurons visually describe the actions of others, a
backward projections, present in large number in the pari-
finding confirmed by several fMRI studies showing a similar
role for STS in humans (8, 319, 407). However, if there is no eto-frontal circuits.
doubt that the action analysis carried out by STS neurons
represents a fundamental step for understanding others’ In contrast, the notion that top-down mechanism plays an
behavior, it also seems highly implausible that STS may important role in action understanding is one of the big
accomplish the process of action understanding by itself assets of the “prediction hypothesis” of mirror neurons ac-
(337). In fact, crucial for action understanding are those tivity. However, as recently stated by Kilner et al. (199,
areas that contain neurons that encode the goal with a high 200), this hypothesis requires a prior expectation about the
degree of generality. With the problem posed in these terms, goal and the intention of the observed action. To solve this
the evidence shows that generalization is an aspect that problem, Kilner (199) suggested that this prior expectation
characterizes the parieto-frontal mirror network rather is mediated by a circuit that includes middle temporal gyrus
than STS neurons. In fact, although STS neurons may en- and area 47. While there is no doubt that the context and
code motor acts, they are unable to generalize the action the object semantics might be important cues for under-
goal (188, 306). In contrast, parieto-frontal mirror neurons standing the intention underlying a given motor act, this is
are able to encode the goal of the observed motor acts not necessary for understanding the goal of a motor act.
regardless of whether motor acts are performed with the Direct activation of parietal and premotor neurons is suffi-
mouth, the hands, or even with a tool. cient for this purpose.

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 677


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

D. Mirror Neurons in Area VIP targets visible, 2) only the replayed targets visible, and 3) only
the moving cursor visible. It was found that, in both PMd and
The studies on the mirror mechanism we discussed up to F1 there were neurons that discharged in a similar way during
now were focused on how this mechanism contributes to the execution and the observation of the task. No responses or
the understanding of the actions performed by others. Is weak responses were found when the monkey saw the cursor
mirror mechanism also involved in understanding observed without targets without cursor. The authors concluded that
action that an individual does towards his/her own body? these neurons (“mirror-like” neurons) respond to the goal of
This issue was addressed by Ishida et al. (179) by studying the motor act.
the proprieties of area VIP coding peripersonal space. As
mentioned above, in area VIP there are bimodal, visual, and
VII. THE MIRROR NEURON SYSTEM IN
tactile neurons. Their tactile receptive fields are located pre-
HUMANS
dominantly on the face, and the visual receptive fields are in
spatial register with the tactile ones (80, 102).
The possibility to record single neurons in humans is limited to
Single neurons were recorded from monkey VIP. The stud- preoperative recordings in patients undergoing surgery for in-
ied neurons showed tactile receptive fields and visual recep- tractable epilepsy. In such cases single neurons have been re-
tive fields in spatial register with the tactile ones. The extent corded, but, for technical reasons, exclusively from the medial
of the peripersonal space of each neuron was delimited. wall of the hemispheres. As a consequence, the cortical regions
Then the stimulus was presented outside it, at a distance of in which mirror neurons are expected to be (the convexity of
120 cm. No visual response could be elicited at this dis- the parietal and frontal lobes) have not been explored. A group
tance. However, if an experimenter moved inside the mon- of researchers tested single neurons from the medial temporal
key visual field standing at a distance of 120 cm from it, and cortex and the medial frontal cortex for mirror properties.
waved his hand close to his own body parts that corre- They reported neurons that discharged during both execution
sponded to that where was the monkey tactile receptive and observation of congruent actions (271).
field, the response appeared. The responses disappeared
again, if the experimenter moved away (179).
A. Brain Imaging
This study is of great interest because it shows a way in
which individuals might encode the body of others. Note Brain imaging techniques (PET, fMRI) are the gold stan-
that, although Ishida et al. (179) have not studied the motor dard in present-day neuroscience studies for topographical
responses of the recorded neurons, area VIP is strictly con- localization of brain functions. However, two great limita-
nected with area F4 (see above), where peripersonal space is tions are intrinsic to these methods, namely, the lack of
encoded in terms of reaching movements. It is likely there- temporal resolution and the correlative nature of the re-
fore that the observed visual responses actually represent sults. In spite of this, a large number of experiments using
potential motor acts directed towards specific body parts these techniques demonstrated a strikingly consistent pat-
(123, 142). tern of cortical activity evoked by the observation of others’
actions, overlapping with the cortical network that is re-
cruited during action execution. The bulk of studies have
E. “Mirror-like” Neurons in Areas F2 and F1 been recently reviewed in several meta-analyses (63, 157,
268) (see FIGURE 14).
There is evidence that there is a set of neurons located in the
dorsal premotor cortex (PMd or F2) that are active when a These reviews confirmed the early PET and fMRI studies (46,
monkey performs a conditioned task (moving a cursor on a 152, 156, 174, 176, 209, 210, 247, 304, 331) showing that the
computer screen) and when it observes the same task exe- executed and observed goal-directed grasping movements are
cuted by the experimenter. These neurons, unlike “classi- encoded in a circuit formed by the IPL and by the ventral
cal” mirror neurons, did not need to be activated the obser- premotor cortex (PMv) plus the caudal part of the inferior
vation of a hand acting on an object. The observation of frontal gyrus (IFGc). It has been also reported that PMd and
goal achievement is sufficient (76). the SPL could be active during action observation and execu-
tion. It is possible that these activations are due to mirror
Dushanova and Donoghue (104) and Tkach et al. (380) ex- mechanism. It is equally possible, however, that they reflect
tended this observation. In particular, Tkach et al. (380) motor preparation. Monkey single neuron data showing that
trained monkeys to move repetitively a cursor to targets ran- these areas play an important role in covert motor preparation
domly presented. There were two phases in the experiment. In support the last possibility (84, 194).
the first phase (active movement phase), the monkey moved
the cursor, and in the second phase (observation phase), the The issue of areas that are activated both during action obser-
monkey observed the replayed movements of the first phase. vation and execution was addressed by Gazzola and Keysers
The observation phase had three conditions: 1) cursor and (139), using single-subject analyses of unsmoothed fMRI data.

678 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

rostrally. Observing distal motor acts activated voxels that


consistently overlap with the ones recruited by active grasping.

2. Representation of goals: decoupling movements


from motor acts

As already discussed, motor representations in the cerebral


cortex of the monkey range from representations of joint
displacements (i.e., movements) to representations of goals
(motor acts), to representations of motor intention. Evi-
FIGURE 14. Meta-analysis showing the activations during action
dence for representation of goals in human motor areas
observation. Results from meta-analysis are displayed on the left and
right lateral surface view of the MNI single subject template. pMTG, comes from a series of fMRI studies.
posterior middle temporal gyrus; SMA, supplementary motor area
(hidden within the interhemispheric fissure); BA44,45, Broca’s ar- First evidence in this sense derives from the study of move-
ea; BA6, lateral premotor cortex; SI, primary somatosensory cortex ments performed by artificial effectors. Video-clips of a hu-
(BA2); 7A, superior parietal area; PFt, inferior parietal area; hIP3,
man or a robot arm grasping objects were presented to
intraparietal area; V5, extrastriate visual area. [Modified from
Caspers et al. (63). Copyright 2010, with permission from Elsevier.] normal volunteers by Gazzola et al. (140). The data showed
that in spite of shape and kinematics differences, the pari-
eto-frontal mirror network became active in both condi-
The data showed that voxels shared between action observa- tions. This study was extended by Peeters et al. (301) who
tion and action execution were located, in addition to the studies the cortical activations during the observation of
classical parieto-premotor circuit, also in various other corti- motor acts executed by a human being, by a robot or per-
cal areas, namely, dorsal premotor, middle cingulate, somato- formed with a tool. The results showed bilateral activation
sensory, superior parietal, and middle temporal cortex. It is of the mirror network independently of the effector used.
plausible that the activations outside the classical mirror cir- Most interestingly, tool action observation determined ac-
cuit reflect sensory predictions from internal models. These tivation of a specific rostral sector of the left anterior supra-
activations would enrich the information about other individ- marginal gyrus (FIGURE 15). Shimada (360) investigated the
uals’ actions that the mirror mechanism provides. congruency between the shape of the acting effector (human
vs. robot) and movement kinematics (human vs. robot). His
1. Organization of the representation of observed results showed that there was an interaction between these
movements: reaching and grasping factors, with strong “deactivations” in sensorimotor areas
when the individuals saw a human agent performing ac-
A classical distinction in motor organization is the one be- tions with robotic kinematics.
tween the two parallel “modules” of reaching and grasping
(see Ref. 187). In agreement with the notion of the mirror Second evidence in favor of goal coding is provided by
mechanism, observed proximal reaching movements are typ- aplasic subjects. Two individuals born without arms and
ically associated with the activation of dorsal parieto-frontal hands and normal controls were scanned while they were
circuit linking the convexity of the superior parietal lobule to watching videos showing hand actions. In addition, the two
the dorsal premotor cortex. However, only few studies (97, aplasic individuals also made actions using their feet and
118, 245) investigated the representation of observed reaching mouth. It was found that in the aplasic individuals the pa-
motor acts isolated from the distal component of grasping. rieto-frontal mirror circuit active during the execution of
These studies found activations in a dorsal parieto-frontal cir- foot and mouth movements was also active when they ob-
cuit spatially overlapping with the areas activated during exe- served hand motor acts that they had never executed but the
cution of reaching. A study addressed this issue categorizing goals of which they could obtain using their feet or mouth
the observed movements in distal limb, proximal limb, and (141). Similar observations on individuals with limb ampu-
axial movements (348). It was found that axial movements tation were provided by Aziz-Zadeh et al. (234).
activated the medial premotor cortex (SMA), proximal limb
movements activated the dorsal part of the convexity of the Third evidence that confirmed goal coding in the mirror
premotor cortex and, finally, distal limb movements activated circuit is based on presentation of stimuli in a sensory mo-
ventral premotor cortex. dality different from the visual one. An early study showed
that listening to piano tunes activated the parieto-frontal
The bulk of studies on action observation investigated object mirror system in expert pianists (161). More recently, Lewis
grasping (63, 157, 268). These studies confirmed the monkey et al. (228) demonstrated that the middle portion of the
data that the execution of object-centered visually guided hand STG was activated bilaterally during listening and catego-
actions relies on a ventral parieto-frontal circuit including the rization of animal vocalizations while, in contrast, during
anterior intraparietal area (AIP) and the IPL caudally and the listening and categorization of sounds of tools manipulated
ventral premotor cortex plus the posterior part of the IFG by humans, the activated areas were those forming the pa-

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 679


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

A limb movements located medially, followed in the lateral


direction by the upper limb and by the face/head move-
ments. In contrast, as mentioned above, the premotor cor-
tex encodes mostly motor acts, largely independent of indi-
vidual movements, but still maintaining a certain somato-
topic organization, congruent with that of the primary
motor cortex. This double coding has produced apparently
contradictory results in imaging studies that addressed the
issue of whether observed motor acts are encoded in a so-
matotopic (following the spatial representation of a body
map) or actotopic way (clustering together different motor
acts with the same goal, irrespective of the body part used).
B
In a pioneer study, Buccino et al. (46) presented normal
volunteers with video clips showing transitive motor acts
performed with the mouth, hand, and foot. The activations
during the observation of these motor acts were contrasted
with those obtained during the observation of the same
stimuli presented statically. They found a trend towards a
somatotopic organization both in the parietal and in the
frontal lobes, although with considerable overlap. A subse-
quent study (401) used meaningless intransitive movements
of the face, hand, or leg and found a similar somatotopic
arrangement in the frontal and parietal lobes, but present
C only in the right hemisphere. Note that also Buccino et al.
(46) found a prevalence of mouth movements (ingestive) on
the right premotor cortex. Recently, Abdollahi et al. (1)
investigated the cortical representation of manipulation, lo-
comotion, and climbing. The result showed that observa-
tion of climbing activated the rostro-dorsal part of SPL,
locomotion the same region, although weakly, and finally,
manipulation AIP.

A subsequent, more detailed, study addressed specifically


the issue of somatotopy versus actotopy (182). Volunteers
FIGURE 15. Cortical region specifically activated by tool action obser-
were presented with video clips showing four different mo-
vation. A: cortical regions activated by the observation of hand-grasping tor acts (dragging, dropping, grasping, and pushing) per-
actions relative to their static controls, rendered on lateral views of the formed with three different effectors (foot, hand, and
standard (MNI) human brain. B: cortical regions activated by the observa- mouth). The results showed that the general picture of mo-
tion of mechanical hand actions relative to their static control. C: cortical
tor act representation was in favor of goal coding. In the
regions more active during the observation of grasping performed with a
mechanical implement than during the observation of the same action frontal lobe a mixed picture was evident in which, although
done by a hand, each relative to its static control. The purple arrow a clear somatotopy was not found, there was some spatial
indicates the activation in aSMG. [Modifed from Peeters et al. (301).] segregation of the three effectors, with foot being more
represented dorsally and the other two effectors overlap-
ping one with the others. In the parietal lobe no clear dis-
rieto-frontal mirror network. The same network was also tribution according to effectors was found. In the IPL, the
found to be activated by Gazzola et al. (138) by presenta- motor acts clustered according to the valence of the motor
tion of sound of hand and mouth actions. The activation act, regardless of the effector. Grasping and dragging (the
was somatotopically organized. two motor acts bringing the object toward the agent) pro-
duced activation in the ventral part of the putative human
3. Somatotopy or actotopy? AIP (phAIP), whereas pushing and dropping (the two motor
acts moving the object away from the agent) produced ac-
It is a well-established that the cortical motor system of tivation dorsally within the same area. This result suggests a
primates contains a double coding of motor behavior. In the specialization of the parietal lobe in representing action
primary motor cortex, neurons code mostly movements goals according to their behavioral valence, while the pre-
and they are organized in a somatotopic fashion with lower motor cortex encoding follows a certain degree of somato-

680 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

topy probably functional to the actualization of the motor there are no other mechanisms involved in this function.
acts into movements. Some mechanisms may rely on the association between a
given stimulus and its effect. For example, one can under-
4. Motor knowledge and the mirror mechanism stand a gesture conveying a threat, without necessarily
transforming it into a motor format. A monkey is scared
Is there a relation between personal expertise in performing
when it sees the experimenter throwing a stone towards it
some specific actions and the mirror responses to the obser-
even when the way in which the stone is thrown does not
vation of the same acts? Evidence for this was provided by
correspond to the way in which the monkey would throw it
Buccino et al. (47) who investigated in an fMRI experiment
(404, 405). This is not surprising because what counts here
the differences between observed motor acts performed by
is the painful effect of being hit by the stone rather than the
different species: human, monkey, and dog. Two types of
precise gesture mirroring.
actions were shown: biting or oral communicative actions
(speaking, lip-smacking, barking). Biting, regardless of
On the other hand, there is a long tradition that accounts
whom was the agent of the actions, produced bilateral ac-
for action understanding by referring to the capability of
tivations in the IPL-IFG circuit, virtually identical for three
individuals to “read” the mind of others, that is, to attribute
species especially in the left hemisphere. Communicative
a causal role to their mental states (e.g., beliefs and desires)
actions produced activation in the left caudal IFG that was
in executing actions. The nature and the format of this
dependent on species. Speech produced a strong activation
“mindreading” are still a matter of controversy (see Refs.
in Broca’s area and related areas, while barking produced
61, 246). There is, however, no doubt that human beings
none. These data show that owned motor behavior (in this
have this capability. There, however, is a fundamental dif-
case, biting) is represented in the observer’s motor system
ference between the mirror-based action understanding and
via a mirror mechanism even when performed by non-con-
the understanding of others’ behavior relying either on a
specifics. On the contrary, behavior that does not belong to
lower-order associative mechanism or on a higher-order
one’s own repertoire (as is the case of barking) does not
meta-representational capability. The mirror mechanism is,
produce mirror activations.
in fact, the only mechanism that allows understanding the
actions of others by activating the same circuits that the
In a further study, Calvo-Merino et al. (56) investigated the
individuals use when executing the observed action. This
mirror responses of three different groups of participants:
kind of understanding has been referred to as “understand-
classical dancers, teachers of Capoeira, and people naive to
ing from the inside” (337). The experiment by Buccino et al.
dancing. Stimuli consisted in videos of either Capoeira or
(47) that demonstrated the difference between action un-
classical dance steps. A clear double dissociation was found
derstanding belonging to human motor repertoire and that
between the two expert groups. Mirror responses to Ca-
of only other species clearly illustrates these different types
poeira steps were stronger in the Capoeira experts, and vice
of action understanding.
versa, mirror responses to classical dance steps were stron-
ger in ballet experts. In a further experiment the same re-
6. fMRI and repetition suppression technique
searchers (57) tried to disentangle the visual familiarity with
the dance steps from the motor expertise. They studied
Some studies used the repetition suppression technique dur-
classical ballet and examined the activation in men and
ing action observation and action execution in an attempt
women determined by the observation of steps done by
to characterize the proprieties of the mirror neuron system
dancers of the same and different gender. They found that
in humans. This technique assumes that stimulus repetition
the mirror system was activated more strongly by steps
reflects neuronal adaptation and that this effect can be de-
executed by individuals of the same gender of the observer.
tected using fMRI. One has to make a fundamental distinc-
This finding indicates that the motor expertise and not vi-
tion between fMRI experiments in which adaptation was
sual experience is crucial in activating the mirror system.
investigated by showing two visual actions in sequence (in-
The data by Calvo-Merino were extended by Cross et al.
tra-modal testing) and those in which the presentation of an
(87). In their study expert dancers learned new dance se-
action was either preceded or followed by an active execu-
quences for 5 wk. fMRI was weekly recorded while dancers
tion of that action by the participant (trans-modal testing).
observed the new dance sequences. The results showed that
the observation of another dancer’s movements activated
The results of intra-modal testing were reported in detail in
the mirror circuit. Critically, the activation of IPL and pre-
three studies. In one, the stimuli consisted of a series of
motor cortex correlated with dancers’ ratings of their abil-
intransitive motor acts (235). The results showed adapta-
ity to perform the new learned dance sequences.
tion in the left lateral occipital lobe and the left extrastriate
5. Understanding actions without using the mirror body area. None of the areas of the parietal and frontal lobe
mechanism known to be endowed with the mirror mechanism showed
repetition suppression. The main aim of the second study
The claim that the mirror mechanism plays a crucial role in (363) was to establish functional dissociations between the
understanding the behavior of others does not imply that dorsal and ventral visual pathways. It was tested whether

