10.1177/1534582304267187 ARTICLEBEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWSDecety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY
The Functional Architecture of Human Empathy
Philip L. Jackson
University of Washington
Empathy accounts for the naturally occurring subjective experi-
ence of similarity between the feelings expressed by self and others
without loosing sight of whose feelings belong to whom. Empathy
involves not only the affective experience of the other person’s
actual or inferred emotional state but also some minimal recog-
nition and understanding of another’s emotional state. In light
of multiple levels of analysis ranging from developmental psy-
chology, social psychology, cognitive neuroscience, and clinical
neuropsychology, this article proposes a model of empathy that
involves parallel and distributed processing in a number of
dissociable computational mechanisms. Shared neural represen-
tations, self-awareness, mental flexibility, and emotion regula-
tion constitute the basic macrocomponents of empathy, which are
underpinned by specific neural systems. This functional model
may be used to make specific predictions about the various empa-
thy deficits that can be encountered in different forms of social
and neurological disorders.
Key Words: self-awareness, intersubjectivity, affective
sharing, perspective taking, executive inhibition,
shared representations, emotion regulation
You are peacefully reading your favorite newspaper
while your child is playing with others in a playpen
nearby, when suddenly, she cries. It does not take long to
orient your attention toward her, perceive her distressed
state, and understand what she feels. Not only do you
perceive her plight, but you also actively want to comfort
her. This natural ability to understand the emotions and
feelings of others, whether one actually witnessed his or
her situation, perceived it from a photograph, read
about it in fiction book, or merely imagined it, refers to
the phenomenological experience of empathy. This
“every-day mind reading,” to borrow Ickes’s (2003) met-
aphor, is not something one needs to learn. Rather, the
basic building blocks are hardwired in the brain and
await development through interaction with others.
Such a capacity to understand others and experience
their feelings in relation to oneself illustrates the social
nature of the self, inherently intersubjective. Humans
are indeed social animals, and virtually all of their
actions (including their thoughts and desires) are
directed toward or are produced in response to others
Empathy denotes, at a phenomenological level of
description, a sense of similarity between the feelings
one experiences and those expressed by others. This
sharing of the feelings of another person does not neces-
sarily imply that one will act or even feel impelled to act
in a supportive or sympathetic way. Given the complexity
of this construct, we believe that only a multidisci-
plinary approach can help to better understand the
information-processing mechanisms that give rise to this
subjective psychological phenomenon. Our ambition in
this article is to articulate different domains of research,
including developmental science, cognitive and social
psychology, and neuroscience. In addition, instead of
addressing each of these research domains separately,
we integrate data from these different approaches with
the guidance of a putative model. This model should
be considered as a heuristic tool for future research,
especially to foster new investigations of social behav-
ior disorders (e.g., anti-social personality disorders),
whether they are clinical observations or empirical stud-
ies, as well as to cast some light into empathy deficits
observed in brain damaged patients. We also hope that
this endeavor adequately illustrates the emerging field of
social-cognitive neuroscience (Ochsner & Lieberman,
Authors’ Note: Dr. Philip L. Jackson is supported by a grant from the
Canadian Institute for Health Research. Please address correspon-
dence to Professor Jean Decety, Head of Social Cognitive Neuroscience
at the Institute for Learning and Brain Sciences, University of Wash-
ington, Box 357988, Seattle, WA 98195-7988, USA; e-mail: decety@
Behavioral and Cognitive Neuroscience Reviews
Volume 3 Number 2, June 2004 71-100
© 2004 Sage Publications
It is unlikely that empathy is the product of random
mutation and just happened in humans without any evo-
lutionary history. During the evolution of the mammal
and primate brain, the organization of the neural activity
has been shaped by the need for rapid evaluation of the
motivations of others (Brothers, 1989). Indeed, affective
communication is widely distributed in the animal king-
dom (e.g., Buck & Ginsburg, 1997; Preston & de Waal,
2002). It has a survival value and contributes to inclusive
fitness because it assists individuals in gathering and
hunting for food, detecting predators, courtship, and
ensuring reproductive success (Plutchik, 1987).
It is not, however, evident how selective pressures tai-
lor such superordinate categories as empathy or social
cognition. Selection operates at the level of function, not
at the level of physical structures or behaviors that sub-
serve the function. Evolution does not create specific
behaviors; it creates mental organizations and infer-
ence systems that make people behave in particular ways
(Boyer, 2001). Tooby and Cosmides (1996) proposed
that humans have a specialized computational device—
an implicit theory of human nature—that models what
motivations and mental representations others would
develop when placed in various evolutionarily recurrent
situations. But this does not mean there is a single mod-
ule in the brain for such a device. Rather, there is a collec-
tion of separate systems whose combination produces
typically human “mind reading” and the hypertrophied
social intelligence (Boyer & Barrett, 2004). Another
interesting argument considers that social complexity
(often indexed by group size) has been a driving force in
brain evolution and that the demands of navigating
more complex social landscapes constitutes a unique
selection pressure among the ancestral apes for in-
creased brain size and cognitive abilities (Dunbar, 1998).
Advanced levels of social cognition may have arisen as an
emergent property of powerful executive functioning
assisted by the representational properties of language
(Barrett, Henzi, & Dunbar, 2003). In addition, the bal-
ance of cost and benefits for the individual who
expresses feelings and the observer who interacts with
this individual has implications for survival in social
groups. Empathic concern is often associated with pro-
social behaviors, such as helping a kin, and has been con-
sidered by Batson (1991a) as a chief enabling process
to altruism. Evolutionary biologists such as Hamilton
(1964) and Wilson (1988) suggested that empathic help-
ing behavior has evolved because of its contribution to
genetic fitness (kin selection). In humans and other
mammals, an impulse to care for offspring is almost cer-
tainly genetically hardwired. It is far less clear that an
impulse to care for siblings, more remote kin, and simi-
lar nonkin is genetically hardwired (Batson, in press).
The emergence of altruism, of empathizing with and car-
ing for those who are not kin, is thus not easily explained
within the framework of neo-Darwinian theories of natu-
ral selection (Eisler & Levine, 2002). Social learning
explanations of kinship patterns in human helping
behavior are highly plausible. However, one of the most
striking aspects of human empathy is that it can be felt
for virtually any target—even targets of a different spe-
cies. We favor the view championed by Cummins and
Cummins (1999) that a viable evolutionary cognitive psy-
chology requires neither extreme nativism nor modular-
ity. There are, however, evolved biological predisposi-
tions (e.g., the capacity to distinguish agents from other
objects and to engage in reciprocal interactions with the
former but not the latter) that are necessary for the full
maturation of empathy. But without social interaction
and emotional bonds with others, it is unlikely that em-
It seems evident from the descriptions of compara-
tive psychologists and ethologists that some behaviors
homologous to empathy can be found in animals (Buck
& Ginsburg, 1997; Plutchik, 1987). For de Waal (1996),
empathy is not an all-or-nothing phenomenon, and
many forms of empathy exist intermediate between the
extremes of mere agitation at the distress of another
and full understanding of their predicament. However,
many other comparative psychologists view empathy as a
kind of induction process by which emotions, both posi-
tive and negative, are shared and by which the probabili-
ties of similar behavior are increased in the participants.
In our view, this is not a sufficient mechanism to account
for human empathy. Feelings may be shared, but
humans are able to intentionally “feel for” and act on
behalf of other people whose experiences differ greatly
from their own (Batson, 1991a, in press; Decety &
We believe that self-other awareness and self-regulation
of emotions are vital components of human empathy
(see sections titled Self-Other Awareness and Mental
Flexibility and Self-Regulation). These components may
well steer us toward a clear distinction between humans
and other mammals when referring to empathy. In addi-
tion, as emphasized by Harris (2000), humans, unlike
other primates, can put their emotions into words, allow-
ing them not only to express emotion but to report on
current, as well as past, emotions. These reports provide
an opportunity to share and explain emotional experi-
ence with others that is not found in other species. Con-
versation helps to develop empathy, for it is often here
that one learns of shared experiences and feelings.
Moreover, this self-reflexive capability (including emo-
72 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
tion reappraisal) may be a crucial difference between
humans and other animals (Povinelli, 2001).
Batson (1991a) has put forward an attractive empathy-
altruism hypothesis. This hypothesis claims that the pro-
social motivation evoked by empathy is directed toward
the ultimate goal of increasing the welfare of the person
in need. This hypothesis seems necessary to explain why
some people hold helping intentions that are not
explained by egoistic motivations, such as relief of per-
sonal distress, the relief of sadness, and the desire to
make oneself happy (Batson, 1991b). An egoistic expla-
nation of the empathy-altruism hypothesis was proposed
by Smith, Keating, and Stotland (1989). They suggested
that empathically aroused individuals help to gain the
good feeling of sharing vicariously in the needy person’s
joy at improvement (or, in other words, the empathically
concerned witness to the distress of others helps to be
happy). However, other authors have suggested that
rather than empathy, it is the sense of self-other overlap
between the helper and the person in need that moti-
Neuberg, 1997). Helping others with whom one feels
some level of commonality would not be selfless because
it leads to a more favorable mental state. Thus, as demon -
strated by recent research from Kruger (2003), psycho-
logically altruistic and egoistic pathways seem to operate
simultaneously in empathic concern.