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 681


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

the repetitive presentation of object gasping would decrease easy to explain. Finally, it is worth stressing the lack of
the hemodynamic responses of the areas in the dorsal adaptation of visual responses in monkey F5 mirror as re-
stream (anterior intraparietal sulcus and premotor cortex) cently shown by Caggiano et al. (55).
related to action observation, while object identify those in
the ventral stream regions as the fusiform gyrus. The results 7. Imitation
supported the classical dichotomy between the two streams.
As far as the dorsal stream is concerned, they showed clear Is the mirror system involved in imitation? A recent meta-
adaptation in the condition “same object-same grasp,” but analysis of 139 imaging experiments (63) addressed this
a much weaker one in the condition “different object-same issue. The authors compared the parieto-frontal activations
grasp.” There are rich connections between inferior tempo- produced by passive observation of actions with those pro-
ral lobe and areas of the inferior parietal lobule (37, 282). duced by observing them to imitate. A largely overlapping
Thus it is plausible that the described fMRI adaptation circuit emerged, comprising the “classical” mirror areas
effect might have depended essentially on information on AIP/IPL in the parietal lobe and the premotor cortex and
objects and took place before information reached the pa- caudal IFG in the frontal lobe.
rietal lobe. In the third paper (99) in which adaptation
paradigm was employed, the participants played a game Note that there is ambiguity in the term imitation and, as a
(the “rock-paper-scissor” game) in the scanner, against a consequence, heterogeneity in the experimental tasks de-
videotaped opponent. A variety of visual and visuomotor vised to investigate its neural basis (see Ref. 172). One main
areas were found to decrease their activity with repetition of distinction to be made is that between immediate imitative
the same stimulus. It is not obvious, however, why these behavior and imitation learning. Immediate imitation refers
to online replication of an observed behavior. In contrast,
adaptation effects should be attributed to the activity of
imitation learning consists in building new motor memo-
mirror neurons. What counts in the “rock-paper-scissor”
ries, based on action observation. A second relevant distinc-
game is not the kinematics of the observed movements or
tion is that between imitation of action outcomes (also
the observed goal of the motor act (e.g., grasping), but only
called emulation, see Ref. 381) compared with imitation of
the final static hand shape (rock, scissor, or paper). Shape
the precise movements used to produce a given outcome. A
discrimination tasks have nothing to do with the “transfor-
final distinction is that between anatomical imitation and
mation” of observed motor acts or observed visual move-
specular imitation.
ments into a motor format, the core feature of the mirror
mechanism.
Prinz and co-workers (316, 318) extensively studied imita-
tion of simple motor acts. They found that tendency to
Rather surprisingly, all fMRI studies aiming at demonstrat-
repeat a motor act is greater when the motor act to be
ing the “existence” of mirror neurons employing adapta- copied is well represented in the observer’s motor reper-
tion have claimed that mirror neurons, if they exist, should toire. They proposed that there is a “common representa-
show “trans-modal adaptation.” For example, Lingnau et tional domain” for action perception and execution. The
al. (235) wrote: “Direct evidence for the existence of mirror discovery of mirror neurons suggested an alternative expla-
neurons would require finding adaptation in the condition nation for their findings, namely, that a motor mechanism,
in which execution is followed by observation.” This state- rather than an abstract, amodal domain is activated during
ment is rather curious because it postulates that the physi- the observed actions that have to imitated (317).
ological basis of the adaptation mechanism is cellular fa-
tigue, i.e., a decrease in the capacity of neurons to generate As already mentioned, the human mirror system encodes
action potentials. Aside from the fact that cellular fatigue is movements as well as goal-directed motor acts. Evidence on
a rather unlikely event in the case of a presentation of nat- this point was provided by Iacoboni et al. (176). They car-
ural stimuli, the data by Sawamura et al. (350) showed that ried out an fMRI study in which there were two experimen-
after repetitive presentation of the same stimulus, another tal conditions: “observation only” and “observation execu-
stimulus only slightly different from the first one is able to tion.” In the first condition, participants were shown a mov-
excite the putatively “fatigued” neuron. These findings are ing finger, a cross on a stationary finger, or a cross on empty
utterly incompatible with the assumption of the hypothesis background. They were instructed to observe passively the
of cellular fatigue. The mechanism leading to adaptation is stimuli. In the second condition, the same stimuli were pre-
synaptic fatigue of a specific input to a given neurons rather sented, but the participants were instructed to lift the right
than the general incapacity of neurons to respond to any finger in response to their presentation. The fundamental
stimulus. Hence, fMRI experiments trying to reveal the ex- comparison was between trials in which the subjects per-
istence of mirror neurons by resorting to trans-modal ad- formed the movement in response to an observed action
aptation paradigms should be considered very cautiously. (“imitation”) and trials in which the movement was trig-
Note that the findings of the existence of trans-modal ad- gered by the cross (a “nonimitative” behavior). Activation
aptation in human areas endowed with the mirror mecha- of the mirror system was significantly stronger during “im-
nism (202), although very interesting, are far from being itation” than during nonimitative behavior and stimuli pas-

682 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

sive observation. These results were subsequently con- (175). They carried out an fMRI study consisting of three
firmed by Koski et al. (210) and Grèzes et al. (155). conditions: 1) “context”; 2) “action”; and 3) “intention.”
In the first condition participants were presented with pic-
The mirror mechanism is also involved in imitation learn- tures of objects arranged as if a person was either ready to
ing. Some years ago Byrne (51) proposed an interesting have breakfast or had just finished it; in the second condi-
model based on his studies of ape behavior. According to tion, the participants saw the picture of a hand grasping a
his model, two distinct processes are at the basis of learning mug without any context; in the third, the participants saw
by imitation. First, the observer segments the action to be a picture of the same hand grasping a mug within the two
imitated into its individual elements, transforming it into a contexts. The context in which the action was performed
string of motor acts already present in the observer’s motor allowed the participants to understand the agent’s inten-
repertoire. Then, the observer organizes these motor acts tion. The data indicated that in both “action” and “inten-
into a sequence that replicates the observed one. It is plau- tion” conditions the mirror system became active. The com-
sible that an analogous process is also at the basis of learn- parison between conditions showed that the intention un-
ing nonsequential motor patterns. derstanding determined the strongest increase in the activity
of the mirror system.
Buccino et al. (48) investigated the neural basis of imitation
learning experimentally. In a fMRI study, participants who The neural substrate underlying our capacity to disentangle
never played guitar before were asked to imitate guitar between actions that reflect he intention of the agent (in-
chords played by an expert guitarist. The analysis was car- tended actions) and actions that did not reflect it (nonin-
ried out on four epochs: 1) observation of the chords made tended actions) was assessed in a fMRI study. Video clips
by the teacher, 2) pause, 3) execution of the observed showing a series of actions were shown to normal partici-
chords, and 4) rest. During passive observation there was pants in a double version each. In the first version, the actor
activation of IPL, PMv, plus the pars opercularis of IFG. achieved the goal of his/her action (e.g., pour the wine); in
During pause, the same circuit as during observation was the other, the actor failed to achieve it because of a motor
activated but, most interestingly, there was also blood flow slip or a clumsy movement (e.g., spill the wine) (45). It
increase in the middle frontal cortex (area 46) and in the turned out that the mirror circuits become active in both
anterior mesial cortex. Following Byrne (51), the authors conditions. However, if the activations present during the
proposed a two-step processing in imitation learning: first, intended actions were subtracted from those present in the
“mirror” activation of motor act representations in the pa- nonintended actions, increase of blood flow was found in
rietal and frontal lobe; second, the recombination, thanks the right temporo-parietal junction, left supramarginal
to prefrontal lobe (area 46), of these motor acts, so to fit the gyrus, and mesial prefrontal cortex. The opposite contrast
observed model. The same authors carried out a subsequent did not show any significant activation. The authors con-
fMRI study in expert and naive guitarists. This new study cluded that the understanding of nonintended action de-
confirmed the previous data. In particular, the data showed pends on the activation, besides the mirror circuit, of atten-
again the fundamental role of area 46 in combining differ- tion-related areas signaling unexpected events in temporal
ent motor acts in a new motor pattern (48). and spatial domains.

The issue of specular imitation versus anatomical imitation


was thoroughly addressed by Shmuelof and Zohary in two B. EEG Studies and Mirror Neurons
papers in which it was clearly shown that in AIP the identity
of the observed hand determines which of the two hemi- The activity of human sensorimotor cortex is characterized
spheres is predominantly activated (364, 365). The match- by a specific EEG rhythm called mu rhythm (136). This
ing between actor and observer is in this case strictly ana- rhythm belongs to the alpha frequency band (8 –13Hz) and
tomical, since the vision of a right hand activates systemat- exhibits a characteristic archlike appearance. The charac-
ically the left AIP, and vice versa, regardless of the left or terizing functional property of mu rhythm is its reactivity to
right field of presentation. Visual areas encoding body active movements (136). Furthermore, mu rhythm is
parts, but devoid of motor properties such as the occipito- blocked during action observation (137) and during motor
temporal cortex and the posterior superior temporal cortex, imagery (309). Because of its functional properties, it was
on the contrary, showed a laterality of activation that is proposed that mu rhythm could be used as an electrophys-
related to the visual field in which the hand is presented, iological marker of the mirror mechanism in humans. This
irrespective of the hand identity. possibility was first tested by Altschuler et al. (9) and thereafter
by many other researchers. These investigations supported the
8. Intention understanding hypothesis that mu rhythm dynamics reflect the activity of
human mirror system. A strong confirmation of this hypothe-
As in the monkey, the mirror system of humans is involved sis was also provided by an fMRI study in which a clear cor-
in understanding the intention of others. Evidence in favor relation was found between mu desynchronization and blood
of this conclusion was first provided by Iacoboni et al. flow increase in the motor region (15).

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 683


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

1. The reactivity of the mu rhythm: studies in adults modulation reflects a mechanism that filters information
into the mirror system allowing socially relevant informa-
There is vast literature on reactivity of mu rhythm to the ob- tion to pass.
servation of motor behavior. Before reviewing some (not all
for space reason) of these studies, it is important to have a Two issues, addressed with fMRI technique, were also in-
global picture of the time course of Rolandic cortical rhythms vestigated using EEG: the effect of motor practice on mirror
modifications during action observation. FIGURE 16 shows activity and that of actions describing sounds. As shown in
this picture based on a recent study by Avanzini et al. (18). fMRI studies, motor expertise of a given motor behavior
Four types of hand movements were presented to normal produces, during the observation of that behavior, in-
adults: two were target-directed motor acts (grasping and creased blood flow in the areas that belong to the mirror
pointing), and two were no-target directed movement (supi- system. Orgs et al. (292) addressed this issue presenting
nating and clenching). FIGURE 16 shows the results of the professional dancers and nondancers with videos showing
study. All stimuli determined a desynchronization of alpha dance movements and movements of everyday life. The re-
and beta rhythm in central and parietal regions. Note also the sults showed that power in alpha and beta frequency bands
large poststimulus power rebound present for all bands. Note was reduced when dancers watched dance movements, but
also the correlation between the velocity profile of the ob- not when nondancers watched the same movements. Dur-
served movement and beta band modulation, a finding indi- ing observation of everyday movements, no group differ-
cating a direct matching of the stimulus parameter on motor ence was evident. De Lucia et al. (91) recorded auditory
activity evoked potentials in response to sounds of man-made ob-
jects and in response of sounds that were not linked to
A very interesting EEG research branch concerns the effect motor actions. Analysis of source estimation identified dif-
of social interactions on Rolandic reactivity. Oberman et al. ferential activity within premotor and inferior frontal re-
(290) collected data while subjects viewed videos showing gion in response to sounds of actions typically cueing a
three actions: 1) an individual tossed a ball up in the air to responsive action.
themselves, 2) individuals tossed a ball to each other, and 3)
individuals tossed, occasionally, a ball toward the viewer. 2. Further EEG/MEG studies
The result showed that the mu rhythm was modulated by
the degree of social interaction. Actions where the ball was Another issue that was addressed using electrophysiological
tossed towards the viewer produced the strongest desyn- methods is imitation. In a pioneer work on this issue, Nishi-
chronization. Similar effects of social stimuli were also tani and Hari performed two studies in which they investi-
found by Perry et al. (307). A related issue was addressed by gated MEG activation during imitation of hand-grasping
Kilner et al. (201) using MEG. They recorded cortical ac- actions and facial movements, respectively. The first study
tivity of subjects while they watched a series of videos of an (285) showed the importance of the left IFG in imitation. In
actor making a movement recorded from different view- the second study (286), volunteers were asked to observe
points. They showed that the cortical response to action still pictures of lip forms, to imitate them, or to make similar
observation was modulated by the spatial relationship be- lip forms spontaneously. The mere observation activated
tween the observer and the actor. According to them, this sequentially a circuit formed by the visual cortex, STS, IPL,

2.0
fix still movement post-movement alpha
lower beta
FIGURE 16. EEG rhythms during action obser-
upper beta vation. The graph shows the EEG power time
course for each frequency band: alpha band
1.5 (8 –13 Hz) in green, lower beta (13–18 Hz) in
relative power

red, and upper beta (18 –25 Hz) in cyan. Each


epoch (fix, still, movement, and post movement)
is labeled by a different color. Significant differ-
ences between adjacent time bins are indicated
by asterisks whose color-code corresponds to
1.0 that used for different bands. At the top of the
figure, a film strip shows an example illustrating
the different epochs of the observed video clips.
[Modified from Avanzini et al. (18).]

0.5
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
time

684 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

IFG, and finally the primary motor cortex. The same se- stimulus. During quiet visual attention, a prominent central
quence was observed during both imitation and observa- rhythm was present at the age of 10, 14, and 24 mo with a
tion. During the spontaneous execution of the lip forms, peak frequency of 7– 8 Hz. By 51 mo of age, the peak
only IFG and the motor cortex were activated. frequency of the central rhythm had shifted to 9 Hz.

An original approach to the study of mirror system using The problem of the age at which infant mu rhythms respond
electrical recordings from cerebral cortex was used by Kil- to hand actions was addressed by Nystrom (287). He pre-
ner et al. (203). Instead of standard EEG recording, they sented 6-mo-old infants and adults with videos of a person
used slow potentials (“readiness potentials,” Ref. 208) to carrying out goal-directed actions (reaching and grasping
investigate whether the observation of grasping move- an object). While, as expected, adults showed greater mu
ments, would determine the occurrence of these potentials, rhythm desynchronization to the observation of these stim-
similarly to what observed during motor preparation. The uli compared with the observation of an inanimate moving
results showed that readiness potentials occur during action stimulus, there was no evidence of a similar differential
observation and that they appear before action execution. response in the infant EEG. More recently, Nystrom et al.
These data indicate that the knowledge of an upcoming (288) investigated EEG responses in 8-mo-old infants dur-
movement is sufficient to excite one’s own motor system. ing the observation of an experimenter performing a grasp-
ing action. Compared with the observation of a hand move-
In a recent EEG study, based on microstates analysis (220), ment not directed toward a particular goal, grasping obser-
Ortigue et al. (293) examined the temporal dynamics of vation produced a significant desynchronization in the 5- to
brain activations following observation of motor acts in 9-Hz frequency band. The desynchronization occurred
individuals instructed to understand the agent’s intention. around the time that the object was grasped.
The results revealed four sequential temporal steps during
intention understanding: 1) bilateral posterior cortical acti- Van Elk et al. (393) addressed the issue of mu reactivity in
vations, 2) strong activation of the left posterior temporal infants using another approach. Instead of showing hand
and inferior parietal cortices with almost a complete disap- action, they presented to 14- to 16-mo-old infants videos of
pearance of activations in the right hemisphere, 3) a signif- infants who were crawling or walking. Spectral power in
icant increase of the activations of the right temporo-pari-
the 7- to 9-Hz band desynchronized more during the obser-
etal region, and 4) a global decrease of cortical activity with
vation of crawling than walking (FIGURE 17). These find-
appearance of activation in the orbito-frontal cortex. The
ings clearly indicate that the motor experience has a signif-
authors concluded that the early activation of the left hemi-
icant influence on the mirror system responses. A well-prac-
sphere reflects the engagement of a lateralized mirror net-
ticed motor behavior as crawling has much stronger effect
work mediating goal understanding, while the subsequent
than the not yet learned walking behavior. Compelling ev-
later activation of the right hemisphere is due to the role
idence that infants are able to link sound to motor activity
that this hemisphere plays in others’ intention understand-
was provided by Paulus et al. (299). They trained 8-mo-old
ing.
infants to use a novel rattle that produced a specific sound
3. The reactivity of the cortical motor rhythm: when shaken. Infants were also presented with another
studies in infants sound not related to an action. EEG responses to the two
sounds were recorded. The results showed a stronger mu
A very interesting field in the research on mirror mechanism desynchronization during listening to the action-related
is that concerning the reactivity of the Rolandic rhythms to sound.
action observation and execution in infants. Relatively to
other imaging technique, EEG appears to be a very suitable The just reviewed studies provide an important insight into
and also easily available technique. There is, however, a the reactivity of the infant EEG to action observation. Their
preliminary problem. Are the rhythms recorded over corti- experimental design did not include, however, an active
cal central sites in children functionally homologous to condition. Two studies addressed this point. Fecteau et al.
those recorded on the same sites in adults? (113) used a subdural electrode grid to record EEG from a
36-mo-old child with epilepsy who was awaiting neurosur-
Early studies of the developing EEG have documented that gery. They examined two conditions: in one the child made
an alpha-like rhythm is present in children Rolandic areas. a drawing, in the other, the child watched an adult carrying
Its frequency is lower than that in adults and has the same out a similar action. The main finding was that alpha-range
frequency range as that of the infant alpha rhythm (6 –9 Hz) power at the electrodes over somatosensory cortex showed
(217; see also Ref. 371) A systematic investigation of the significant decreases in both conditions relative to a resting
infant mu rhythm was carried out by Marshall et al. (252) baseline. The second study that investigated action execu-
who examined the development of the EEG signal from tion was carried out by Lepage and Theoret (227). These
infancy to early childhood in a longitudinal study compar- authors recorded EEG during observation of repetitive
ing periods of quiet attention with periods of abstract visual hand actions either goal-directed (hand grasping) or not

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 685


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

FIGURE 17. Different EEG reactivity in response


to crawling and walking in infants. A: relative per-
centage of looking time for designated areas of in-
terest for crawling (left) and walking videos (right).
Values indicate looking times for observation the
videos. B: grand averaged EEG power for walking
B Mu-power (7-9Hz) Beta-power (17-19Hz) videos (dark bars) and crawling videos (light bars) at
different electrode sites (abscissae) for the mu-fre-
70 * Mu-power Crawl
14
Beta-power Crawl * quency band (7–9 Hz) and the beta-frequency band
Mu-power Walk 12 Beta-power Walk (17–19 Hz). Ordinate: voltage. [Modified from Vaa-
60 * dia et al. (393). Copyright 2009, with permission
10
* from Elsevier.]
50
µV2

µV2

6
40
4
30
2

20 0
FCz Cz C3 C4 LP RP FCz Cz C3 C4 LP RP

goal-directed (moving a hand kept flat) as well as during cortex. This excitability increase was interpreted as due to
active hand grasping. Observation of grasping was associ- the activation of mirror neurons located in the premotor
ated with greater mu rhythm desynchronization at central areas and connected with neurons of the motor cortex.
sites compared with observation of flat hand movements.
Action execution was also associated with a significant de- 1. Cortical origin of the mirror responses
synchronization of the mu rhythm over central sites. The
magnitude of mu desynchronization during action execu-
tion was greater than during action observation. The demonstration that the effects of action observation on
corticospinal excitability occurs at cortical level has been
An attempt to investigate the effects of social interactions provided by paired-pulse paradigms, which specifically test
on Rolandic reactivity in infants was done by Saby et al. the intracortical circuitry. In paired-pulse paradigms, the
(345). They explored the neural processes related to ten- conditioning stimulus (“cond-TMS”) and the test stimulus
dency of infants to prefer others who act like the self in (“test-TMS”) are delivered through the same coil on the
14-mo-old infants. EEG was recorded while infants were motor cortex. Several different paradigms based on the
observing actions that either matched or mismatched the same principle were employed. The classical one, originally
action the infant had just executed. Desynchronization of described by Kujirai et al. (212), investigated the effects of a
the EEG mu rhythm was greater when infants observed an sub-motor threshold “cond-TMS” on a suprathreshold
action that matched their own executed action. This effect test-TMS. At interstimulus intervals (ISIs) below 5 ms
was strongest immediately prior to the culmination of the “cond-TMS” produce short-latency intracortical inhibition
goal of the observed action. (SICI). At ISIs above 7 ms, “cond-TMS” produces short-
latency intracortical facilitation (SICF). Another paradigm
C. Transcranial Magnetic Stimulation and is the “long-latency paired pulse paradigm” (279). In this
Mirror Neurons paradigm both “cond-TMS” and “test-TMS” are suprath-
reshold. At ISIs around 100 ms, “cond-TMS” exerts long-
Magnetic stimulation transcranially applied (TMS) to the latency intracortical inhibition (LICI). Finally, the “I-wave
motor cortex evokes a volley of action potentials along the facilitation” paradigm (411) consists in stimulating the mo-
corticospinal pathways ultimately resulting in a contraction tor cortex with a suprathreshold “cond-TMS” followed by
of the muscles controlled by the stimulated cortical area. a subthreshold test-TMS at ISIs from 1 to 5 ms at submilli-
The electrical activity recorded from the activated muscles second steps. The interaction between the two pulses pro-
is referred to as motor evoked potential (MEP). Fadiga et al. duces peaks of facilitation that test the interneurons that
(111) showed that the observation of motor acts done by produce indirect waves (I-waves). For a review, see Refer-
another individual increases the excitability of the motor ence 410.