Even though empathy provides obvious benefits at
both the individual and societal level by allowing people
to coordinate their behavior and care for the other, it
also has its cost in terms of maintaining an expanded
self—that is, a self that is linked to others (Hodges &
Klein, 2001). One example of such cost is the tendency
to assume that others will feel the same way the self does,
which is referred to as the false consensus effects (Marks
& Miller, 1987). Another example is the anxiety that can
result from watching an unpleasant situation happening
to another person. It has also been argued that some
aspects of psychopathology may be in part regarded as
the evolutionary cost of humankind for the develop-
ment of our advanced capacity to empathize (Brüne,
2001). How this cost-benefit equation is solved within
each individual depends on regulatory mechanisms as
well as several personality and situational characteristics.
EMPATHY AND ITS
It is not an easy task to bridge our intuitive folk con-
ception of empathy with the explanations offered by
social psychology, developmental science, and neurosci-
ence (the three domains we want to articulate). The real
challenge, as insightfully expressed by Adolphs (2003),
is that research in social cognition faces both theoretical
and methodological problems. For instance, how do we
unite the vocabularies of social psychology and cognitive
psychology, the latter being easier to link with neurosci-
ence data? Is social cognition domain specific or domain
general? How are social-cognitive processes related to
non-social-cognitive processes? These difficulties hold
particularly true for the concept of empathy, for it is a
psychological construct, which accounts for a super-
ordinate category of behaviors. This construct has been
the main focus so far of extensive research in social and
developmental psychology (see, e.g., Batson, 1987;
Eisenberg & Strayer,1987; Ickes, 1997) and much less in
cognitive neuroscience. Moreover, we think it is useless
to seek “the” neural substrate of empathy in the same way
that it appears futile to identify one specific empathy def-
icit in neurological or psychiatric patients. Rather, we
suggest that a more successful approach would come
from breaking down this concept into its constitutive
components or processes and examining their respec-
tive neural instantiations. Thus, a clear definition of em-
pathy is needed (for a history of the concept of empathy,
see Wispé, 1987).
Empathy is a complex form of psychological infer-
ence in which observation, memory, knowledge, and
reasoning are combined to yield insights into the
thoughts and feelings of others (Ickes, 1997). As such,
empathy involves not only some minimal recognition
and understanding of another’s emotional state (or
most likely emotional state) but also the affective experi-
ence of the other person’s actual or inferred emotional
state. There are many other definitions of empathy,
almost as many as there are researchers in this field. For
many psychologists, empathy implies at least three differ-
ent processes: feeling what another person is feeling,
knowing what another person is feeling, and having the
intention to respond compassionately to another per-
son’s distress. But regardless of the particular terminol-
ogy that is used, there is broad agreement on three pri-
mary components: (a) an affective response to another
person, which often, but not always, entails sharing that
person’s emotional state; (b) a cognitive capacity to take
the perspective of the other person; and (c) some regula-
tory mechanisms that keep track of the origins of self-
and other-feelings (e.g., Batson, 1991a, 1991b; Davis,
1996; Decety, 2002b; Decety & Hodges, 2004; Eisenberg,
2000; Hodges & Wegner, 1997; Ickes, 2003). Thus, empa-
thy requires both the ability to share the emotional expe-
rience of the other person (affective component) and an
understanding of the other person’s experience (cogni-
tive component). Some scholars favor in their definition
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 73
one aspect or the other. For instance, Hoffman (1981)
views empathy as a largely involuntary vicarious response
to affective cues from another person or her situation,
whereas Davis (1996) or Batson (1991b, in press) stress
the conscious role-taking ability, which taps mainly into
We favor Ickes’s definition presented above because it
best captures the multidimensional nature of empathy
and makes explicit reference to some minimal mental-
izing capacity. This latter concept refers to the broad
social-cognitive ability used by humans to explain and
predict their own behavior and that of others by attribut-
ing to them independent mental states, such as belief,
desires, emotions, or intentions (Gallagher & Frith,
2003). This mentalizing ability is considered to set us
apart from other primates (Povinelli, Bering, &
Giambrone, 2000), with perhaps the exception of apes.
Whether or not the concepts of empathy and mental-
izing overlap remains an unsolved theoretical issue.
Although emotions and feelings are generally included
in the definition of mentalizing (also dubbed “theory of
mind”), it is often considered that the recognition of the
emotional state of others is a sort of direct, automatic
process that does not require psychological inference
and metarepresentation. Although this may be true for
the recognition of basic emotions, more complex ones,
such as self-conscious emotions, are likely to require cog-
nitive processing. In addition, the extent to which attri-
bution of desires, intentions, and emotions rely on dis-
tinct or common functional or anatomical substrates
remains open. The neurophysiological evidence is
scarce. For instance, most neuroimaging studies on
these types of processes have used mentalizing tasks that
did not include any affective components.
Of all the sources from which one can draw insight as
to the constituents of human empathy, psychotherapeu-
tic schools provide the most interesting, experience-
related knowledge. Indeed, empathy is appreciated to
play a central role in psychotherapies as almost all psy-
chotherapy involves intersubjective communication
between at least two individuals for the clinician to
understand the client sufficiently to proceed along a
treatment path (Bohart & Greenberg, 1997). For in-
stance, Freud (1921) wrote that empathy was indispens-
able when it came to taking a position regarding another
person’s mental life and considered it as the process that
plays the largest part in our understanding of what is
inherently foreign to our ego in other people. In Jokes
and Their Relation to the Unconscious (Freud, 1905), Freud
used the concept of empathy (influenced by the work of
Lipps, which he profoundly admired) to designate the
process of putting oneself into another’s position, either
consciously or unconsciously. A number of analysts have
pointed out that empathy involves resonating with the
other’s unconscious affect and experiencing the experi-
ence with this person while the empathizer maintains
the integrity of his self intact (see Basch, 1983). Accord-
ing to Beres and Arlow (1974), a therapist can empathize
with how the patient would feel if and when he or she
could become consciously aware of the unconscious
wishes, conflicts, fantasies, and other mental contents
that are being warded off.
The psychoanalyst Theodor Reik (1949) offered a
definition of the processes involved in empathy that is
especially relevant to our view. He described the four fol-
•Identification: focusing one’s own attention to another
and allowing oneself to become absorbed in contempla-
tion of that person.
•Incorporation: making the other’s experience one’s own
via internalizing the other.
•Reverberation: experiencing the other’s experience while
attending to one’s own cognitive and affective associa-
tions to that experience.
•Detachment: moving back from the merged inner rela-
tionship to a position of separate identity, which permits
a response to be made that reflects both understanding
of others as well as separateness from them.
More than to anyone else, the concept of empathy was
an important part of the counseling technique devel-
oped by Carl Rogers. For him, empathy was one of the
central conditions for therapeutic change. The therapist
experiences an empathic understanding of the client’s
internal frame of reference and endeavors to communi-
cate this experience to the client. By empathy, Rogers
(1959) meant “to perceive the internal frame of refer-
ence of another person with accuracy and with the emo-
tional components and meanings which pertain thereto
as if one were the person, but without losing the as if con-
dition.” This last component reflects an active and con-
trolled mechanism during which the person remains
aware of the merging between the self and the other.
The model we propose here, as a heuristic frame-
work, considers that the basic mechanism for empathy
rests on the innate ability to recognize that the self and
the other can be the same but also can be teased apart.
Moreover, empathic understanding requires a minimal
mental flexibility for the subjective viewpoint of the
other to be adopted. Our model is strongly influenced by
theories of psychotherapy and is compatible with both
the humanistic and psychodynamic theories, as well as
the behavioral approach. The former views empathy as
an innate ability to experience the inner life of another
while retaining objectivity, whereas the latter views
empathy as a communication skill (Carlozzi, Bull, Stein,
Ray, & Barnes, 2002).
74 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
We propose that three major functional components
dynamically interact to produce the experience of empa-
thy in humans:
•affective sharing between the self and the other, based
on perception-action coupling that lead to shared repre-
•self-other awareness. Even when there is some tempo-
rary identification, there is no confusion between self
•mental flexibility to adopt the subjective perspective of
the other and also regulatory processes.