686 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

The first application of a paired-pulse paradigm to test the point of the motor act (4, 5, 390). In these studies the
motor cortex during action observation was carried out by significant experimental manipulation consisted in con-
Strafella and Paus (370) who used both the short-latency trasting factors as “movement type,” “spatial end point,”
and long-latency paradigms. Participants received TMS and “observer’s posture.” In Urgesi et al. (390), participants
while they were watching ongoing actions involving distal were shown movements of the index finger or of the little
muscles or proximal muscles. Recordings were made from finger of actors with their palm down or palm up, so that
an intrinsic hand muscle (1DI) and from a proximal muscle each of the two fingers could be on the left side or on the
(the biceps brachii). The results showed that MEP ampli- right side of the hand. The subjects’ muscular activity was
tudes from test-TMS pulses alone were modulated by action recorded from the muscles that contribute to the move-
observation with an effector specific pattern. The most in- ments of the index or of the little finger, namely, the first
teresting result came from the effects of the “cond-TMS.” interosseus and the abductor digiti minimi muscles. A sim-
An effector-specific change in short-latency intracortical in- ilar manipulation was used in Alaerts et al. (4, 5) but, in
hibition and facilitation was observed, thus showing that their protocol, the movements consisted in flexion or exten-
action mirroring is an exquisitely cortical phenomenon. sion of the wrist, in a palm-up or palm-down position. In
this way, the extrinsic spatial end point was the same for all
A further study (205) found specific cortical effects of action acts, and it was positioned along a far-near axis. The spatial
observation using the “I-wave facilitation” paradigm. The endpoints of the movements were therefore not defined in
results showed a clear increase in the second I-wave facili- terms of retinotopic space (left or right to the body part). All
tatory peak. Modulations of this peak reflect intrahemi- these studies obtained similar results, namely, the predom-
spheric corticocortical connections to the motor cortex inance of muscle-specific encoding of observed actions,
(72), presumably from the premotor cortex. Taken to- with a minimal or no contribution of spatial and postural
gether, these results unequivocally indicate that the site of features.
modulation of the corticospinal system during action obser-
vation is intracortical. The question of whether also a mod- A different picture emerges from studies in which the ob-
ulation at spinal level exists has been explored in a very servation of transitive (goal-directed) actions was investi-
limited number of studies and no definitive conclusions can
gated. These studies showed that, as in the monkey, human
be drawn at the moment (39, 298). Also the issue of hemi-
motor cortex encodes motor acts (68, 70, 101, 105, 106).
spheric specialization for action observation has been the
An experimental paradigm that has helped to disentangle the
object of few studies. They all point at a predominant role
goal of the motor act from the movements used to achieve it
of the left motor cortex (20, 21, 114).
has been the observation of actions done using tools. This
paradigm, derived from monkey experiments (387), was ad-
2. Single pulse TMS of the motor cortex
opted in humans in a study with single-pulse TMS over M1
A) THE BODILY AND SPATIAL FRAMES OF REFERENCE OF MIRROR
(68). The experimental setup involved two types of tools (“di-
ACTIVATION. As mentioned above, the modulation of the rect” and “inverse” pliers, FIGURE 18) that required opposite
observer’s motor system mimics the muscular pattern movements to achieve the same goal (grasping and releasing).
used by the agent to perform the movements. The range The results showed that when observing goal-directed grasp-
of muscle specificity varies depending on the experimen- ing, the observers’ motor system encoded its goal regardless of
tal paradigms. One study found an exact replica of the the tool used and, therefore, irrespective of the movements
observed muscular pattern in hand and forearm muscles performed to achieve it (FIGURE 18). These results were not
(244). Other studies reported a less strict observation- found when observing intransitive movements. In this case the
execution picture. They showed that motor mirroring observer’s motor system mirrored the activity of single mus-
was muscle specific, but the effect of movement observa- cles, irrespective of the tool used. These findings were recently
tion was not limited exclusively on the “prime mover” confirmed in a second study employing a similar “inverse pli-
muscle but also involved other muscles along a gradient ers” paradigm (70).
of functional proximity to the observed movement (251).
Finally, a TMS study by Gangitano et al. (134) showed A related topic is that of self-attribution of observed stimuli
that the observers’ motor cortex is modulated with a that produce motor mirroring. It was shown that the ob-
muscle-specific pattern when observing different phases server’s sense of ownership of an observed body part re-
of a reach-to-grasp act. duced and even turned into inhibition the phenomenon of
motor mirroring (354, 355).
A more complex issue is whether motor resonance reflects
the observed behavior in terms of movements or in terms of B) MOTOR IMAGERY AND MIRROR ACTIVATION. There are two
action goal. There is evidence that in certain conditions, types of mental imagery of actions (13): visual and motor
namely, during the observation of intransitive movements, imagery. In the first case, one imagines seeing himself or
the net output of the observer’s motor system is expression another person performing actions from an “external view-
of the observed movements rather than of the spatial end point.” In the second case, one imagines himself performing

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 687


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

2.0

1.8

mean MEP amplitude (mV)


1.6

1.4
pliers

1.2

1.0

0.8

0.6

0.4
extend flex

extend fingers flex fingers 2.2

2.0

mean MEP amplitude (mV)


1.8

1.6

1.4
pliers

1.2

1.0

0.8

0.6

0.4

0.2
extend flex
FIGURE 18. Evidence for action goal coding by the mirror mechanism. Participants watched the experi-
menter grasping objects with two types of pliers: direct pliers and reverse pliers (left panel). Finger flexion
produced closure of the direct pliers, but opening of the reverse ones (right part of the left panel). Finger
extension produced opening of the direct pliers, but closing of the reverse ones (left part of the left panel). The
right panel shows MEP amplitudes obtained from the opponens pollicis muscle during the observation of
different movement phases. Right panel, top: MEPs during the observation of grasping with direct pliers. Right
panel, bottom: MEPs during the observation of grasping with reverse pliers. The observers’ motor systems
encode the goal of the action, regardless of whether it is achieved by closing or opening the hand. [Modified
from Cattaneo et al. (68).]

actions. In both cases mental imagery of actions is a volun- C) CHRONOMETRY OF MIRROR RESPONSES. TMS has a temporal
tary process. This must be kept distinguished from the au- resolution of submillisecond magnitude (411). It is there-
tomatic motor mirroring that occurs during action obser- fore an optimal tool for establishing the chronometry of
vation (185). cortical processes. This advantage of TMS has been ex-
ploited to investigate the precise time course of motor mod-
Various studies using TMS over the motor cortex showed ulation from the onset of an observed movement. What is
that observing or imagining the same acts produces a vari- the delay between visual presentation of an action and mo-
ety of different patterns of motor modulation in the observ- tor system activation? Convergent data on this issue come
er/imaginer (117, 226, 232, 384, 385). A crucial demon- from four independent studies. In the first, Catmur et al.
stration of such difference was provided by Cattaneo et al. (64) showed that mirror activation occurs 200 ms from the
(68). Employing the inverse pliers paradigm (see above), they onset of the observed movement and is still present at 300
reported a clear difference between imagery and observation ms. In the second, Barchiesi et al. (26) reported that motor
of the same goal-directed actions. The observation of goal- facilitation, absent at 150 ms from stimulus onset, was pres-
directed actions produced a motor response that mimicked the ent at 250 ms as well as 320 ms. The above-mentioned
goal of the action. On the contrary, the imagination of the works tested the observers’ motor systems during passive
same goal-directed actions produced a motor facilitation that observation of actions. A recent work (386) showed that if
mimicked the muscular activity required to perform the task. the mirror phenomenon is probed during a task that re-
These data indicate that motor imagery is essentially a covert quires a hand responses to observed actions, a response can
preparation of movement and, as such, is more closely linked be evident as early as 150 ms from the observed movement
to motor execution than motor observation. onset.

688 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

The knowledge of time of occurrence of mirror responses nitive decisions) that has the capacity to establish arbitrary
during action observation is of great importance for studies associations between visual cues and motor responses
that investigate whether mirror responses might be modi- (FIGURE 19, bottom). The timing of appearance of motor
fied by the so called “counter-mirror” training. This train- responses following action presentation would therefore be
ing consists of instructing subjects to respond as fast as an important physiological signature that identifies which
possible to an observed movement with a different move- of the two systems is recruited.
ment. In the first study with this task (65), it was found that
motor responses to action observation, evaluated the day Ubaldi et al. (386) recently confirmed the existence of the
after the counter-mirror training, displayed an inverted biphasic response pattern in the observer’s motor system
mapping relative to the original response. A subsequent during action observation. In this work, rather than the
study by Catmur et al. (64) confirmed these findings. In this after-effects of counter mirror training, the authors tested
study, the motor responses were recorded 300 ms following the motor modulations to action observation during coun-
the onset of the stimulus presentation (64). Barchiesi and ter-mirror training. The results confirmed that a mirror re-
Cattaneo (26) tested the “reconfiguration” hypothesis ap- sponse, independent of the arbitrary visuomotor rule, was
plying stimuli early and late following stimulus onset. They evident in an early time window (150 ms) from the onset of
found that in an early time segment corresponding to 250 an observed movement, while rule-dependent responses ap-
ms from the onset of the observed movement, the partici- peared at 300 ms from movement onset. The authors fur-
pants’ responses were not influenced by training. Only in a ther extended the concept that the early stimulus-driven
later time segment (at 320 ms) the participant’s responses (mirror) response and the late rule-driven (counter-mirror)
became “counter-mirror” (FIGURE 19, top). These data response are two independent phenomena by applying re-
show that two neural systems concur in producing re- petitive TMS (rTMS) to the posterior parietal cortex and to
sponses to action observation: a fast and stable system (the the prefrontal cortex in two separate groups of subjects. A
mirror system) that is not influenced by recent visuomotor clear double dissociation was found between the target of
experience and a slower system (most likely related to cog- rTMS and the early and late components of the motor re-

A baseline
post-training
CW
observer’s wrist
acceleration

FIGURE 19. Early and late motor responses to ac-


tion observation. A: TMS-evoked accelerations ob-
tained from participants while watching a hand rotat-
ing a lid clockwise (CW) or counter-clockwise (CCW). In
CCW

baseline (without training) conditions (red segments), a


mirror response (CW accelerations when observing a
CW movement and CCW accelerations when observ-
CCW CW CCW CW CCW CW CCW CW
ing a CCW movement) is present at 250 and 320 ms
from the onset of the observed movement. After the
0 ms 100 ms 150 ms 250 ms 320 ms “counter-mirror” associative training (blue segments),
the mirror responses persist at 250 ms from move-
ment onset but become “counter-mirror” at 320 ms.
B B: model explaining the time course of motor re-
sponses to action observation after the “counter mir-
counterimitative

ror” training. Two separate processes interact and


response

compete for the motor output. A fast automatic pro-


cess (referred to as mirror function, dotted line) oc-
curs earlier than 250 ms from stimulus onset and
motor output

produces mirror responses regardless of the training.


An associative process due to the training determines
the counter-mirror response. This response (executive
function response, dashed line) occurs later and over-
comes the mirror function response. The arithmetic
summation of the two functions produces a net motor
response
mitative

output (solid line) that has a biphasic time course as


the one observed in the empirical data described in A.
mirror function [Modified from Barchiesi and Cattaneo (26), by per-
executive function mission of Oxford University Press.]
net motor output

0 ms 100 ms 150 ms 250 ms 320 ms

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 689


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

sponse. Parietal rTMS influenced only the early stimulus- sual expertise. The results from the literature show that
driven response; prefrontal rTMS affected only the late both these trainings can induce long-term changes in mir-
rule-driven response. roring (2, 119, 191).

In a related topic, several studies investigated whether mir- The second approach is to investigate the effects of brief
ror responses recorded during predictable on-going natu- sessions of motor training in specific tasks that were previ-
ralistic movements anticipate the observed movements or if ously unknown to participants. Motor practice is known to
they lag passively behind the observed action. All studies induce online effects and short-latency aftereffects on ac-
employing on-going observed actions obtained a clear indi- tion observation (27, 62, 67; for a review, see Ref. 356).
cation that whenever the action is predictable, the motor Such effects are generally attributed to the mirror neuron
activation has anticipatory dynamics. One of these studies mechanism. One study investigated the evolution in time of
employed cyclic intransitive wrist movements (270). Motor practice-induced changes in action mirroring in a popula-
responses recorded from the observer’s wrist extensors and tion of amateur piano players after a short practice of a
flexors fitted into a sinusoidal function that matched the piano composition. Participants exhibited a consistently
sinusoidal variations of the observed wrist angle. More in- greater activation of their motor system when listening to
terestingly, when comparing the observer’s and the agent’s the rehearsed piano tunes relative to nonrehearsed piano
sinusoidal functions, a phase shift was evident. The observ- tunes. Such auditory-motor facilitation occurred early after
er’s sinusoidal function anticipated that of the agent by 60° rehearsal and increased over days with consolidation of the
to 90° of a full oscillatory cycle. In another experiment, the motor expertise (88). Stefan and co-workers (368, 369)
time course of the observers’ mirroring of an actor’s grip investigated the issue of motor memory formation by exe-
force was tested during observation of a reach-grasp-lift cution alone, observation alone, or simultaneous observa-
sequence. It was shown that the observer’s motor activation tion and execution. It was shown that a training period in
predicted the actor’s grip force well before the occurrence of which participants simultaneously performed and observed
the actual grasp (3). These results further expand the find- congruent movements potentiated motor cortex plasticity
ings by Gangitano et al. (133) who showed that the MEPs with respect to motor training alone, as shown by the kine-
recorded during observation of predictable actions actually matics of the movement evoked by TMS (see FIGURE 20).
mirror the predicted trajectory of the observed hand rather
than the movement that is actually happening. E) ARTICULATORY RESONANCE TO SPEECH SOUNDS. TMS studies
showed that the hand-related motor system responds to
D) MOTOR EXPERIENCE AND MOTOR RESONANCE. Experience of a sounds produced by manual actions (20, 88). A particular
certain motor skill appears to be of importance in influenc- case of motor activation to sounds is that occurring in the
ing the mirror responses to observation of actions requiring articulatory system when people listen to speech sounds.
that skill. In the literature, two different experimental ap-
proaches have been taken to handle this problem. The first In adults exposed to verbal-related sounds, there is a fast
is to investigate populations of experts, (e.g., dancers, elite automatic activation of a corresponding orofacial motor
sportsmen) who are skilled in determinate actions. Experi- program. Articulatory resonance to speech has been de-
ments dealing with mirroring in hyperspecialized popula- scribed in several papers by recording changes in MEPs to
tion have to disentangle the contribution of motor and vi- single-pulse TMS applied to the representation of tongue

TMS-evoked extension execution of observation of


movements flexion movements flexion movements
40 *
movements towards
the trained direction
% of TMS-evoked

30 *
20

10

0
motor observational
training training

baseline individual individual motor cortex tuning


motor or observational training
motor cortex tuning after training
FIGURE 20. Formation of motor memory by action observation. Participants were tested with single-pulse
TMS over the motor cortex (first panel: baseline conditions; thumb extension). They underwent subsequently
either a motor training or an observational training in which they had to perform or to observe movements
opposite to the baseline motor tuning, respectively (second panel: thumb flexion). After training, the partici-
pants’ motor cortex tuning was assessed by single-pulse TMS over the motor cortex. Both movement and
observation produced a shift of TMS-evoked movements towards the trained direction (open columns, base-
line; closed column, training effect). The results indicate that action observation has a direct access to the
motor system. [Modifed from Stefan et al. (369).]

690 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

movements (110, 342) and of lip movements (372, 399, main finding of the study was that virtual lesions of IFG
400). Speech motor resonance is effector specific, i.e., it produced by rTMS suppressed mirror facilitation of
matches the articulatory gestures that are required to pro- MEPs when observing possible movements, thus estab-
duce the heard sounds. It occurs very early during speech lishing a causal role between the IFG and the mirror-
processing, and evidence indicates that this early resonance related motor activation. A similar approach was used in
(100 –200 ms) does not require semantic comprehension of another experiment (198) in which the authors investi-
the verbal material (342), nor does it reflect the final per- gated the effects of rTMS, applied over the IFG, on the
ceptual stage. Finally, motor responsiveness correlates neg- reactivity of the mu rhythm to action observation. The
atively with the perceptual clarity of the phonological stim- results showed that rTMS reduced the suppression of the
uli (89, 274), i.e., it is enhanced by adding noise to the cortical rhythm to action observation.
stimuli.
The role of the IFG and of PMv in the genesis of mirror
3. Beyond the motor cortex: definition of the mirror responses has been also investigated in a study in which
network with TMS rTMS was applied to the pars opercularis of the IFG bilat-
erally. The results showed that rTMS impaired an imitative
The experimental approach of measuring MEPs to TMS on task but not a nonimitative visuomotor task (165). Pobric
the motor cortex offers the great advantage of unequivo- and Hamilton (310) performed a study in which partici-
cally probing visuomotor integration. However, it cannot pants were asked to make a weight judgment task on a box
provide any information on the neural circuits outside the that was lifted by an actor with different kinematics or on a
motor cortex. One possible way to overcome this problem ball bouncing in the absence of a biological agent. It was
is to apply TMS to other cortical areas so to condition the
shown that rTMS on the IFG produced a deficit in inferring
response of the motor cortex to single-pulse TMS. In these
the object’s weight from the kinematics of the hand move-
paradigms, an effect of the conditioning stimuli on the
ments. On the contrary, the judgment on the kinematics of
MEPs is considered as an index of functional connectivity
the bouncing ball remained unaffected. Along the same line,
between the stimulated area and the motor cortex. Further-
another study found that rTMS of left BA44 impaired indi-
more, TMS can be used to stimulate areas different from the
vidual performance in evaluating only biological actions,
motor cortex with a subsequent measurement of the behav-
and more specifically object-oriented actions (78). Stimula-
ioral effects of such interventions.
tion of PMv produced deficits in an action matching task
(391) but, interestingly, did not impair a body-form match-
A) THE “DUAL COIL” TMS PARADIGMS: TESTING CORTICO-MOTOR
CONNECTIVITY DURING ACTION OBSERVATION. The “dual coil” ing task.
paradigms investigate the effect of a “cond-TMS” pulse,
C) VIRTUAL LESION PARADIGMS: POSTERIOR TEMPORAL CORTEX.
delivered with one coil to one cortical site, on the responses
to a “test-TMS” pulse delivered with another coil to the The posterior superior temporal cortex, and in particular
motor cortex. This paradigm tests direct cortico-cortical the cortex in the banks of the posterior superior temporal
pathways from the site of “cond-TMS” to the motor cortex. sulcus (pSTS), contains cells that encode biological mo-
The ISIs, commonly employed, are between 4 and 8 ms tion (see above). Coherently with these data, TMS ap-
when intrahemispheric pathways are tested (66) and plied to the pSTS region disrupts the processing of up-
around 10 ms when transcallosal pathways are investigated right point-light displays of whole bodies moving (158).
(115). A study using this paradigm examined the effects of Because the pSTS region is the main source of visual
a “cond-TMS” applied to the ventral premotor cortex or to information feeding into the parieto-frontal mirror sys-
the anterior intraparietal sulcus on “test-TMS” applied to tem, TMS applied to the pSTS modulates the parieto-
the motor cortex during action observation (205). Partici- frontal mirror mechanism (60). One paper explored the
pants were shown movies of a hand grasping either a small functional connections between the pSTS and the pari-
or a large object. In half of the movies the hand shape was etal-premotor regions by temporarily perturbing the ac-
appropriate to the object size, while in the other half it was tivity in the pSTS cortex with 1-Hz rTMS and measuring
not. In those cases in which the hand shape was appropriate the hemodynamic responses to action observation (14).
to the object size, there was a specific activation of the The data showed that the effects of rTMS over the pSTS
cortico-cortical pathways, while there was no effect when were evident in 1) the lateral occipito-temporal cortex, 2)
there was an inappropriate hand shape. in the anterior intraparietal cortex, and 3) in the premo-
tor cortex, thus confirming that the pSTS occupies a hub
b) VIRTUAL LESION PARADIGMS: IFG, PMV AND MOTOR CORTEX. position between the early visual areas and the parieto-
Avenanti et al. (19) applied prolonged repetitive 1-Hz frontal mirror circuit.
TMS (rTMS) over the caudal sector of IFG and the so-
matosensory cortex (S1). Subjects were subsequently D) TMS-ADAPTATION PARADIGMS. TMS-adaptation paradigms
tested with TMS of the motor cortex while watching are designed to test the presence of subpopulations of neu-
biomechanically possible or impossible movements. The rons that are spatially overlapping but with different func-