In our view, empathy entails these components, and
none of them can account solely for the potential of
human empathy. The three are intertwined and must
interact with one another to produce the subjective
experience of empathy. For instance, sharing emotion
without self-awareness corresponds to the phenomenon
of emotional contagion, which takes the form of “total
identification without discrimination between one’s
feelings and those of the other” (de Waal, 1996). Models
or theories based only on the affect-sharing component
of empathy may well reflect the continuity observed
across species (in particular, emotion communication,
see Preston & de Waal, 2002). We argue that there are
unique features of empathy (e.g., perspective-taking,
self-awareness, and emotion reappraisal) emerging in
the course of evolution that distinguish human and non-
human species. This model of empathy combines both
representational aspects (i.e., memories that are local-
ized in distributed neural networks that encode informa-
tion and, when temporarily activated, enable access to
this stored information) and processes (i.e., computa-
tional procedures that are localized and are indepen-
dent of the nature or modality of the stimulus that is
being processed). Like many emotion-related processes,
some components involved in empathy occur implicitly
and sometimes without awareness. This is the case of the
emotion-sharing aspect. Other components require
explicit processing, such as perspective taking, repre-
senting our own thoughts and feelings as well as those of
others, and also some aspects of emotion regulation. It is
unfortunately beyond the scope of this article to review
the current knowledge of the neuroscience of emotion
(see Davidson, Pizzagalli, Nitschke, & Kalin, 2003).
Each of these macrocomponents of empathy can fur-
ther be fragmented into their constitutive elements as
well as associated with their neural implementation. In
the rest of the article, we address each of the macro-
components separately; then, we review the evidence
from various neurological disorders that can lead to a
lack of empathy and discuss how these deficits enlighten
our understanding of this process. Finally, we raise some
questions for future research that we think can be best
tackled by a multidisciplinary approach.
BETWEEN SELF AND OTHERS
Perception and Action Coupling
Investigations of the brain substrates involved in the
perception of actions are directly relevant to the explora-
tion of the mechanisms subserving empathy because
bodily expressions constitute an external, perceivable
indication of people’s intentions and emotions. The
notion of shared representations between self and other
(Decety & Sommerville, 2003; Jeannerod, 1999) is at the
core of our theoretical framework. This notion reflects
the idea that the perception of a given behavior in
another individual automatically activates one’s own
representations of that behavior (Knoblich & Flach,
2003; Preston & de Waal, 2002; Prinz, 1997). Such a view
is grounded in the fundamental physiological properties
of the nervous system regarding the continuity between
action and cognition, which is primarily based on per-
ception/action cycles. These processes are functionally
intertwined—that is, perception is a means to action,
and action is a means to perception, and they operate
right after birth. Indeed, the vertebrate brain has
evolved for the purpose of governing motor activity by
transforming sensory patterns into patterns of motor
coordination (Sperry, 1952). Spontaneous neural activ-
ity in developing networks of the vertebrate nervous sys-
tem may well be at the very origin of these cycles. Re-
search over the past 10 years has established that
spontaneous activity (in many cortical and subcortical
areas) is a characteristic feature of the embryonic ner-
vous system (O’Donovan, 1999). Observations of
preterm-born infants and fetuses show that sensory and
motor activities are linked and that such linkages are
prewired (Bloch, 1997). Thus, the human infant, like
other young mammals, is born with instruments that can
ensure relations with the external world. The newborn
spontaneous motor activity provides the necessary and
sufficient conditions for the natural interactions with
others (Decety, 2002a).
Gibson (1966) proposed the metaphor of “afford-
ance” to account for the direct link between perception
and action. Affordances are properties of objects or
events in the surroundings that respond to the needs of
the perceiver. They are both physical and psychological,
as well as ecological (see McArthur & Baron, 1983, for an
ecological theory of social perception). Later, Shepard
(1984) argued that as a result of biological evolution
and individual learning, the organism is, at any given
moment, tuned to resonate to the incoming patterns
that correspond to the invariants that are significant for
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 75
it. These patterns, according to Shepard, have become
most deeply internalized (i.e., represented), and even in
the complete absence of external information, the sys-
tem can be excited entirely from within (while imagin-
ing, for example). Thus, unlike Gibson, Shepard makes
explicit reference to internal representation and, in our
opinion, makes it possible to articulate the notion of res-
onance with that of shared representations. In addition,
humans actively seek information about themselves and
others. This aspect is compatible with contemporary the-
ory of motor representation, which stresses the auton-
omy of the individual with respect to the external milieu
and views his or her actions as a consequence of trig-
gering by the environment or as a consequence of an
internal process (Jeannerod, 1994).