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 691


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

tional properties (366). They take advantage of the well- sented in the left than in the right hemisphere (see above),
established psychophysical phenomenon of perceptual ad- clinical studies may provide some information about a pos-
aptation. After adaptation, a subject’s perception is biased sible causative link between motor and perceptual deficits
against the adapted stimulus. A TMS pulse over the adapted following left hemisphere lesions.
neurons enhances selectively the perception of the adapted
features. This technique allows one to localize on the cortex Evidence for associations between motor and perceptual
the subpopulation of neurons encoding that feature. The disturbances has been reported in patients with diagnosis of
neural mechanisms by which TMS produces state-depen- apraxia. Apraxia, however, is a higher order cognitive/mo-
dent effects are still debated (263, 305, 357). tor disturbance that only occasionally is associated with
perceptual disturbances. Thus it is important to discuss the
One study employing a TMS-adaptation paradigm ex- fundamental characteristics of apraxia and the location of
plored the possibility to induce cross-modal motor-visual lesions producing perceptual deficits in patients with this
adaptation (67). Repeated performance of the same mo- syndrome, before concluding a link between the mirror
tor act produced adaptation-like phenomena in the visual mechanism and apraxic deficits.
categorization of actions. Event-related TMS applied to
the PMv abolished the adaptation effects, while sham The term apraxia designates the impaired ability to perform
stimulation and active TMS over M1 did not. TMS-ad- motor activity, in spite of preserved motor, somatosensory,
aptation paradigms have been also used with purely vi- and coordination functions in the effectors engaged in the
sual action stimuli, to investigate the tuning properties of action (93, 164). The characterizing feature of apraxia is an
the different nodes of the action observation system. In automatic-voluntary dissociation. The same gesture that
one TMS study (71) participants underwent visual adap- the patient is unable to perform on the examiner’s request
tation to observed actions by repeatedly seeing one of is, typically, correctly performed if done in response to a
two actions (pushing or pulling) that were performed habitual stimulus or context. There are various classifica-
with one of two different effectors. Subjects were subse- tions of apraxias. A fundamental distinction lies between
quently tested in a match-to-sample action recognition the ideational apraxia (IA) and the ideomotor apraxia
task, while they were stimulated over different cortical
(IMA). Patients with IA do not know what to do, i.e., they
areas. TMS selectively enhanced the subjects’ perfor-
are unable to retrieve the movement schema typical for
mance when judging the adapted action irrespectively of
certain object; in contrast, patients with IMA are unable to
the effector when it was delivered over the left or right
translate the observed or instructed movement into appro-
PMv or over the left supramarginal gyrus. On the con-
priate motor pattern.
trary, when TMS was delivered onto the pSTS region, it
improved performance only in processing stimuli that
The form of apraxia that has gained most attention in the
shared both the same action and the same effector with
last years as possibly related to dysfunction of the mirror
the adaptor sequence. It was concluded that the premotor
mechanism is IMA (221). The core IMA symptoms are 1)
and parietal cortices contain neurons that were adapted
deficits in performing tool-use pantomimes on verbal com-
(and therefore encode) the actions in an abstract way,
irrespective of the effector that performs them. On the mand or imitation, e.g., pantomiming to hammer a nail or
contrary, the temporal cortex contains neurons that rep- to brush one’s teeth, and 2) deficits in performing symbolic
resent actions only linked to the very same body part that communicative gestures, as in waving good-bye, on verbal
is performing them. It was concluded that action repre- command or imitation. Patients can be impaired in panto-
sentation is hierarchically organized, with a low-level mime/gesture performance for both verbal command and
pictorial representation in the pSTS cortex and a more imitation or for one of this task only (94, 302). The capacity
abstract representation in the parieto-frontal system. to perform actions on real objects spontaneously or on com-
mand is typically spared, although it can be occasionally
impaired (49, 183).
D. Disorders of Mirror Function
Typically, apraxia is not accompanied by disturbances in
1. Acquired disorders of the mirror mechanism action recognition (93). Heilman and co-workers (164,
341) described, however, cases of patients with IMA that
An important issue is whether lesions of the frontal and showed, besides the classical symptoms, also disturbances
parietal areas endowed with the mirror mechanism are in recognizing the actions done by others. This observation
causative of incapacity to understand actions done by oth- raises two issues. First, why in most cases apraxia is not
ers. In the monkey, the mirror system is equally represented accompanied by perceptual deficits? Second, what may be
in both the right and the left hemisphere. Lesion studies are the neural basis of these deficits?
therefore impossible to carry out without determining se-
vere cognitive deficits. The same difficulty is present in hu- To answer the first question, let us examine what is the core
mans. However, being the mirror mechanism more repre- deficit in apraxia. Regardless of whether the required action

692 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

is elicited by a verbal command or by imitation, the initial it. As stated by Jeannerod (185), motor imagery is nothing
operation that the patient has to do is that of forming a else but implicit motor preparation.
motor imagery of the requested behavior. Motor imagery is
not a mere visual copy of an action, but includes the concept The hypothesis about the circuits underlying IMA well
of self (“I am doing the action”). This profoundly differen- explains why in this syndrome only deficits in the motor
tiates motor imagery from visual imagery (see also above). sphere are typically observed. Deficits in action under-
Once the motor imagery is generated, it has to be translated standing should appear only when lesions in IMA pa-
into an overt behavior. Evidence from fMRI studies showed tients would extend beyond IMA core circuit. If we dis-
that when an individual is required to generate motor im- regard, now, the apraxia framework and consider more
agery of symbolic, mimed, or meaningless gestures, there is simply the presence or the absence of association of mo-
a strong activation of Broca’s area and of the anterior pole tor and perceptual deficits in patients with parietal or
of the temporal lobe (FIGURE 21). These activations are frontal lesions, there is clear evidence that such an asso-
accompanied by activation of the ventrocaudal part of the ciation does exist. A correlation, in fact, has been repeat-
supramarginal gyrus. In addition, there is an activation of edly found between deficits in gesture/pantomime under-
the dorsal part of supramarginal gyrus around the intrapa- standing and gesture/pantomime production (50, 281,
rietal sulcus and of the premotor cortex. Note that, unlike 300). In particular, Pazzaglia et al. (300) examined an
the ventrocaudal parietal activation, the dorsal supramar- action recognition task (hand and mouth action-related
ginal activation, as well the premotor one, are also present sound) in patients with limb or buccofacial apraxia. The
when the individuals simply observe the same gestures with- analysis of the lesion showed that left frontoparietal cor-
out any further behavioral request. tex has a fundamental role in the recognition of the sound
due to limb movements. On the contrary, left IFG and the
On the basis of these findings, the following hypothesis on adjacent insular cortex were causatively associated with
the genesis of IMA can be advanced. The deficit in IMA recognition of sounds generated by the buccofacial appa-
patient resides in the incapacity to produce the motor im- ratus. This double dissociation demonstrates that differ-
agery of the requested action or to translate the imagined ent sites endowed with the mirror mechanism encode
actions into related motor program. Lesions of Broca’s re- limb and mouth action-related sounds and that these sites
gion would be responsible of the incapacity to generate are also involved in limb and mouth execution.
motor imagery, while damage to the ventrocaudal part of
the supramarginal gyrus, or, as already postulated by Liep- In conclusion, IMA is a syndrome related to the incapac-
mann (233), disconnections between Broca’s area and the ity to elicit a motor imagery of the requested action and
parietal lobe, would be the cause of the incapacity to trans- to execute it. Automatic-voluntary dissociation is its hall-
late motor imagery into a motor program. This explanation mark, as already stressed by Liepmann (233). The pres-
is valid not only for action elicited on verbal command, but ence of perceptual deficits (e.g., action understanding)
also in the case of imitation. Also, in this case the patient are not related to apraxia itself, but rather to an exten-
tends, before executing the action, to imaging himself doing sion of the lesion to parietal and premotor areas endowed

all

pantomime
transitive FIGURE 21. Activation during motor imagery. Maps of cor-
tical areas during imaging meaningless intransitive action,
symbolic gesture, and pantomime of goal-directed action. Top
panel indicates the activations obtained pulling together the
symbolic activation of the three conditions mentioned above.
gesture

meaningless
intransitive

p<0.001 unc.

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 693


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

with the mirror mechanism. Thus the deficit in action 2. Developmental disorders of mirror function:
understanding and action production, described in pa- autistic spectrum disorders
tients with parietal and frontal damage, represent clear
evidence for a causative relation between action produc- Autistic spectrum disorder (ASD) refers to a wide contin-
tion and action recognition in humans. uum of associated cognitive and neurobehavioral disorders
characterized by impairment in communication and social
B) HYPERACTIVITY OF THE MIRROR MECHANISM. The potential interactions, and restricted repetitive and stereotyped be-
motor acts resulting from activation of the mirror mecha- haviors (10, 17, 195). In ASD, a variety of nervous struc-
nism ought to be subject to an inhibitory control. Other- tures are affected, ranging from brain stem to cerebellum
wise, movements triggered by the observation of motor acts and cerebral cortex (42, 83, 124, 273). As far as the cortical
abnormalities are concerned, of particular interest is the
done by others should appear. Evidence for the existence of
connectivity deficit found in the parieto-frontal network,
a top-down executive control on the mirror mechanism
i.e., in the circuit whose areas are endowed with the mirror
comes from experiments on healthy volunteers (40, 135,
mechanism (159). It has been proposed that a malfunction-
269, 386), but the most compelling evidence for such con-
ing of the mirror mechanism is one of the factors underlying
trol derives from clinical observations of the release of au-
the cognitive aspects of the deficit (402).
tomatic imitative behaviors. In such conditions, motor mir-
roring becomes an overt, uncontrollable response that is Most studies on the relationship between mirror mecha-
automatically elicited by seeing other people acting. nism and autism have employed EEG and MEG techniques.
They revealed that the cortical rhythms that desynchronize
Patients with damage to the prefrontal lobes may exhibit in typically developing children (TD) during both execution
behavioral disturbances characterized by the appearance and observation of hand movements do desynchronize in
of forced motor behavior triggered by external stimuli. children with autism only during active hand movements
This condition is broadly referred to as “environmental (255, 289, 321).
dependency syndrome” (230, 231). A manifestation of it
is the “imitation behavior.” Typically this symptom is Recently, in a MEG study, Honaga et al. (170) explored
observed in patients with unilateral or, more frequently, neural activity in ASD focusing on power increase in beta
bilateral prefrontal lesions (92, 169, 231, 374) but has frequency band (see EEG section, beta “rebound”) rather
been also described in patients with different lesion to- than the rhythm desynchronization. ASD and control chil-
pography (148, 266). Echopraxia is a similar, but distinct dren were asked to observe and later execute object-related
clinical phenomenon. It consists of forced and uncritical hand actions. They found significantly reduced rebound in
imitation of others’ movements sustained through endog- ASD only during the observation condition. The absence of
enous mechanisms and resulting in its perseveration. Im- motor cortex activation during the observation of move-
itation behavior in broad sense is not a frequent symp- ments done by others strongly suggests a deficit of mirror
tom. The few descriptions, however, seem to agree on the system.
fact that it contains a spectrum of manifestations accord-
ing to how much the behavior is voluntary or automatic. Oberman et al. (291) investigated the sensitivity of mu
While some forms of imitation behavior could seem to be rhythm desynchronization in relation to the familiarity of
dependent somehow on the patient’s will, other forms, the agent performing the observed action. There were four
conditions: an unfamiliar hand grasping an object
including echopraxia, appear to be totally involuntary
(“stranger”), the hand of the child’s relatives performing
(92, 362, 374).
the same action (“familiar”), the participant’s own hand
(“own”), and, finally, bouncing balls (“control condition”).
An interesting clinical phenomenon related to a release of
Both TD and ASD children showed greater suppression to
motor plans from an executive control is the so-called “an-
the observation of grasping made by familiar hand com-
archic hand.” It consists in the appearance of unintended pared with that of strangers. The desynchronization ap-
complex movements of one hand that are directed towards peared to be absent in ASD children, but only with unfamil-
a goal. Interestingly, the awareness of the patients of the iar stimuli.
movements is preserved. What is lost is the sense of agency
of the limb that behaves as if it had “a will of its own” (96, Imitation ability has been frequently found to be impaired
213). The anarchic hand is associated with lesions of the in individuals with autism (see Refs. 339, 402). It was sug-
anterior corpus callosum or the mesial frontal cortex (249). gested that a dysfunction of mirror mechanism might ac-
The anarchic hand should be kept distinct from the so- count for this impairment (see above). In a study performed
called “alien hand,” which is characterized by a degraded on ASD individuals and age-matched typical adults, Bernier
sense of ownership of the affected body part and is typically et al. (33) found that while both groups showed a clear
associated with parietal rather than frontal lesions (249, attenuation of the mu rhythm during action execution, TD
294). individuals, but not ASD individuals, also showed signifi-

694 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

cant attenuation during action observation. The degree of A different approach to the problem of possible impairment
mu wave attenuation during observation correlated with of the mirror mechanism in ASD was employed by Cattaneo
imitation capacities. A deficit in imitation was also found in et al. (69). They recorded electromyogram (EMG) from
a MEG study on adults with ASD (284). In contrast to these mylohyoideus muscle (MH), one of the muscles involved in
findings, Raymaekers et al. (326) were unable to find differ- the mouth opening, in TD and ASD children during the
ential mu suppression between ASD and TD children. These execution and observation of two actions. One action con-
results are surprising because they used the same paradigm sisted of grasping a piece of chocolate and bringing it to the
and stimuli employed by Oberman et al. (289). The authors mouth (eating action), while the second consisted of grasp-
explained this discrepancy highlighting the heterogeneity of ing a piece of paper and putting it into a container (placing
ASD population. action). It was found that during eating action, the MH
muscle became active as soon as the action starts in TD
The possibility of abnormalities of mirror mechanism in children, while no early activation was observed in children
ASD children was also investigated with fMRI. In a highly with ASD (FIGURE 22). In TD children, MH muscle was also
active during the observation of the eating action. In con-
cited study, Dapretto et al. (90) studied brain activations in
trast, in children with ASD, the observation of another child
high-functioning children with ASD and in TD children
grasping and eating chocolate did not activate MH muscle.
while they observed and imitated emotional facial expres-
In other terms, the observation of actions of others did not
sions. The results showed that ASD children, in contrast to
“intrude” into the motor system of ASD children. The au-
TD children, presented poor activation in the frontal mirror thors proposed that a basic deficit in ASD consists of im-
node (pars opercularis of IFG). Furthermore, the activation pairment of the motor system, which in turn prevents the
of IFG was inversely related to the severity of symptoms in organization of intentional motor action. This last deficit is
the social domain. A dysfunctional mirror neuron mecha- responsible for impaired capacity of children with ASD in
nism was also reported by Martineau et al. (254). In con- understanding others’ intention.
trast, no abnormalities in the parieto-frontal circuit was
found by Dinstein et al. (100) in adults with ASD during the Additional evidence in favor of a deficit of the intention
observation and execution of intransitive hand movements. understanding in ASD based on the mirror mechanism has
been provided by Boria et al. (36). They tested the capacity
Another technique that was used for studying the mirror of TD and ASD children to report the goal of the observed
mechanism in autism is TMS. Theoret et al. (378) investi- motor acts, i.e., what the actor was doing, and the intention
gated the neural mechanism matching action observation underlying it, i.e., why he/she was doing it. It was found that
and execution in adults with ASD and normal controls. children with ASD recognize the what of the motor acts, but
They applied TMS over the primary motor cortex (M1) they fail to recognize the why, i.e., the intention behind the
during observation of intransitive, meaningless finger observed action.
movements. They showed that overall modulation of M1
excitability during action observation was significantly Evidence that children with ASD have motor impairment in
lower in individuals with ASD compared with matched con- action organization was provided by Fabbri-Destro et al.
trols. A more recent TMS study tested the cortico-spinal (109). TD and ASD children were asked to perform two
excitability in adults during the observation of transitive actions consisting each of two motor acts, the first identical
hand gesture, relative to observation of the static hand. in both actions, the second varying in its difficulty. It was
They found a significantly reduced cortical excitability in found that the task difficulty modulates the kinematics of
ASD relative to controls. Interestingly, among the ASD the first motor act, in TD children (190, 253), while it did
group there was a negative association between the activity not affect it in children with ASD.
of the mirror mechanism and self-reported social impair-
The data of Boria et al. (36) show that ASD children are able
ments (107). An original approach to the study of autism
to recognize the what of the motor act was confirmed by
was used by Puzzo et al. (320). Instead of testing individuals
Hamilton et al. (160). These authors, however, considered
with ASD, they investigated the cortico-spinal excitability
this finding as evidence against a role of mirror neurons in
in individuals with high and low traits of autism assessed autism deficit. This criticism does not take into consider-
with the autistic quotient (AQ) test (29). Videos of hand ation that the what of an action can be recognized in differ-
and mouth actions and static images with hand or mouth ent ways even without the involvement of the mirror mech-
were used as stimuli. The results showed that individuals anism (see above).
with low trait of autism presented large MEPs during the
observation of the hand and mouth actions compared with
MEPs obtained during the observation of static stimuli. In E. Conclusions
contrast, participants with high traits of autism showed
MEPs with similar amplitude during the observation of In this review we summarized the present state of our
both static and dynamic stimuli. knowledge concerning the mirror mechanism. This

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 695


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

FIGURE 22. EMG activity during action


execution and action observation. A: sche-
matic representation of the task. Top: the
B individual reaches for a piece of food lo-
Typically Developing Children Children with Autism cated on a touch-sensitive plate, grasps it,
0.08 0.08 brings it to the mouth, and eats it. Bottom:
the individual reaches for a piece of a paper
rectified mylohyioid EMG

rectified mylohyioid EMG


0.07 0.07
located on the same plate, grasps it, and
eat eat
0.06 place 0.06 puts into a container placed on the shoul-
place
Execution

0.05 0.05
der. B, top: time course of the rectified
EMG activity of mylohyoideus (MH) muscle
0.04 0.04 during the execution of the bringing-to-the-
0.03 0.03 mouth action (red) and the placing action
(blue) in typically developing children (left)
0.02 0.02 and in children with autism (right). Bottom:
0.01 0.01 time course of the rectified EMG activity of
-2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 MH muscle during observation of the bring-
time (s) time (s) ing-to-the-mouth (red) and placing actions
(blue) in typically developing children (left)
rectified mylohyioid EMG

rectified mylohyioid EMG

0.06 0.06 and in children with autism (right). All


curves are aligned with the moment of ob-
Observation

0.05 0.05
ject lifting from the touch-sensitive plate.
eat eat
0.04 place 0.04
place [Modified from Cattaneo et al. (69).]
0.03 0.03

0.02 0.02

0.01 0.01
-2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0
time (s) time (s)

mechanism appears to play a fundamental role in under- mentalizing network, appears to be a first step in this
standing actions and intentions of others. This conclu- direction.
sion is corroborated by the presence of the mirror mech-
anism also in species, evolutionarily distant from pri- It will be also of great interest to continue the pioneer work
mates, like birds (197, 314). But what about mammals? by Mukhamel et al. (271) on the properties of single neu-
Is the mirror mechanism limited to primates, or is it also rons in humans. The technology used in that and other
present in other mammalian families? Future research studies on human single neurons enables single neurons
will clarify this issue. Of particular interest would be to recording only from the mesial brain structures (see Ref.
establish the presence of the mirror mechanism in ro- 272). It is likely, however, that in the near future electrodes
dents. If the mirror mechanism is present in these species, will be available for recording single neuron activity from
this will open the possibility to characterize the mirror neural structure located in other parts of the cerebral cor-
mechanism at a biochemical level and to establish its tex. This possibility will be instrumental in solving the pres-
genetic bases. Both these approaches are hardly achiev- ent dichotomy between the mirror neuron-based cognition
able in primates. and mentalizing.
Another fundamental development of mirror neuron re-
Finally, as shown in this review, the mirror mechanism is
search should be that of defining the relationships be-
the only mechanism that allows us to understand others
tween the basic mirror circuits and other circuits related
“from inside” (337), as individuals similar to ourselves. It is
to it and in particular those that could exert a top-down
control on it. Even more fascinating will be to establish very likely that malfunctioning of this mechanism should be
the anatomical and functional relationship between the responsible for symptoms of psychiatric disorders charac-
areas endowed with the mirror mechanism and the areas terized by what Baron-Cohen called “zero degrees of em-
involved, according to fMRI studies, in mentalizing. The pathy” (28). Psychiatry will be, most likely, one of the most
study of Yoshida et al. (409) that showed mirror neurons important “mirror fields” of the future. The discovery of
in the monkey medial frontal cortex, an area part of the relations between impairment in the mirror mechanism and

696 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

some autistic symptoms has been an encouraging begin- 14. Arfeller C, Schwarzbach J, Ubaldi S, Ferrari P, Barchiesi G, Cattaneo L. Whole-brain
haemodynamic after-effects of 1-Hz magnetic stimulation of the posterior superior
ning. temporal cortex during action observation. Brain Topogr 26: 278 –291, 2013.

ACKNOWLEDGMENTS 15. Arnstein D, Cui F, Keysers C, Maurits NM, Gazzola V. Mu-suppression during action
observation and execution correlates with BOLD in dorsal premotor, inferior pari-
etal, and SI cortices. J Neurosci 31: 14243–14249, 2011.
Address for reprint requests and other correspondence: G. Rizzo-
16. Asanuma H, Rosen I. Topographical organization of cortical efferent zones projecting
latti, Dept. of Neuroscience, Univ. of Parma, Via Volturno, 39-E, to distal forelimb muscles in the monkey. Exp Brain Res 14: 243–256, 1972.
Parma, Italy (e-mail: giacomo.rizzolatti@unipr.it).
17. Asperger H. Die ‘Autistischen Psychopathen’ im Kindesalter. Arch fur Psychiatrie Ner-
venkrankheiten 117: 76 –136, 1944.
GRANTS
18. Avanzini P, Fabbri-Destro M, Dalla Volta R, Daprati E, Rizzolatti G, Cantalupo G. The
dynamics of sensorimotor cortical oscillations during the observation of hand move-
This study was supported by the European Grant “Cog- ments: an EEG study. PLoS ONE 7: e37534, 2012.
systems” and by Rete Multidisciplinare Tecnolgica
(RTM-IIT). 19. Avenanti A, Bolognini N, Maravita A, Aglioti SM. Somatic and motor components of
action simulation. Curr Biol 17: 2129 –2135, 2007.

20. Aziz-Zadeh L, Iacoboni M, Zaidel E, Wilson S, Mazziotta J. Left hemisphere motor


DISCLOSURES facilitation in response to manual action sounds. Eur J Neurosci 19: 2609 –2612, 2004.

21. Aziz-Zadeh L, Maeda F, Zaidel E, Mazziotta J, Iacoboni M. Lateralization in motor


No conflicts of interest, financial or otherwise, are declared facilitation during action observation: a TMS study. Exp Brain Res 144: 127–131, 2002.
by the authors.
22. Bakola S, Gamberini M, Passarelli L, Fattori P, Galletti C. Cortical connections of
parietal field PEc in the macaque: linking vision and somatic sensation for the control
of limb action. Cereb Cortex 20: 2592–2604, 2010.
REFERENCES
23. Barbas H. Anatomic organization of basoventral and mediodorsal visual recipient
prefrontal regions in the rhesus monkey. J Comp Neurol 276: 313–342, 1988.
1. Abdollahi RO, Jastorff J, Orban GA. Common and segregated processing of observed
actions in human SPL. Cereb Cortex 11: 2734 –2753, 2012. 24. Barbas H, Pandya DN. Architecture and frontal cortical connections of the premotor
cortex (area 6) in the rhesus monkey. J Comp Neurol 256: 211–228, 1987.
2. Aglioti SM, Cesari P, Romani M, Urgesi C. Action anticipation and motor resonance in
elite basketball players. Nat Neurosci 11: 1109 –1116, 2008. 26. Barchiesi G, Cattaneo L. Early and late motor responses to action observation. Soc
Cogn Affect Neurosci 8: 711–719, 2013.
3. Alaerts K, de Beukelaar TT, Swinnen SP, Wenderoth N. Observing how others lift
light or heavy objects: time-dependent encoding of grip force in the primary motor 27. Barchiesi G, Wache S, Cattaneo L. The frames of reference of the motor-visual
cortex. Psychol Res 76: 503–513, 2012. aftereffect. PLoS One 7: e40892, 2012.

4. Alaerts K, Heremans E, Swinnen SP, Wenderoth N. How are observed actions 28. Baron-Cohen S. Zero Degrees of Empathy. London: Penguin Books Limited, 2011.
mapped to the observer’s motor system? Influence of posture and perspective. Neu-
ropsychologia 47: 415– 422, 2009. 29. Baron-Cohen S, Wheelwright S, Skinner R, Martin J, Clubley E. The autism-
spectrum quotient (AQ): evidence from Asperger syndrome/high-functioning au-
5. Alaerts K, Swinnen SP, Wenderoth N. Is the human primary motor cortex activated tism, males and females, scientists and mathematicians. J Autism Dev Disord 31:
by muscular or direction-dependent features of observed movements? Cortex 45: 5–17, 2001.
1148 –1155, 2009.
30. Baumann MA, Fluet MC, Scherberger H. Context-specific grasp movement rep-
6. Alegre M, Rodriguez-Oroz MC, Valencia M, Perez-Alcazar M, Guridi J, Iriarte J,
resentation in the macaque anterior intraparietal area. J Neurosci 29: 6436 – 6448,
Obeso JA, Artieda J. Changes in subthalamic activity during movement observation in
2009.
Parkinson’s disease: is the mirror system mirrored in the basal ganglia? Clin Neuro-
physiol 121: 414 – 425, 2010. 31. Belmalih A, Borra E, Contini M, Gerbella M, Rozzi S, Luppino G. A multiarchitectonic
approach for the definition of functionally distinct areas and domains in the monkey
7. Alexander GE, Crutcher MD. Preparation for movement: neural representations of
frontal lobe. J Anat 211: 199 –211, 2007.
intended direction in three motor areas of the monkey. J Neurophysiol 64: 133–150,
1990. 32. Belmalih A, Borra E, Contini M, Gerbella M, Rozzi S, Luppino G. Multimodal archi-
tectonic subdivision of the rostral part (area F5) of the macaque ventral premotor
8. Allison T, Puce A, McCarthy G. Social perception from visual cues: role of the STS
region. Trends Cogn Sci 4: 267–278, 2000. cortex. J Comp Neurol 512: 183–217, 2009.

9. Altschuler EL, Vankov A, Wang V, Ramachandran VS, Pineda JA. Person see, person 33. Bernier R, Dawson G, Webb S, Murias M. EEG mu rhythm and imitation impairments
do: human cortical electrophysiological correlates of monkey see monkey do cells. In: in individuals with autism spectrum disorder. Brain Cogn 64: 228 –237, 2007.
Annual Meeting of the Society for Neuroscience. New Orleans: Soc. Neurosci., 1997, p.
34. Bonini L, Rozzi S, Serventi FU, Simone L, Ferrari PF, Fogassi L. Ventral premotor and
719.717.
inferior parietal cortices make distinct contribution to action organization and inten-
10. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders tion understanding. Cereb Cortex 20: 1372–1385, 2010.
(5th ed.). Washington, DC: Am. Psychiatr. Assoc., 2013.
35. Bonini L, Ugolotti Serventi F, Bruni S, Maranesi M, Bimbi M, Simone L, Rozzi S, Ferrari
11. Amunts K, Zilles K. Architecture and organizational principles of Broca’s region. PF, Fogassi L. Selectivity for grip type and action goal in macaque inferior parietal and
Trends Cogn Sci 16: 418 – 426, 2012. ventral premotor grasping neurons. J Neurophysiol 108: 1607–1619, 2012.

12. Andersen RA, Asanuma C, Essick G, Siegel RM. Corticocortical connections of ana- 36. Boria S, Fabbri-Destro M, Cattaneo L, Sparaci L, Sinigaglia C, Santelli E, Cossu G,
tomically and physiologically defined subdivisions within the inferior parietal lobule. J Rizzolatti G. Intention understanding in autism. PLoS ONE 4: e5596, 2009.
Comp Neurol 296: 65–113, 1990.
37. Borra E, Belmalih A, Calzavara R, Gerbella M, Murata A, Rozzi S, Luppino G. Cortical
13. Annett J. Motor imagery: perception or action? Neuropsychologia 33: 1395–1417, connections of the macaque anterior intraparietal (AIP) area. Cereb Cortex 18: 1094 –
1995. 1111, 2008.

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 697


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

38. Borra E, Belmalih A, Gerbella M, Rozzi S, Luppino G. Projections of the hand field of 59. Campbell AW. Histological Studies on the Localisation of Cerebral Function. Cambridge,
the macaque ventral premotor area F5 to the brainstem and spinal cord. J Comp MA: Cambridge Univ. Press, 1905.
Neurol 518: 2570 –2591, 2010.
60. Candidi M, Stienen BM, Aglioti SM, de Gelder B. Event-related repetitive transcranial
39. Borroni P, Montagna M, Cerri G, Baldissera F. Cyclic time course of motor excitability magnetic stimulation of posterior superior temporal sulcus improves the detection of
modulation during the observation of a cyclic hand movement. Brain Res 1065: 115– threatening postural changes in human bodies. J Neurosci 31: 17547–17554, 2011.
124, 2005.
61. Carruthers P, Smith PK. Theories of Theories of Mind. Cambridge, MA: Cambridge
40. Bortoletto M, Mattingley JB, Cunnington R. Action intentions modulate visual pro- Univ. Press, 2007.
cessing during action perception. Neuropsychologia 49: 2097–2104, 2011.
62. Casile A, Giese MA. Nonvisual motor training influences biological motion perception.
41. Boussaoud D, Ungerleider LG, Desimone R. Pathways for motion analysis: cortical Curr Biol 16: 69 –74, 2006.
connections of the medial superior temporal and fundus of the superior temporal
visual areas in the macaque. J Comp Neurol 296: 462– 495, 1990. 63. Caspers S, Zilles K, Laird AR, Eickhoff SB. ALE meta-analysis of action observation and
imitation in the human brain. Neuroimage 50: 1148 –1167, 2010.
42. Brambilla P, Hardan A, di Nemi SU, Perez J, Soares JC, Barale F. Brain anatomy and
development in autism: review of structural MRI studies. Brain Res Bull 61: 557–569, 64. Catmur C, Mars RB, Rushworth MF, Heyes C. Making mirrors: premotor cortex
2003. stimulation enhances mirror and counter-mirror motor facilitation. J Cogn Neurosci
23: 2352–2362, 2011.
43. Breveglieri R, Galletti C, Gamberini M, Passarelli L, Fattori P. Somatosensory cells in
area PEc of macaque posterior parietal cortex. J Neurosci 26: 3679 –3684, 2006. 65. Catmur C, Walsh V, Heyes C. Sensorimotor learning configures the human mirror
system. Curr Biol 17: 1527–1531, 2007.
44. Brodmann K. Vergleichende Lokalisationslehre der Grosshirnrinde in ihren Prinzipien
dargestellt auf Grund des Zellenbaues. Leipzig: Johann Ambrosius Barth Verlag, 1909. 66. Cattaneo L, Barchiesi G. Transcranial magnetic mapping of the short-latency modu-
lations of corticospinal activity from the ipsilateral hemisphere during rest. Front
45. Buccino G, Baumgaertner A, Colle L, Buechel C, Rizzolatti G, Binkofski F. The neural Neural Circuits 5: 14, 2011.
basis for understanding non-intended actions. Neuroimage 36 Suppl 2: T119 –127,
2007. 67. Cattaneo L, Barchiesi G, Tabarelli D, Arfeller C, Sato M, Glenberg AM. One’s motor
performance predictably modulates the understanding of others’ actions through
46. Buccino G, Binkofski F, Fink GR, Fadiga L, Fogassi L, Gallese V, Seitz RJ, Zilles K, adaptation of premotor visuo-motor neurons. Soc Cogn Affect Neurosci 6: 301–310,
Rizzolatti G, Freund HJ. Action observation activates premotor and parietal areas in a 2011.
somatotopic manner: an fMRI study. Eur J Neurosci 13: 400 – 404, 2001.
68. Cattaneo L, Caruana F, Jezzini A, Rizzolatti G. Representation of goal and movements
47. Buccino G, Lui F, Canessa N, Patteri I, Lagravinese G, Benuzzi F, Porro CA, Rizzolatti without overt motor behavior in the human motor cortex: a transcranial magnetic
G. Neural circuits involved in the recognition of actions performed by nonconspecif- stimulation study. J Neurosci 29: 11134 –11138, 2009.
ics: an FMRI study. J Cogn Neurosci 16: 114 –126, 2004.
69. Cattaneo L, Fabbri-Destro M, Boria S, Pieraccini C, Monti A, Cossu G, Rizzolatti G.
48. Buccino G, Vogt S, Ritzl A, Fink GR, Zilles K, Freund HJ, Rizzolatti G. Neural circuits Impairment of actions chains in autism and its possible role in intention understanding.
underlying imitation learning of hand actions: an event-related fMRI study. Neuron 42: Proc Natl Acad Sci USA 104: 17825–17830, 2007.
323–334, 2004.
70. Cattaneo L, Maule F, Barchiesi G, Rizzolatti G. The motor system resonates to the
49. Buxbaum LJ, Johnson-Frey SH, Bartlett-Williams M. Deficient internal models for distal goal of observed actions: testing the inverse pliers paradigm in an ecological
planning hand-object interactions in apraxia. Neuropsychologia 43: 917–929, 2005. setting. Exp Brain Res 231: 37– 49, 2013.

50. Buxbaum LJ, Kyle KM, Menon R. On beyond mirror neurons: internal representations 71. Cattaneo L, Sandrini M, Schwarzbach J. State-dependent TMS reveals a hierarchical
subserving imitation and recognition of skilled object-related actions in humans. Brain representation of observed acts in the temporal, parietal, and premotor cortices.
Res Cogn Brain Res 25: 226 –239, 2005. Cereb Cortex 20: 2252–2258, 2010.

51. Byrne RW. Imitation as behaviour parsing. Philos Trans R Soc Lond B Biol Sci 358: 72. Cattaneo L, Voss M, Brochier T, Prabhu G, Wolpert DM, Lemon RN. A cortico-
529 –536, 2003. cortical mechanism mediating object-driven grasp in humans. Proc Natl Acad Sci USA
102: 898 –903, 2005.
52. Caggiano V, Fogassi L, Rizzolatti G, Casile A, Giese MA, Thier P. Mirror neurons
encode the subjective value of an observed action. Proc Natl Acad Sci USA 109: 11848 – 73. Cavada C, Goldman-Rakic PS. Posterior parietal cortex in rhesus monkey. I. Parcel-
11853, 2012. lation of areas based on distinctive limbic and sensory corticocortical connections. J
Comp Neurol 287: 393– 421, 1989.
53. Caggiano V, Fogassi L, Rizzolatti G, Pomper JK, Thier P, Giese MA, Casile A. View-
based encoding of actions in mirror neurons of area f5 in macaque premotor cortex. 74. Cavada C, Goldman-Rakic PS. Posterior parietal cortex in rhesus monkey. II. Evidence
Curr Biol 21: 144 –148, 2011. for segregated corticocortical networks linking sensory and limbic areas with the
frontal lobe. J Comp Neurol 287: 422– 445, 1989.
54. Caggiano V, Fogassi L, Rizzolatti G, Thier P, Casile A. Mirror neurons differentially
encode the peripersonal and extrapersonal space of monkeys. Science 324: 403– 406, 75. Cerri G, Shimazu H, Maier MA, Lemon RN. Facilitation from ventral premotor cortex
2009. of primary motor cortex outputs to macaque hand muscles. J Neurophysiol 90: 832–
842, 2003.
55. Caggiano V, Pomper JK, Fleischer F, Fogassi L, Giese M, Thier P. Mirror neurons in
monkey area F5 do not adapt to the observation of repeated actions. Nat Commun 4: 76. Cisek P, Kalaska JF. Neural correlates of mental rehearsal in dorsal premotor cortex.
1433, 2013. Nature 431: 993–996, 2004.

56. Calvo-Merino B, Glaser DE, Grezes J, Passingham RE, Haggard P. Action observation 77. Cisek P, Kalaska JF. Neural mechanisms for interacting with a world full of action
and acquired motor skills: an FMRI study with expert dancers. Cereb Cortex 15: choices. Annu Rev Neurosci 33: 269 –298, 2010.
1243–1249, 2005.
78. Clerget E, Winderickx A, Fadiga L, Olivier E. Role of Broca’s area in encoding sequen-
57. Calvo-Merino B, Grezes J, Glaser DE, Passingham RE, Haggard P. Seeing or doing? tial human actions: a virtual lesion study. Neuroreport 20: 1496 –1499, 2009.
Influence of visual and motor familiarity in action observation. Curr Biol 16: 1905–
1910, 2006. 79. Colby CL, Duhamel JR. Spatial representations for action in parietal cortex. Brain Res
Cogn Brain Res 5: 105–115, 1996.
58. Caminiti R, Ferraina S, Johnson PB. The sources of visual information to the primate
frontal lobe: a novel role for the superior parietal lobule. Cereb Cortex 6: 319 –328, 80. Colby CL, Duhamel JR, Goldberg ME. Ventral intraparietal area of the macaque:
1996. anatomic location and visual response properties. J Neurophysiol 69: 902–914, 1993.

698 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

81. Colby CL, Gattass R, Olson CR, Gross CG. Topographical organization of cortical 104. Dushanova J, Donoghue J. Neurons in primary motor cortex engaged during action
afferents to extrastriate visual area PO in the macaque: a dual tracer study. J Comp observation. Eur J Neurosci 31: 386 –398, 2010.
Neurol 269: 392– 413, 1988.
105. Enticott PG, Kennedy HA, Bradshaw JL, Rinehart NJ, Fitzgerald PB. Motor cortico-
82. Cooke DF, Taylor CS, Moore T, Graziano MS. Complex movements evoked by spinal excitability during the observation of interactive hand gestures. Brain Res Bull
microstimulation of the ventral intraparietal area. Proc Natl Acad Sci USA 100: 6163– 85: 89 –95, 2011.
6168, 2003.
106. Enticott PG, Kennedy HA, Bradshaw JL, Rinehart NJ, Fitzgerald PB. Understanding
83. Courchesne E. Brainstem, cerebellar and limbic neuroanatomical abnormalities in mirror neurons: evidence for enhanced corticospinal excitability during the observa-
autism. Curr Opin Neurobiol 7: 269 –278, 1997. tion of transitive but not intransitive hand gestures. Neuropsychologia 48: 2675–2680,
2010.
84. Crammond DJ, Kalaska JF. Prior information in motor and premotor cortex: activity
during the delay period and effect on pre-movement activity. J Neurophysiol 84: 986 – 107. Enticott PG, Kennedy HA, Rinehart NJ, Tonge BJ, Bradshaw JL, Taffe JR, Daskalakis
1005, 2000. ZJ, Fitzgerald PB. Mirror neuron activity associated with social impairments but not
age in autism spectrum disorder. Biol Psychiatry 71: 427– 433, 2012.
85. Crippa A, Cerliani L, Nanetti L, Roerdink JB. Heuristics for connectivity-based brain
parcellation of SMA/pre-SMA through force-directed graph layout. Neuroimage 54: 108. Evarts EV. Relation of pyramidal tract activity to force exerted during voluntary move-
2176 –2184, 2011. ment. J Neurophysiol 31: 14 –27, 1968.

86. Critchley M. Tactile thought, with special reference to the blind. Brain 76: 19 –35, 109. Fabbri-Destro M, Cattaneo L, Boria S, Rizzolatti G. Planning actions in autism. Exp
1953. Brain Res 192: 521–525, 2009.

87. Cross ES, Hamilton AF, Grafton ST. Building a motor simulation de novo: observation 110. Fadiga L, Craighero L, Buccino G, Rizzolatti G. Speech listening specifically modulates
of dance by dancers. Neuroimage 31: 1257–1267, 2006. the excitability of tongue muscles: a TMS study. Eur J Neurosci 15: 399 – 402, 2002.

88. D’Ausilio A, Altenmuller E, Olivetti Belardinelli M, Lotze M. Cross-modal plasticity of 111. Fadiga L, Fogassi L, Pavesi G, Rizzolatti G. Motor facilitation during action observation:
the motor cortex while listening to a rehearsed musical piece. Eur J Neurosci 24: a magnetic stimulation study. J Neurophysiol 73: 2608 –2611, 1995.
955–958, 2006.
112. Fagg AH, Arbib MA. Modeling parietal-premotor interactions in primate control of
89. D’Ausilio A, Bufalari I, Salmas P, Fadiga L. The role of the motor system in discrimi- grasping. Neural Netw 11: 1277–1303, 1998.
nating normal and degraded speech sounds. Cortex 48: 882– 887, 2012.
113. Fecteau S, Carmant L, Tremblay C, Robert M, Bouthillier A, Theoret H. A motor
90. Dapretto M, Davies MS, Pfeifer JH, Scott AA, Sigman M, Bookheimer SY, Iacoboni M. resonance mechanism in children? Evidence from subdural electrodes in a 36-month-
Understanding emotions in others: mirror neuron dysfunction in children with autism old child. Neuroreport 15: 2625–2627, 2004.
spectrum disorders. Nat Neurosci 9: 28 –30, 2006.
114. Fecteau S, Lassonde M, Theoret H. Modulation of motor cortex excitability during
91. De Lucia M, Camen C, Clarke S, Murray MM. The role of actions in auditory object action observation in disconnected hemispheres. Neuroreport 16: 1591–1594, 2005.
discrimination. Neuroimage 48: 475– 485, 2009.
115. Ferbert A, Priori A, Rothwell JC, Day BL, Colebatch JG, Marsden CD. Interhemi-
92. De Renzi E, Cavalleri F, Facchini S. Imitation and utilisation behaviour. J Neurol Neu- spheric inhibition of the human motor cortex. J Physiol 453: 525–546, 1992.
rosurg Psychiatry 61: 396 – 400, 1996.
116. Ferrari PF, Gallese V, Rizzolatti G, Fogassi L. Mirror neurons responding to the
93. De Renzi E, Faglioni P. Apraxia. In: Handbook of Clinical and Experimental Neuropsy- observation of ingestive and communicative mouth actions in the monkey ventral
chology, edited by Denes G and Pizzamiglio L. Hove, UK: Psychology Press, 1999, p. premotor cortex. Eur J Neurosci 17: 1703–1714, 2003.
421– 440.
117. Feurra M, Bianco G, Polizzotto NR, Innocenti I, Rossi A, Rossi S. Cortico-Cortical
94. De Renzi E, Faglioni P, Sorgato P. Modality-specific and supramodal mechanisms of Connectivity between Right Parietal and Bilateral Primary Motor Cortices during Imagined
apraxia. Brain 105: 301–312, 1982. and Observed Actions: A Combined TMS/tDCS study. Front Neural Circuits 5: 10, 2011.

95. Deecke L, Kornhuber HH. An electrical sign of participation of the mesial “supple- 118. Filimon F, Nelson JD, Hagler DJ, Sereno MI. Human cortical representations for
mentary” motor cortex in human voluntary finger movement. Brain Res 159: 473– reaching: mirror neurons for execution, observation, and imagery. Neuroimage 37:
476, 1978. 1315–1328, 2007.

96. Della Sala S, Marchetti C, Spinnler H. Right-sided anarchic (alien) hand: a longitudinal 119. Fiorio M, Cesari P, Bresciani MC, Tinazzi M. Expertise with pathological actions
study. Neuropsychologia 29: 1113–1127, 1991. modulates a viewer’s motor system. Neuroscience 167: 691– 699, 2010.

97. Di Dio C, Di Cesare G, Higuchi S, Roberts N, Vogt S, Rizzolatti G. The neural 120. Fluet MC, Baumann MA, Scherberger H. Context-specific grasp movement repre-
correlates of velocity processing during the observation of a biological effector in the sentation in macaque ventral premotor cortex. J Neurosci 30: 15175–15184, 2010.
parietal and premotor cortex. Neuroimage 64: 425– 436, 2013.
121. Fogassi L, Ferrari PF, Gesierich B, Rozzi S, Chersi F, Rizzolatti G. Parietal lobe: from
98. Di Pellegrino G, Fadiga L, Fogassi L, Gallese V, Rizzolatti G. Understanding motor action organization to intention understanding. Science 308: 662– 667, 2005.
events: a neurophysiological study. Exp Brain Res 91: 176 –180, 1992.
122. Fogassi L, Gallese V, Buccino G, Craighero L, Fadiga L, Rizzolatti G. Cortical mecha-
99. Dinstein I, Hasson U, Rubin N, Heeger DJ. Brain areas selective for both observed and nism for the visual guidance of hand grasping movements in the monkey: a reversible
executed movements. J Neurophysiol 98: 1415–1427, 2007. inactivation study. Brain 124: 571–586, 2001.

100. Dinstein I, Thomas C, Humphreys K, Minshew N, Behrmann M, Heeger DJ. Normal 123. Fogassi L, Gallese V, Fadiga L, Luppino G, Matelli M, Rizzolatti G. Coding of periper-
movement selectivity in autism. Neuron 66: 461– 469, 2010. sonal space in inferior premotor cortex (area F4). J Neurophysiol 76: 141–157, 1996.

101. Donne CM, Enticott PG, Rinehart NJ, Fitzgerald PB. A transcranial magnetic 124. Frith U. Autism: Explaining the Enigma. Oxford, UK: Blackwell, 2003.
stimulation study of corticospinal excitability during the observation of meaning-
less, goal-directed, and social behaviour. Neurosci Lett 489: 57– 61, 2011. 125. Fujii N, Mushiake H, Tanji J. Rostrocaudal distinction of the dorsal premotor area
based on oculomotor involvement. J Neurophysiol 83: 1764 –1769, 2000.
102. Duhamel JR, Colby CL, Goldberg ME. Ventral intraparietal area of the macaque:
congruent visual and somatic response properties. J Neurophysiol 79: 126 –136, 126. Fulton J. A note on the definition of the “motor” and “premotor” areas. Brain 58:
1998. 311–316, 1935.

103. Dum RP, Strick PL. The origin of corticospinal projections from the premotor areas in 127. Gallese V, Fadiga L, Fogassi L, Rizzolatti G. Action recognition in the premotor cortex.
the frontal lobe. J Neurosci 11: 667– 689, 1991. Brain 119: 593– 609, 1996.

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 699


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

128. Gallese V, Keysers C, Rizzolatti G. A unifying view of the basis of social cognition. 151. Goodale MA, Foreman NP, Milner AD. Visual orientation in the rat: a dissociation of
Trends Cogn Sci 8: 396 – 403, 2004. deficits following cortical and collicular lesions. Exp Brain Res 31: 445– 457, 1978.

129. Gallese V, Murata A, Kaseda M, Niki N, Sakata H. Deficit of hand preshaping after 152. Grafton ST, Arbib MA, Fadiga L, Rizzolatti G. Localization of grasp representations in
muscimol injection in monkey parietal cortex. Neuroreport 5: 1525–1529, 1994. humans by positron emission tomography. 2. Observation compared with imagina-
tion. Exp Brain Res 112: 103–111, 1996.
130. Galletti C, Fattori P, Kutz DF, Gamberini M. Brain location and visual topography of
cortical area V6A in the macaque monkey. Eur J Neurosci 11: 575–582, 1999. 153. Graziano M. The organization of behavioral repertoire in motor cortex. Annu Rev
Neurosci 29: 105–134, 2006.
131. Galletti C, Gamberini M, Kutz DF, Fattori P, Luppino G, Matelli M. The cortical
connections of area V6: an occipito-parietal network processing visual information. 154. Graziano MS, Reiss LA, Gross CG. A neuronal representation of the location of nearby
Eur J Neurosci 13: 1572–1588, 2001. sounds. Nature 397: 428 – 430, 1999.
132. Gamberini M, Passarelli L, Fattori P, Zucchelli M, Bakola S, Luppino G, Galletti C. 155. Grezes J, Armony JL, Rowe J, Passingham RE. Activations related to “mirror” and
Cortical connections of the visuomotor parietooccipital area V6Ad of the macaque “canonical” neurones in the human brain: an fMRI study. Neuroimage 18: 928 –937,
monkey. J Comp Neurol 513: 622– 642, 2009. 2003.
133. Gangitano M, Mottaghy FM, Pascual-Leone A. Modulation of premotor mirror neuron 156. Grezes J, Costes N, Decety J. The effects of learning and intention on the neural
activity during observation of unpredictable grasping movements. Eur J Neurosci 20:
network involved in the perception of meaningless actions. Brain 122: 1875–1887,
2193–2202, 2004.
1999.
134. Gangitano M, Mottaghy FM, Pascual-Leone A. Phase-specific modulation of cortical
157. Grosbras MH, Beaton S, Eickhoff SB. Brain regions involved in human movement
motor output during movement observation. Neuroreport 12: 1489 –1492, 2001.
perception: a quantitative voxel-based meta-analysis. Hum Brain Mapp 33: 431– 454,
135. Gangitano M, Mottaghy FM, Pascual-Leone A. Release of premotor activity after 2012.
repetitive transcranial magnetic stimulation of prefrontal cortex. Soc Neurosci 3: 289 –
158. Grossman ED, Battelli L, Pascual-Leone A. Repetitive TMS over posterior STS dis-
302, 2008.
rupts perception of biological motion. Vision Res 45: 2847–2853, 2005.
136. Gastaut H, Terzian H, Gastaut Y. Etude d’une activite electroencephalographique
meconnue: le rythme rolandique en arceau. Mars Med 89: 296 –310, 1952. 159. Hadjikhani N, Joseph RM, Snyder J, Tager-Flusberg H. Anatomical differences in the
mirror neuron system and social cognition network in autism. Cereb Cortex 16: 1276 –
137. Gastaut HJ, Bert J. EEG changes during cinematographic presentation: moving picture 1282, 2006.
activation of the EEG. Electroencephalogr Clin Neurophysiol 6: 433– 444, 1954.
160. Hamilton AF. Goals, intentions and mental states: challenges for theories of autism. J
138. Gazzola V, Aziz-Zadeh L, Keysers C. Empathy and the somatotopic auditory mirror Child Psychol Psychiatry 50: 881– 892, 2009.
system in humans. Curr Biol 16: 1824 –1829, 2006.
161. Haslinger B, Erhard P, Altenmuller E, Schroeder U, Boecker H, Ceballos-Baumann
139. Gazzola V, Keysers C. The observation and execution of actions share motor and AO. Transmodal sensorimotor networks during action observation in professional
somatosensory voxels in all tested subjects: single-subject analyses of unsmoothed pianists. J Cogn Neurosci 17: 282–293, 2005.
fMRI data. Cereb Cortex 19: 1239 –1255, 2009.
162. He SQ, Dum RP, Strick PL. Topographic organization of corticospinal projections
140. Gazzola V, Rizzolatti G, Wicker B, Keysers C. The anthropomorphic brain: the mirror from the frontal lobe: motor areas on the lateral surface of the hemisphere. J Neurosci
neuron system responds to human and robotic actions. Neuroimage 35: 1674 –1684, 13: 952–980, 1993.
2007.
163. He SQ, Dum RP, Strick PL. Topographic organization of corticospinal projections
141. Gazzola V, van der Worp H, Mulder T, Wicker B, Rizzolatti G, Keysers C. Aplasics from the frontal lobe: motor areas on the medial surface of the hemisphere. J Neurosci
born without hands mirror the goal of hand actions with their feet. Curr Biol 17: 15: 3284 –3306, 1995.
1235–1240, 2007.
164. Heilman KM, Rothi LJ, Valenstein E. Two forms of ideomotor apraxia. Neurology 32:
142. Gentilucci M, Fogassi L, Luppino G, Matelli M, Camarda R, Rizzolatti G. Functional 342–346, 1982.
organization of inferior area 6 in the macaque monkey. I. Somatotopy and the control
of proximal movements. Exp Brain Res 71: 475– 490, 1988. 165. Heiser M, Iacoboni M, Maeda F, Marcus J, Mazziotta JC. The essential role of Broca’s
area in imitation. Eur J Neurosci 17: 1123–1128, 2003.
143. Georgopoulos AP, Ashe J, Smyrnis N, Taira M. The motor cortex and the coding of
force. Science 256: 1692–1695, 1992. 166. Henneman E. Organization of the spinal cord and its reflexes. In: Medical Physiology,
edited by Mountcastle V. St. Louis, MO: Mosby, 1980.
144. Gerbella M, Belmalih A, Borra E, Rozzi S, Luppino G. Cortical connections of the
anterior (F5a) subdivision of the macaque ventral premotor area F5. Brain Struct Funct 167. Hepp-Reymond MC, Husler EJ, Maier MA, Ql HX. Force-related neuronal activity in
216: 43– 65, 2011. two regions of the primate ventral premotor cortex. Can J Physiol Pharmacol 72:
571–579, 1994.
145. Gerbella M, Belmalih A, Borra E, Rozzi S, Luppino G. Cortical connections of the
macaque caudal ventrolateral prefrontal areas 45A and 45B. Cereb Cortex 20: 141– 168. Hikosaka O, Nakahara H, Rand MK, Sakai K, Lu X, Nakamura K, Miyachi S, Doya K.
168, 2010.
Parallel neural networks for learning sequential procedures. Trends Neurosci 22: 464 –
146. Gerbella M, Borra E, Tonelli S, Rozzi S, Luppino G. Connectional heterogeneity of the 471, 1999.
ventral part of the macaque area 46. Cereb Cortex 23: 967–987, 2013.
169. Hoffmann MW, Bill PL. The environmental dependency syndrome, imitation behav-
147. Geyer S, Matelli M, Luppino G, Schleicher A, Jansen Y, Palomero-Gallagher N, Zilles iour and utilisation behaviour as presenting symptoms of bilateral frontal lobe infarc-
K. Receptor autoradiographic mapping of the mesial motor and premotor cortex of tion due to moyamoya disease. S Afr Med J 81: 271–273, 1992.
the macaque monkey. J Comp Neurol 397: 231–250, 1998.
170. Honaga E, Ishii R, Kurimoto R, Canuet L, Ikezawa K, Takahashi H, Nakahachi T, Iwase
148. Ghika J, Tennis M, Growdon J, Hoffman E, Johnson K. Environment-driven responses M, Mizuta I, Yoshimine T, Takeda M. Post-movement beta rebound abnormality as
in progressive supranuclear palsy. J Neurol Sci 130: 104 –111, 1995. indicator of mirror neuron system dysfunction in autistic spectrum disorder: an MEG
study. Neurosci Lett 478: 141–145, 2010.
149. Gibson JJ. The Ecological Approach to Visual Perception. Boston, MA: Houghton Mifflin,
1979. 171. Hoshi E, Tanji J. Integration of target and body-part information in the premotor
cortex when planning action. Nature 408: 466 – 470, 2000.
150. Godschalk M, Lemon RN, Nijs HG, Kuypers HG. Behaviour of neurons in monkey
peri-arcuate and precentral cortex before and during visually guided arm and hand 172. Hurley S, Chater N. Perspectives on Imitation: From Neuroscience to Social Science
movements. Exp Brain Res 44: 113–116, 1981. Cambridge, MA: MIT Press, 2005.

700 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

173. Hyvarinen J. Posterior parietal lobe of the primate brain. Physiol Rev 62: 1060 –1129, 195. Kanner L. Autistic disturbances of affective contact. Nervous Child 2: 217–250, 1943.
1982.
196. Keizer K, Kuypers HG. Distribution of corticospinal neurons with collaterals to the
174. Iacoboni M, Koski LM, Brass M, Bekkering H, Woods RP, Dubeau MC, Mazziotta JC, lower brain stem reticular formation in monkey (Macaca fascicularis). Exp Brain Res 74:
Rizzolatti G. Reafferent copies of imitated actions in the right superior temporal 311–318, 1989.
cortex. Proc Natl Acad Sci USA 98: 13995–13999, 2001.
197. Keller GB, Hahnloser RH. Neural processing of auditory feedback during vocal prac-
175. Iacoboni M, Molnar-Szakacs I, Gallese V, Buccino G, Mazziotta JC, Rizzolatti G. Grasp- tice in a songbird. Nature 457: 187–190, 2009.
ing the intentions of others with one’s own mirror neuron system. PLoS Biol 3: e79,
2005. 198. Keuken MC, Hardie A, Dorn BT, Dev S, Paulus MP, Jonas KJ, Den Wildenberg WP,
Pineda JA. The role of the left inferior frontal gyrus in social perception: an rTMS
176. Iacoboni M, Woods RP, Brass M, Bekkering H, Mazziotta JC, Rizzolatti G. Cortical study. Brain Res 1383: 196 –205, 2011.
mechanisms of human imitation. Science 286: 2526 –2528, 1999.
199. Kilner JM. More than one pathway to action understanding. Trends Cogn Sci 15:
177. Ingle D. Two visual systems in the frog. Science 181: 1053–1055, 1973. 352–357, 2011.

178. Isa T, Ohki Y, Alstermark B, Pettersson LG, Sasaki S. Direct and indirect cortico- 200. Kilner JM, Friston KJ, Frith CD. Predictive coding: an account of the mirror neuron
motoneuronal pathways and control of hand/arm movements. Physiology 22: 145– system. Cogn Process 8: 159 –166, 2007.
152, 2007.
201. Kilner JM, Marchant JL, Frith CD. Modulation of the mirror system by social relevance.
179. Ishida H, Nakajima K, Inase M, Murata A. Shared mapping of own and others’ bodies Soc Cogn Affect Neurosci 1: 143–148, 2006.
in visuotactile bimodal area of monkey parietal cortex. J Cogn Neurosci 22: 83–96,
2010. 202. Kilner JM, Neal A, Weiskopf N, Friston KJ, Frith CD. Evidence of mirror neurons in
human inferior frontal gyrus. J Neurosci 29: 10153–10159, 2009.
180. Iwamura Y, Tanaka M. Representation of reaching and grasping in the monkey post-
central gyrus. Neurosci Lett 214: 147–150, 1996. 203. Kilner JM, Vargas C, Duval S, Blakemore SJ, Sirigu A. Motor activation prior to obser-
vation of a predicted movement. Nat Neurosci 7: 1299 –1301, 2004.
181. Jahanshahi M, Jenkins IH, Brown RG, Marsden CD, Passingham RE, Brooks DJ. Self-
initiated versus externally triggered movements. I. An investigation using measure- 204. Kinoshita M, Matsui R, Kato S, Hasegawa T, Kasahara H, Isa K, Watakabe A, Yamamori
ment of regional cerebral blood flow with PET and movement-related potentials in T, Nishimura Y, Alstermark B, Watanabe D, Kobayashi K, Isa T. Genetic dissection of
normal and Parkinson’s disease subjects. Brain 118: 913–933, 1995. the circuit for hand dexterity in primates. Nature 487: 235–238, 2012.
182. Jastorff J, Begliomini C, Fabbri-Destro M, Rizzolatti G, Orban GA. Coding observed 205. Koch G, Versace V, Bonni S, Lupo F, Gerfo EL, Oliveri M, Caltagirone C. Resonance
motor acts: different organizational principles in the parietal and premotor cortex of of cortico-cortical connections of the motor system with the observation of goal
humans. J Neurophysiol 104: 128 –140, 2010. directed grasping movements. Neuropsychologia 48: 3513–3520, 2010.
183. Jax SA, Buxbaum LJ, Moll AD. Deficits in movement planning and intrinsic coordinate 206. Kohler E, Keysers C, Umilta MA, Fogassi L, Gallese V, Rizzolatti G. Hearing sounds,
control in ideomotor apraxia. J Cogn Neurosci 18: 2063–2076, 2006. understanding actions: action representation in mirror neurons. Science 297: 846 –
848, 2002.
184. Jeannerod M. Motor Cognition: What Actions Tell to the Self. New York: Oxford Univ.
Press, 2006.
207. Kornhuber HH. Attention, readiness for action, and the stages of voluntary decision:
185. Jeannerod M. Neural simulation of action: a unifying mechanism for motor cognition. some electrophysiological correlates in man. In: Sensory-Motor Integration in the Ner-
Neuroimage 14: S103–109, 2001. vous System, edited by Creutzfeldt O, Schmidt R, and Willis WD. Berlin: Springer,
1984 p. 420 – 429.
186. Jeannerod M, Arbib MA, Rizzolatti G, Sakata H. Grasping objects: the cortical mech-
anisms of visuomotor transformation. Trends Neurosci 18: 314 –320, 1995. 208. Kornhuber HH, Deecke L. [Changes in the Brain Potential in Voluntary Movements
and Passive Movements in Man: Readiness Potential and Reafferent Potentials].
187. Jeannerod M, Paulignan Y, Weiss P. Grasping an object: one movement, several Pflügers Arch Gesamte Physiol Menschen Tiere 284: 1–17, 1965.
components. Novartis Found Symp 218: 5–20, 1998.
209. Koski L, Iacoboni M, Dubeau MC, Woods RP, Mazziotta JC. Modulation of cortical
188. Jellema T, Maassen G, Perrett DI. Single cell integration of animate form, motion and activity during different imitative behaviors. J Neurophysiol 89: 460 – 471, 2003.
location in the superior temporal cortex of the macaque monkey. Cereb Cortex 14:
781–790, 2004. 210. Koski L, Wohlschlager A, Bekkering H, Woods RP, Dubeau MC, Mazziotta JC, Iaco-
boni M. Modulation of motor and premotor activity during imitation of target-di-
189. Johansen-Berg H, Behrens TE, Robson MD, Drobnjak I, Rushworth MF, Brady JM, rected actions. Cereb Cortex 12: 847– 855, 2002.
Smith SM, Higham DJ, Matthews PM. Changes in connectivity profiles define func-
tionally distinct regions in human medial frontal cortex. Proc Natl Acad Sci USA 101: 211. Kraskov A, Dancause N, Quallo MM, Shepherd S, Lemon RN. Corticospinal neurons
13335–13340, 2004. in macaque ventral premotor cortex with mirror properties: a potential mechanism
for action suppression? Neuron 64: 922–930, 2009.
190. Johnson-Frey SH, McCarty M, Keen R. Reaching beyond spatial perception: ef-
fects of intended future actions on visually-guided prehension. Visual Cogn 11: 212. Kujirai T, Caramia MD, Rothwell JC, Day BL, Thompson PD, Ferbert A, Wroe S,
371–399, 2004. Asselman P, Marsden CD. Corticocortical inhibition in human motor cortex. J Physiol
471: 501–519, 1993.
191. Jola C, Abedian-Amiri A, Kuppuswamy A, Pollick FE, Grosbras MH. Motor simulation
without motor expertise: enhanced corticospinal excitability in visually experienced 213. Kumral E. Compulsive grasping hand syndrome: a variant of anarchic hand. Neurology
dance spectators. PLoS One 7: e33343, 2012. 57: 2143–2144, 2001.

192. Kalaska JF, Cohen DA, Hyde ML, Prud’homme M. A comparison of movement direc- 214. Kurata K. Distribution of neurons with set- and movement-related activity before
tion-related versus load direction-related activity in primate motor cortex, using a hand and foot movements in the premotor cortex of rhesus monkeys. Exp Brain Res
two-dimensional reaching task. J Neurosci 9: 2080 –2102, 1989. 77: 245–256, 1989.

193. Kalaska JF, Cohen DA, Prud’homme M, Hyde ML. Parietal area 5 neuronal activity 215. Kurata K, Tanji J. Premotor cortex neurons in macaques: activity before distal and
encodes movement kinematics, not movement dynamics. Exp Brain Res 80: 351–364, proximal forelimb movements. J Neurosci 6: 403– 411, 1986.
1990.
216. Kwan HC, MacKay WA, Murphy JT, Wong YC. Spatial organization of precentral
194. Kalaska JF, Crammond DJ. Deciding not to GO: neuronal correlates of response cortex in awake primates. II. Motor outputs. J Neurophysiol 41: 1120 –1131, 1978.
selection in a GO/NOGO task in primate premotor and parietal cortex. Cereb Cortex
5: 410 – 428, 1995. 217. Laoprasert P. Atlas of Pediatric EEG. New York: McGraw-Hill, 2011.

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 701


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

218. Lawrence DG, Kuypers HG. The functional organization of the motor system in the 241. Lurija AR. The Working Brain: An Introduction to Neuropsychology. London: Penguin
monkey. I. The effects of bilateral pyramidal lesions. Brain 91: 1–14, 1968. Press, 1973.

219. Lehericy S, Ducros M, Krainik A, Francois C, Van de Moortele PF, Ugurbil K, Kim DS. 242. Mach E. Knowledge and Error: Sketches on the Psychology of Enquiry (English translation
3-D diffusion tensor axonal tracking shows distinct SMA and pre-SMA projections to 1967). Dordrecht: Reidel, 1905.
the human striatum. Cereb Cortex 14: 1302–1309, 2004.
243. MacNeilage PF. The frame/content theory of evolution of speech production. Behav
220. Lehmann D, Ozaki H, Pal I. EEG alpha map series: brain micro-states by space- Brain Sci 21: 499 –511, 1998.
oriented adaptive segmentation. Electroencephalogr Clin Neurophysiol 67: 271–288,
244. Maeda F, Kleiner-Fisman G, Pascual-Leone A. Motor facilitation while observing hand
1987.
actions: specificity of the effect and role of observer’s orientation. J Neurophysiol 87:
221. Leiguarda RC, Marsden CD. Limb apraxias: higher-order disorders of sensorimotor 1329 –1335, 2002.
integration. Brain 123: 860 – 879, 2000.
245. Malfait N, Valyear KF, Culham JC, Anton JL, Brown LE, Gribble PL. fMRI activation
222. Leinonen L, Hyvarinen J, Nyman G, Linnankoski II. Functional properties of neurons in during observation of others’ reach errors. J Cogn Neurosci 22: 1493–1503, 2010.
lateral part of associative area 7 in awake monkeys. Exp Brain Res 34: 299 –320, 1979.
246. Malle BF, Moses JL, Baldwin DA. Intentions and Intentionality: Foundations of Social
Cognition. Cambridge, MA: The MIT Press, 2001.
223. Lemon RN. Descending pathways in motor control. Annu Rev Neurosci 31: 195–218,
2008. 247. Manthey S, Schubotz RI, von Cramon DY. Premotor cortex in observing erroneous
action: an fMRI study. Brain Res Cogn Brain Res 15: 296 –307, 2003.
224. Lemon RN, Griffiths J. Comparing the function of the corticospinal system in different
species: organizational differences for motor specialization? Muscle Nerve 32: 261– 248. Marceglia S, Fiorio M, Foffani G, Mrakic-Sposta S, Tiriticco M, Locatelli M, Caputo E,
279, 2005. Tinazzi M, Priori A. Modulation of beta oscillations in the subthalamic area during
action observation in Parkinson’s disease. Neuroscience 161: 1027–1036, 2009.
225. Lemus L, Hernandez A, Romo R. Neural encoding of auditory discrimination in ventral
premotor cortex. Proc Natl Acad Sci USA 106: 14640 –14645, 2009. 249. Marchetti C, Della Sala S. Disentangling the alien and anarchic hand. Cogn Neuropsy-
chiatry 3: 191–207, 1998.
226. Leonard G, Tremblay F. Corticomotor facilitation associated with observation, imag-
ery and imitation of hand actions: a comparative study in young and old adults. Exp 250. Marconi B, Genovesio A, Battaglia-Mayer A, Ferraina S, Squatrito S, Molinari M,
Brain Res 177: 167–175, 2007. Lacquaniti F, Caminiti R. Eye-hand coordination during reaching. I. Anatomical rela-
tionships between parietal and frontal cortex. Cereb Cortex 11: 513–527, 2001.
227. Lepage JF, Theoret H. EEG evidence for the presence of an action observation-
execution matching system in children. Eur J Neurosci 23: 2505–2510, 2006. 251. Marconi B, Pecchioli C, Koch G, Caltagirone C. Functional overlap between hand and
forearm motor cortical representations during motor cognitive tasks. Clin Neuro-
228. Lewis JW, Brefczynski JA, Phinney RE, Janik JJ, DeYoe EA. Distinct cortical pathways physiol 118: 1767–1775, 2007.
for processing tool versus animal sounds. J Neurosci 25: 5148 –5158, 2005.
252. Marshall PJ, Bar-Haim Y, Fox NA. Development of the EEG from 5 months to 4 years
229. Lewis JW, Van Essen DC. Corticocortical connections of visual, sensorimotor, and of age. Clin Neurophysiol 113: 1199 –1208, 2002.
multimodal processing areas in the parietal lobe of the macaque monkey. J Comp
Neurol 428: 112–137, 2000. 253. Marteniuk RG, MacKenzie CL, Jeannerod M, Athenes S, Dugas C. Constraints on
human arm movement trajectories. Can J Psychol 41: 365–378, 1987.
230. Lhermitte F. Human autonomy and the frontal lobes. Part II: Patient behavior in
complex and social situations: the “environmental dependency syndrome.” Ann Neu- 254. Martineau J, Andersson F, Barthelemy C, Cottier JP, Destrieux C. Atypical activation
rol 19: 335–343, 1986. of the mirror neuron system during perception of hand motion in autism. Brain Res
1320: 168 –175, 2010.
231. Lhermitte F, Pillon B, Serdaru M. Human autonomy and the frontal lobes. Part I:
Imitation and utilization behavior: a neuropsychological study of 75 patients. Ann 255. Martineau J, Cochin S, Magne R, Barthelemy C. Impaired cortical activation in autistic
Neurol 19: 326 –334, 1986. children: is the mirror neuron system involved? Int J Psychophysiol 68: 35– 40, 2008.

232. Liepert J, Neveling N. Motor excitability during imagination and observation of foot 256. Massion J. Movement, posture and equilibrium: interaction and coordination. Prog
Neurobiol 38: 35–56, 1992.
dorsiflexions. J Neural Transm 116: 1613–1619, 2009.
257. Matelli M, Camarda R, Glickstein M, Rizzolatti G. Afferent and efferent projections of
233. Liepmann H. Apraxie. Ergebnisse der gesamten Medizin 1: 516 –543, 1920.
the inferior area 6 in the macaque monkey. J Comp Neurol 251: 281–298, 1986.
234. Liew SL, Sheng T, Aziz-Zadeh L. Experience with an amputee modulates one’s own
258. Matelli M, Govoni P, Galletti C, Kutz DF, Luppino G. Superior area 6 afferents from
sensorimotor response during action observation. Neuroimage 69: 138 –145, 2013.
the superior parietal lobule in the macaque monkey. J Comp Neurol 402: 327–352,
235. Lingnau A, Gesierich B, Caramazza A. Asymmetric fMRI adaptation reveals no evi- 1998.
dence for mirror neurons in humans. Proc Natl Acad Sci USA 106: 9925–9930, 2009.
259. Matelli M, Luppino G, Rizzolatti G. Architecture of superior and mesial area 6 and the
236. Lu MT, Preston JB, Strick PL. Interconnections between the prefrontal cortex and the adjacent cingulate cortex in the macaque monkey. J Comp Neurol 311: 445– 462, 1991.
premotor areas in the frontal lobe. J Comp Neurol 341: 375–392, 1994.
260. Matelli M, Luppino G, Rizzolatti G. Patterns of cytochrome oxidase activity in the
frontal agranular cortex of the macaque monkey. Behav Brain Res 18: 125–136, 1985.
237. Luppino G, Matelli M, Camarda R, Rizzolatti G. Corticocortical connections of area F3
(SMA-proper) and area F6 (pre-SMA) in the macaque monkey. J Comp Neurol 338: 261. Matsuzaka Y, Aizawa H, Tanji J. A motor area rostral to the supplementary motor area
114 –140, 1993. (presupplementary motor area) in the monkey: neuronal activity during a learned
motor task. J Neurophysiol 68: 653– 662, 1992.
238. Luppino G, Matelli M, Camarda RM, Gallese V, Rizzolatti G. Multiple representations
of body movements in mesial area 6 and the adjacent cingulate cortex: an intracortical 262. Maunsell JH, van Essen DC. The connections of the middle temporal visual area (MT)
microstimulation study in the macaque monkey. J Comp Neurol 311: 463– 482, 1991. and their relationship to a cortical hierarchy in the macaque monkey. J Neurosci 3:
2563–2586, 1983.
239. Luppino G, Murata A, Govoni P, Matelli M. Largely segregated parietofrontal connec-
tions linking rostral intraparietal cortex (areas AIP and VIP) and the ventral premotor 263. Miniussi C, Harris JA, Ruzzoli M. Modelling non-invasive brain stimulation in cognitive
cortex (areas F5 and F4). Exp Brain Res 128: 181–187, 1999. neuroscience. Neurosci Biobehav Rev 37: 1702–1712, 2013.

240. Luppino G, Rozzi S, Calzavara R, Matelli M. Prefrontal and agranular cingulate projec- 264. Mita A, Mushiake H, Shima K, Matsuzaka Y, Tanji J. Interval time coding by neurons in
tions to the dorsal premotor areas F2 and F7 in the macaque monkey. Eur J Neurosci the presupplementary and supplementary motor areas. Nat Neurosci 12: 502–507,
17: 559 –578, 2003. 2009.

702 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

265. Mitz AR, Wise SP. The somatotopic organization of the supplementary motor area: 287. Nystrom P. The infant mirror neuron system studied with high density EEG. Soc
intracortical microstimulation mapping. J Neurosci 7: 1010 –1021, 1987. Neurosci 3: 334 –347, 2008.

266. Mizobuchi K, Takahasi N, Ajima A. [Utilization and imitation behavior following right 288. Nystrom P, Ljunghammar T, Rosander K, von Hofsten C. Using mu rhythm desyn-
parietotemporal lesions]. Rinsho Shinkeigaku 51: 350 –353, 2011. chronization to measure mirror neuron activity in infants. Dev Sci 14: 327–335, 2011.

267. Milner EJ, Goodale MA. Cortical and tectal control of visual orientation in the gerbil: 289. Oberman LM, Hubbard EM, McCleery JP, Altschuler EL, Ramachandran VS, Pineda
evidence for parallel channels. Exp Brain Res 55: 33– 48, 1984. JA. EEG evidence for mirror neuron dysfunction in autism spectrum disorders. Brain
Res Cogn Brain Res 24: 190 –198, 2005.
268. Molenberghs P, Cunnington R, Mattingley JB. Brain regions with mirror properties: a
meta-analysis of 125 human fMRI studies. Neurosci Biobehav Rev 36: 341–349, 2012. 290. Oberman LM, Pineda JA, Ramachandran VS. The human mirror neuron system: a link
between action observation and social skills. Soc Cogn Affect Neurosci 2: 62– 66, 2007.
269. Molenberghs P, Hayward L, Mattingley JB, Cunnington R. Activation patterns during
action observation are modulated by context in mirror system areas. Neuroimage 59: 291. Oberman LM, Ramachandran VS, Pineda JA. Modulation of mu suppression in children
608 – 615, 2012. with autism spectrum disorders in response to familiar or unfamiliar stimuli: the
mirror neuron hypothesis. Neuropsychologia 46: 1558 –1565, 2008.
270. Montagna M, Cerri G, Borroni P, Baldissera F. Excitability changes in human cortico-
292. Orgs G, Dombrowski JH, Heil M, Jansen-Osmann P. Expertise in dance modulates
spinal projections to muscles moving hand and fingers while viewing a reaching and
alpha/beta event-related desynchronization during action observation. Eur J Neurosci
grasping action. Eur J Neurosci 22: 1513–1520, 2005.
27: 3380 –3384, 2008.
271. Mukamel R, Ekstrom AD, Kaplan J, Iacoboni M, Fried I. Single-neuron responses in
293. Ortigue S, Sinigaglia C, Rizzolatti G, Grafton ST. Understanding actions of others: the
humans during execution and observation of actions. Curr Biol 20: 750 –756, 2010.
electrodynamics of the left and right hemispheres. A high-density EEG neuroimaging
272. Mukamel R, Fried I. Human intracranial recordings and cognitive neuroscience. Annu study. PLoS ONE 5: e12160, 2010.
Rev Psychol 63: 511–537, 2012. 294. Pacherie E. The anarchic hand syndrome and utilization behavior: a window onto
agentive self-awareness. Funct Neurol 22: 211–217, 2007.
273. Muller RA, Kleinhans N, Kemmotsu N, Pierce K, Courchesne E. Abnormal variability
and distribution of functional maps in autism: an FMRI study of visuomotor learning. 295. Pardo-Vazquez JL, Leboran V, Acuna C. Neural correlates of decisions and their
Am J Psychiatry 160: 1847–1862, 2003. outcomes in the ventral premotor cortex. J Neurosci 28: 12396 –12408, 2008.
274. Murakami T, Restle J, Ziemann U. Observation-execution matching and action inhi- 296. Pardo-Vazquez JL, Leboran V, Acuna C. A role for the ventral premotor cortex
bition in human primary motor cortex during viewing of speech-related lip move- beyond performance monitoring. Proc Natl Acad Sci USA 106: 18815–18819, 2009.
ments or listening to speech. Neuropsychologia 49: 2045–2054, 2011.
297. Passingham RE. The Frontal Lobe and Voluntary Action. Oxford, UK: Oxford Univ.
275. Murata A, Fadiga L, Fogassi L, Gallese V, Raos V, Rizzolatti G. Object representation Press, 1993.
in the ventral premotor cortex (area F5) of the monkey. J Neurophysiol 78: 2226 –
2230, 1997. 298. Patuzzo S, Fiaschi A, Manganotti P. Modulation of motor cortex excitability in the left
hemisphere during action observation: a single- and paired-pulse transcranial mag-
276. Murata A, Gallese V, Luppino G, Kaseda M, Sakata H. Selectivity for the shape, size, netic stimulation study of self- and non-self-action observation. Neuropsychologia 41:
and orientation of objects for grasping in neurons of monkey parietal area AIP. J 1272–1278, 2003.
Neurophysiol 83: 2580 –2601, 2000.
299. Paulus M, Hunnius S, van Elk M, Bekkering H. How learning to shake a rattle affects
277. Murphy JT, Kwan HC, MacKay WA, Wong YC. Spatial organization of precentral 8-month-old infants’ perception of the rattle’s sound: electrophysiological evidence
cortex in awake primates. III. Input-output coupling. J Neurophysiol 41: 1132–1139, for action-effect binding in infancy. Dev Cogn Neurosci 2: 90 –96, 2012.
1978.
300. Pazzaglia M, Pizzamiglio L, Pes E, Aglioti SM. The sound of actions in apraxia. Curr Biol
278. Nachev P, Kennard C, Husain M. Functional role of the supplementary and pre- 18: 1766 –1772, 2008.
supplementary motor areas. Nat Rev Neurosci 9: 856 – 869, 2008.
301. Peeters R, Simone L, Nelissen K, Fabbri-Destro M, Vanduffel W, Rizzolatti G, Orban
279. Nakamura H, Kitagawa H, Kawaguchi Y, Tsuji H. Intracortical facilitation and inhibi- GA. The representation of tool use in humans and monkeys: common and uniquely
tion after transcranial magnetic stimulation in conscious humans. J Physiol 498: 817– human features. J Neurosci 29: 11523–11539, 2009.
823, 1997.
302. Pena-Casanova J, Roig-Rovira T, Bermudez A, Tolosa-Sarro E. Optic aphasia, optic
apraxia, and loss of dreaming. Brain Lang 26: 63–71, 1985.
279a.Nakamura K, Sakai K, Hikosaka O. Neuronal activity in medial frontal cortex during
learning of sequential procedures. J Neurophysiol 80: 2671–2687, 1998. 303. Penfield W, Welch K. The supplementary motor area of the cerebral cortex: a clinical
and experimental study. AMA Arch Neurol Psychiatry 66: 289 –317, 1951.
280. Neal JW, Pearson RC, Powell TP. The cortico-cortical connections of area 7b, PF, in
the parietal lobe of the monkey. Brain Res 419: 341–346, 1987. 304. Perani D, Fazio F, Borghese NA, Tettamanti M, Ferrari S, Decety J, Gilardi MC.
Different brain correlates for watching real and virtual hand actions. Neuroimage 14:
281. Negri GA, Rumiati RI, Zadini A, Ukmar M, Mahon BZ, Caramazza A. What is the role
749 –758, 2001.
of motor simulation in action and object recognition? Evidence from apraxia. Cogn
Neuropsychol 24: 795– 816, 2007. 305. Perini F, Cattaneo L, Carrasco M, Schwarzbach JV. Occipital transcranial magnetic
stimulation has an activity-dependent suppressive effect. J Neurosci 32: 12361–12365,
282. Nelissen K, Borra E, Gerbella M, Rozzi S, Luppino G, Vanduffel W, Rizzolatti G, Orban 2012.
GA. Action observation circuits in the macaque monkey cortex. J Neurosci 31: 3743–
3756, 2011. 306. Perrett DI, Harries MH, Bevan R, Thomas S, Benson PJ, Mistlin AJ, Chitty AJ, Hietanen
JK, Ortega JE. Frameworks of analysis for the neural representation of animate objects
283. Nelissen K, Luppino G, Vanduffel W, Rizzolatti G, Orban GA. Observing others: and actions. J Exp Biol 146: 87–113, 1989.
multiple action representation in the frontal lobe. Science 310: 332–336, 2005.
307. Perry A, Stein L, Bentin S. Motor and attentional mechanisms involved in social
284. Nishitani N, Avikainen S, Hari R. Abnormal imitation-related cortical activation se- interaction– evidence from mu and alpha EEG suppression. Neuroimage 58: 895–904,
quences in Asperger’s syndrome. Ann Neurol 55: 558 –562, 2004. 2011.

285. Nishitani N, Hari R. Temporal dynamics of cortical representation for action. Proc Natl 308. Petrides M, Pandya DN. Projections to the frontal cortex from the posterior parietal
Acad Sci USA 97: 913–918, 2000. region in the rhesus monkey. J Comp Neurol 228: 105–116, 1984.

286. Nishitani N, Hari R. Viewing lip forms: cortical dynamics. Neuron 36: 1211–1220, 309. Pfurtscheller G, Neuper C. Motor imagery activates primary sensorimotor area in
2002. humans. Neurosci Lett 239: 65– 68, 1997.

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 703


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

310. Pobric G, Hamilton AF. Action understanding requires the left inferior frontal cortex. 335. Rizzolatti G, Matelli M. Two different streams form the dorsal visual system: anatomy
Curr Biol 16: 524 –529, 2006. and functions. Exp Brain Res 153: 146 –157, 2003.

311. Poincaré HJ. Science and Method (English translation 1996). London: Routledge, 1908. 336. Rizzolatti G, Matelli M, Pavesi G. Deficits in attention and movement following the
removal of postarcuate (area 6) and prearcuate (area 8) cortex in macaque monkeys.
312. Porter R, Lemon RN. Corticospinal Function and Voluntary Movement. New York:
Brain 106: 655– 673, 1983.
Oxford Univ. Press, 1995.
337. Rizzolatti G, Sinigaglia C. The functional role of the parieto-frontal mirror circuit:
313. Prabhu G, Shimazu H, Cerri G, Brochier T, Spinks RL, Maier MA, Lemon RN. Mod-
interpretations and misinterpretations. Nat Rev Neurosci 11: 264 –274, 2010.
ulation of primary motor cortex outputs from ventral premotor cortex during visually
guided grasp in the macaque monkey. J Physiol 587: 1057–1069, 2009. 338. Rochat MJ, Caruana F, Jezzini A, Escola L, Intskirveli I, Grammont F, Gallese V,
Rizzolatti G, Umilta MA. Responses of mirror neurons in area F5 to hand and tool
314. Prather JF, Peters S, Nowicki S, Mooney R. Precise auditory-vocal mirroring in neu-
grasping observation. Exp Brain Res 204: 605– 616, 2010.
rons for learned vocal communication. Nature 451: 305–310, 2008.

315. Preuss TM, Goldman-Rakic PS. Connections of the ventral granular frontal cortex of 339. Rogers SJ. An examination of the imitation deficit in autism. In: Imitation in Infancy,
macaques with perisylvian premotor and somatosensory areas: anatomical evidence edited by Nadel J and Butterworth G. Cambridge: Cambridge Univ. Press, 1999, p.
for somatic representation in primate frontal association cortex. J Comp Neurol 282: 254 –283.
293–316, 1989.
340. Romo R, Hernandez A, Zainos A. Neuronal correlates of a perceptual decision in
316. Prinz W. A common coding approach to perception and action. In: Cognition and ventral premotor cortex. Neuron 41: 165–173, 2004.
Motor Processes, edited by Neumann O and Prinz W. Berlin: Springer, 1990, p.
167–201. 341. Rothi LJ, Heilman KM, Watson RT. Pantomime comprehension and ideomotor
apraxia. J Neurol Neurosurg Psychiatry 48: 207–210, 1985.
317. Prinz W. Open Minds: The Social Making of Agency and Intentionality. Cambridge, MA:
MIT Press, 2012. 342. Roy AC, Craighero L, Fabbri-Destro M, Fadiga L. Phonological and lexical motor
facilitation during speech listening: a transcranial magnetic stimulation study. J Physiol
318. Prinz W. Perception and action planning. Eur J Cogn Psychol 9: 129 –154, 1997. 102: 101–105, 2008.

319. Puce A, Perrett D. Electrophysiology and brain imaging of biological motion. Philos 343. Rozzi S, Calzavara R, Belmalih A, Borra E, Gregoriou GG, Matelli M, Luppino G.
Trans R Soc Lond B Biol Sci 358: 435– 445, 2003. Cortical connections of the inferior parietal cortical convexity of the macaque mon-
key. Cereb Cortex 16: 1389 –1417, 2006.
320. Puzzo I, Cooper NR, Vetter P, Russo R, Fitzgerald PB. Reduced cortico-motor facil-
itation in a normal sample with high traits of autism. Neurosci Lett 467: 173–177, 2009. 344. Rozzi S, Ferrari PF, Bonini L, Rizzolatti G, Fogassi L. Functional organization of inferior
parietal lobule convexity in the macaque monkey: electrophysiological characteriza-
321. Ramachandran VS, Oberman LM. Broken mirrors: a theory of autism. Sci Am 295:
tion of motor, sensory and mirror responses and their correlation with cytoarchitec-
62– 69, 2006.
tonic areas. Eur J Neurosci 28: 1569 –1588, 2008.
322. Raos V, Franchi G, Gallese V, Fogassi L. Somatotopic organization of the lateral part of
345. Saby JN, Marshall PJ, Meltzoff AN. Neural correlates of being imitated: an EEG study
area F2 (dorsal premotor cortex) of the macaque monkey. J Neurophysiol 89: 1503–
in preverbal infants. Soc Neurosci 7: 650 – 661, 2012.
1518, 2003.

323. Raos V, Umilta MA, Gallese V, Fogassi L. Functional properties of grasping-related 346. Sakai K, Hikosaka O, Miyauchi S, Sasaki Y, Fujimaki N, Putz B. Presupplementary
neurons in the dorsal premotor area F2 of the macaque monkey. J Neurophysiol 92: motor area activation during sequence learning reflects visuo-motor association. J
1990 –2002, 2004. Neurosci 19: RC1, 1999.

324. Raos V, Umilta MA, Murata A, Fogassi L, Gallese V. Functional properties of grasping- 347. Sakata H, Taira M, Murata A, Mine S. Neural mechanisms of visual guidance of hand
related neurons in the ventral premotor area F5 of the macaque monkey. J Neuro- action in the parietal cortex of the monkey. Cereb Cortex 5: 429 – 438, 1995.
physiol 95: 709 –729, 2006.
348. Sakreida K, Schubotz RI, Wolfensteller U, von Cramon DY. Motion class dependency
325. Rathelot JA, Strick PL. Subdivisions of primary motor cortex based on cortico-mo- in observers’ motor areas revealed by functional magnetic resonance imaging. J Neu-
toneuronal cells. Proc Natl Acad Sci USA 106: 918 –923, 2009. rosci 25: 1335–1342, 2005.

326. Raymaekers R, Wiersema JR, Roeyers H. EEG study of the mirror neuron system in 349. Saleem KS, Kondo H, Price JL. Complementary circuits connecting the orbital and
children with high functioning autism. Brain Res 1304: 113–121, 2009. medial prefrontal networks with the temporal, insular, and opercular cortex in the
macaque monkey. J Comp Neurol 506: 659 – 693, 2008.
327. Rizzolatti G, Arbib MA. Language within our grasp. Trends Neurosci 21: 188 –194,
1998. 350. Sawamura H, Orban GA, Vogels R. Selectivity of neuronal adaptation does not match
response selectivity: a single-cell study of the FMRI adaptation paradigm. Neuron 49:
328. Rizzolatti G, Camarda R, Fogassi L, Gentilucci M, Luppino G, Matelli M. Functional
307–318, 2006.
organization of inferior area 6 in the macaque monkey. II. Area F5 and the control of
distal movements. Exp Brain Res 71: 491–507, 1988. 351. Schall JD, Morel A, Kaas JH. Topography of supplementary eye field afferents to
frontal eye field in macaque: implications for mapping between saccade coordinate
329. Rizzolatti G, Craighero L. The mirror-neuron system. Annu Rev Neurosci 27: 169 –192,
systems. Vis Neurosci 10: 385–393, 1993.
2004.
352. Schlag J, Schlag-Rey M. Evidence for a supplementary eye field. J Neurophysiol 57:
330. Rizzolatti G, Fadiga L, Gallese V, Fogassi L. Premotor cortex and the recognition of
motor actions. Brain Res Cogn Brain Res 3: 131–141, 1996. 179 –200, 1987.

331. Rizzolatti G, Fadiga L, Matelli M, Bettinardi V, Paulesu E, Perani D, Fazio F. Localiza- 353. Schubotz RI, Anwander A, Knosche TR, von Cramon DY, Tittgemeyer M. Anatomical
tion of grasp representations in humans by PET: 1. Observation versus execution. Exp and functional parcellation of the human lateral premotor cortex. Neuroimage 50:
Brain Res 111: 246 –252, 1996. 396 – 408, 2010.

332. Rizzolatti G, Fogassi L, Gallese V. Parietal cortex: from sight to action. Curr Opin 354. Schutz-Bosbach S, Avenanti A, Aglioti SM, Haggard P. Don’t do it! Cortical inhibition
Neurobiol 7: 562–567, 1997. and self-attribution during action observation. J Cogn Neurosci 21: 1215–1227, 2009.

333. Rizzolatti G, Gentilucci M, Camarda RM, Gallese V, Luppino G, Matelli M, Fogassi L. 355. Schutz-Bosbach S, Mancini B, Aglioti SM, Haggard P. Self and other in the human
Neurons related to reaching-grasping arm movements in the rostral part of area 6 motor system. Curr Biol 16: 1830 –1834, 2006.
(area 6a beta). Exp Brain Res 82: 337–350, 1990.
356. Schutz-Bosbach S, Prinz W. Perceptual resonance: action-induced modulation of
334. Rizzolatti G, Luppino G. The cortical motor system. Neuron 31: 889 –901, 2001. perception. Trends Cogn Sci 11: 349 –355, 2007.

704 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
ACTION ORGANIZATION AND ACTION UNDERSTANDING

357. Schwarzkopf DS, Silvanto J, Rees G. Stochastic resonance effects reveal the neural 381. Tomasello M, Call J, Gluckman A. Comprehension of novel communicative signs
mechanisms of transcranial magnetic stimulation. J Neurosci 31: 3143–3147, 2011. by apes and human children. Child Dev 68: 1067–1080, 1997.

358. Seltzer B, Pandya DN. Parietal, temporal, and occipital projections to cortex of the 382. Tomassini V, Jbabdi S, Klein JC, Behrens TE, Pozzilli C, Matthews PM, Rushworth
superior temporal sulcus in the rhesus monkey: a retrograde tracer study. J Comp MF, Johansen-Berg H. Diffusion-weighted imaging tractography-based parcella-
Neurol 343: 445– 463, 1994. tion of the human lateral premotor cortex identifies dorsal and ventral sub-
regions with anatomical and functional specializations. J Neurosci 27: 10259 –
359. Shima K, Tanji J. Neuronal activity in the supplementary and presupplementary motor 10269, 2007.
areas for temporal organization of multiple movements. J Neurophysiol 84: 2148 –
2160, 2000. 383. Tower S. Pyramidal lesion in the monkey. Brain 63: 36 –90, 1940.

360. Shimada S. Deactivation in the sensorimotor area during observation of a human agent 384. Tremblay F, Leonard G, Tremblay L. Corticomotor facilitation associated with ob-
performing robotic actions. Brain Cogn 72: 394 –399, 2010. servation and imagery of hand actions is impaired in Parkinson’s disease. Exp Brain Res
185: 249 –257, 2008.
361. Shimazu H, Maier MA, Cerri G, Kirkwood PA, Lemon RN. Macaque ventral premotor
cortex exerts powerful facilitation of motor cortex outputs to upper limb motoneu- 385. Tsukazaki I, Uehara K, Morishita T, Ninomiya M, Funase K. Effect of observation com-
rons. J Neurosci 24: 1200 –1211, 2004. bined with motor imagery of a skilled hand-motor task on motor cortical excitability:
difference between novice and expert. Neurosci Lett 518: 96 –100, 2012.
362. Shimomura T, Mori E. Obstinate imitation behaviour in differentiation of frontotem-
poral dementia from Alzheimer’s disease. Lancet 352: 623– 624, 1998. 386. Ubaldi S, Barchiesi G, Cattaneo L. Bottom-up and top-down responses to action
observation. Cereb Cortex. In press.
363. Shmuelof L, Zohary E. Dissociation between ventral and dorsal fMRI activation during
object and action recognition. Neuron 47: 457– 470, 2005. 387. Umilta MA, Escola L, Intskirveli I, Grammont F, Rochat M, Caruana F, Jezzini A,
Gallese V, Rizzolatti G. When pliers become fingers in the monkey motor system. Proc
364. Shmuelof L, Zohary E. Mirror-image representation of action in the anterior parietal Natl Acad Sci USA 105: 2209 –2213, 2008.
cortex. Nat Neurosci 11: 1267–1269, 2008.
388. Umilta MA, Kohler E, Gallese V, Fogassi L, Fadiga L, Keysers C, Rizzolatti G. I know
365. Shmuelof L, Zohary E. A mirror representation of others’ actions in the human what you are doing. A neurophysiological study. Neuron 31: 155–165, 2001.
anterior parietal cortex. J Neurosci 26: 9736 –9742, 2006.
389. Ungerleider LG, Desimone R. Cortical connections of visual area MT in the macaque.
366. Silvanto J, Pascual-Leone A. State-dependency of transcranial magnetic stimulation. J Comp Neurol 248: 190 –222, 1986.
Brain Topogr 21: 1–10, 2008.
390. Urgesi C, Candidi M, Fabbro F, Romani M, Aglioti SM. Motor facilitation during action
367. Singer T, Seymour B, O’Doherty J, Kaube H, Dolan RJ, Frith CD. Empathy for pain observation: topographic mapping of the target muscle and influence of the onlook-
involves the affective but not sensory components of pain. Science 303: 1157–1162, er’s posture. Eur J Neurosci 23: 2522–2530, 2006.
2004.
391. Urgesi C, Candidi M, Ionta S, Aglioti SM. Representation of body identity and body
368. Stefan K, Classen J, Celnik P, Cohen LG. Concurrent action observation modu- actions in extrastriate body area and ventral premotor cortex. Nat Neurosci 10:
lates practice-induced motor memory formation. Eur J Neurosci 27: 730 –738, 30 –31, 2007.
2008. 392. Vaadia E, Benson DA, Hienz RD, Goldstein MH Jr. Unit study of monkey frontal
cortex: active localization of auditory and of visual stimuli. J Neurophysiol 56: 934 –952,
369. Stefan K, Cohen LG, Duque J, Mazzocchio R, Celnik P, Sawaki L, Ungerleider L,
1986.
Classen J. Formation of a motor memory by action observation. J Neurosci 25: 9339 –
9346, 2005. 393. Van Elk M, van Schie HT, Hunnius S, Vesper C, Bekkering H. You’ll never crawl alone:
neurophysiological evidence for experience-dependent motor resonance in infancy.
370. Strafella AP, Paus T. Modulation of cortical excitability during action observation: a
Neuroimage 43: 808 – 814, 2008.
transcranial magnetic stimulation study. Neuroreport 11: 2289 –2292, 2000.
394. Van Hooff JARAM. The facial displays of the catarrhine monkeys and apes. In: Primate
371. Stroganova TA, Orekhova EV, Posikera IN. EEG alpha rhythm in infants. Clin Neuro-
Ethology, edited by Morris D. Chicago: Aldine, 1967.
physiol 110: 997–1012, 1999.
395. Vigneswaran G, Philipp R, Lemon RN, Kraskov A. M1 corticospinal mirror neurons
372. Sundara M, Namasivayam AK, Chen R. Observation-execution matching system for
and their role in movement suppression during action observation. Curr Biol 23:
speech: a magnetic stimulation study. Neuroreport 12: 1341–1344, 2001.
236 –243, 2013.
373. Taira M, Mine S, Georgopoulos AP, Murata A, Sakata H. Parietal cortex neurons of the 396. Vogt C, Vogt O. Allgemeine Ergebnisse unserer Hirnforschung. Journal für Psychologie
monkey related to the visual guidance of hand movement. Exp Brain Res 83: 29 –36, und Neurologie 25: 277– 462, 1919.
1990.
397. Von Bonnin G, Bailey P. The Neocortex of Macaca Mulatta. Urbana, IL: Univ. of Illinois
374. Takahashi N, Kawamura M, Yamazaki M. Imitation behavior: clinical and pathophys- Press, 1925.
iological study of seven patients. Higher Brain Funct Res 19: 93–100, 1999.
398. Von Economo K, Koskinas GN. Die Cytoarchitektonik der Hirnrinde des erwachsenen
375. Tanji J. Sequential organization of multiple movements: involvement of cortical motor Mensche. Wien: Springer Verlag, 1925.
areas. Annu Rev Neurosci 24: 631– 651, 2001.
399. Watkins K, Paus T. Modulation of motor excitability during speech perception: the
376. Tanji J, Shima K. Role for supplementary motor area cells in planning several move- role of Broca’s area. J Cogn Neurosci 16: 978 –987, 2004.
ments ahead. Nature 371: 413– 416, 1994.
400. Watkins KE, Strafella AP, Paus T. Seeing and hearing speech excites the motor system
377. Tanji J, Shima K. Supplementary motor cortex in organization of movement. Eur involved in speech production. Neuropsychologia 41: 989 –994, 2003.
Neurol 36 Suppl 1: 13–19, 1996.
401. Wheaton KJ, Thompson JC, Syngeniotis A, Abbott DF, Puce A. Viewing the motion of
378. Theoret H, Halligan E, Kobayashi M, Fregni F, Tager-Flusberg H, Pascual-Leone A. human body parts activates different regions of premotor, temporal, and parietal
Impaired motor facilitation during action observation in individuals with autism spec- cortex. Neuroimage 22: 277–288, 2004.
trum disorder. Curr Biol 15: R84 – 85, 2005.
402. Williams JH, Whiten A, Suddendorf T, Perrett DI. Imitation, mirror neurons and
379. Thier P, Andersen RA. Electrical microstimulation distinguishes distinct saccade- autism. Neurosci Biobehav Rev 25: 287–295, 2001.
related areas in the posterior parietal cortex. J Neurophysiol 80: 1713–1735, 1998.
403. Wise SP, Boussaoud D, Johnson PB, Caminiti R. Premotor and parietal cortex: corti-
380. Tkach D, Reimer J, Hatsopoulos NG. Congruent activity during action and action cocortical connectivity and combinatorial computations. Annu Rev Neurosci 20: 25–
observation in motor cortex. J Neurosci 27: 13241–13250, 2007. 42, 1997.

Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org 705


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.
RIZZOLATTI ET AL.

404. Wood JN, Glynn DD, Phillips BC, Hauser MD. The perception of rational, goal- 409. Yoshida K, Saito N, Iriki A, Isoda M. Representation of others’ action by neurons in
directed action in nonhuman primates. Science 317: 1402–1405, 2007. monkey medial frontal cortex. Curr Biol 21: 249 –253, 2011.

405. Wood JN, Hauser MD. Action comprehension in non-human primates: motor simu- 410. Ziemann U, Rothwell JC. I-waves in motor cortex. J Clin Neurophysiol 17: 397–405, 2000.
lation or inferential reasoning? Trends Cogn Sci 12: 461– 465, 2008.
411. Ziemann U, Tergau F, Wassermann EM, Wischer S, Hildebrandt J, Paulus W. Dem-
406. Woolsey CN, Settlage PH, Meyer DR, Sencer W, Hamuy TP, Travis AM. Pattern of onstration of facilitatory I wave interaction in the human motor cortex by paired
Localization in Precentral and “Supplementary” Motor Areas and Their Relation to the transcranial magnetic stimulation. J Physiol 511: 181–190, 1998.
Concept of a Premotor Area. New York: Raven, 1952.
412. Zilles K, Schlaug G, Geyer S, Luppino G, Matelli M, Qu M, Schleicher A, Schormann T.
407. Wyk BC, Hudac CM, Carter EJ, Sobel DM, Pelphrey KA. Action understanding in the Anatomy and transmitter receptors of the supplementary motor areas in the human
superior temporal sulcus region. Psychol Sci 20: 771–777, 2009. and nonhuman primate brain. Adv Neurol 70: 29 – 43, 1996.

408. Yokochi H, Tanaka M, Kumashiro M, Iriki A. Inferior parietal somatosensory neurons 413. Zilles K, Schlaug G, Matelli M, Luppino G, Schleicher A, Qu M, Dabringhaus A, Seitz R,
coding face-hand coordination in Japanese macaques. Somatosensory motor Res 20: Roland PE. Mapping of human and macaque sensorimotor areas by integrating archi-
115–125, 2003. tectonic, transmitter receptor, MRI and PET data. J Anat 187: 515–537, 1995.

706 Physiol Rev • VOL 94 • APRIL 2014 • www.prv.org


Downloaded from www.physiology.org/journal/physrev by ${individualUser.givenNames} ${individualUser.surname} (187.214.249.146) on June 5, 2018.
Copyright © 2014 American Physiological Society. All rights reserved.

You might also like