The automatic mapping between self and other is sup-
ported by considerable empirical literature in the
domain of perception and action, which has been mar-
shaled under the common-coding theory (Prinz, 1997).
This theory claims parity between perception and
action. Its core assumption is that actions are coded in
terms of the perceivable effects (i.e., the distal percep-
tual events) they should generate (Hommel, Müsseler,
Aschersleben, & Prinz, 2001). This theory also states that
perception of an action should activate action repre-
sentations to the degree that the perceived and the
represented action are similar (Knoblich & Flach, 2003).
As such, these representations may be shared between
individuals. Indeed, the meaning of a given object,
action, or social situation may be common to several peo-
ple and activate corresponding distributed patterns of
neural activation in their respective brains (Decety &
There is both behavioral and neurophysiological evi-
dence for shared representations between perception
and action (see Viviani, 2002, for a review of psycho-
physics data demonstrating that the perception of bio-
logical actions is constrained by the observer’s implicit
knowledge of the movements that he or she is capable to
produce). Imitation in neonates is the best evidence for
this perception-action coupling functioning right from
birth (Meltzoff & Decety, 2003, for a recent review).
However, this early imitation cannot be explained solely
by a simple motor resonance behavior mechanism—that
is, a neural activity that is spontaneously generated dur-
ing the perception of movements, gestures, and actions
made by another person or a general arousal reaction.
The work of Meltzoff and Moore (1994, 1997) and oth-
ers (e.g., Kugiumutzakis, 1999; Nadel & Baudonnière,
1982) show that imitation is representationally mediated
because the infant’s response need not be temporally
coupled to the stimulus and is not compulsory (see Table
1.1 in Meltzoff, 2002, for a complete list of the character-
istics of early imitation). In addition, infants are not only
attracted to people, but they also identify with them
(Hobson, 2002; Trevarthen, 1979). By 2 months, they
imitate human actions but not those of objects because
they implicitly understand other people to be like them
(Legerstee, 1991). Taken together, these findings indi-
cate that imitation is a social response.
This work with newborns has also led a number of
developmental psychologists to propose that the under-
standing of the other person is primarily a form of em-
bodied practice (e.g., Hobson, 1989; Legerstee, 1991;
Meltzoff, 1990; Rogers & Pennington, 1991). Humans
develop and maintain their self-concept through the
process of taking action and then reflecting on what they
have done—that is, the sensory consequences of their
actions—and, later in life, what others tell about what
they have done (Gallagher & Meltzoff, 1996). Meltzoff
and Gopnik (1993) argued that the understanding of
the other person emerges in part from being “like them”
in action, through imitation, and that this provides the
basic mechanism for empathy. However, the recognition
of self-other equivalences would be the starting point for
social cognition, not its culmination (Meltzoff, 2002).
Social psychologists have shown that humans mimic
unintentionally and unconsciously a wide range of
behaviors, such as accents, tone of voice, rate of speech,
posture and mannerisms, as well as moods (e.g.,
Chartrand & Bargh, 1999; Dijksterhuis & Bargh, 2001).
One of the adaptive advantages of such behaviors is that
they bind people together and foster empathy, liking,
and smooth interaction. In an interesting series of re-
cent experiments, Van Baaren, Holland, Kawakami, and
Van Knippenberg (2004) demonstrated that partici-
pants who had been mimicked by the experimenter
were more helpful and generous toward other people
than nonmimicked participants. They also found that
these beneficial consequences of mimicry were not
restricted to behavior directed toward the mimicker but
included behavior directed toward people not directly
involved in the mimicry situation. The authors con-
cluded that the effects of mimicry are not simply the
result of an increased liking for the mimicker but are
due to an increased prosocial orientation in general.
In neuroscience, evidence for this perception/action
coupling ranges from electrophysiological recordings in
monkeys, in which mirror neurons that fire both during
goal-directed actions and observation of actions per-
formed by another individual (Rizzolatti, Fogassi, &
Gallese, 2001, for a review), to functional neuroimaging
experiments in humans, which demonstrate that the
neural circuit involved in action execution overlaps with
that activated when actions are observed (Blakemore &
Decety, 2001, for a review). This neural network includes
the premotor cortex, the parietal lobule, the supple-
mentary motor area, and the cerebellum (see Grèzes &
76 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
Decety, 2001, for a meta-analysis). In addition, several
neuroimaging studies have shown that similar brain
areas, pertaining to the same network in the premotor
and posterior parietal cortex, are activated during imag-
ining one’s own action (e.g., Decety et al., 1994; Hari et
al., 1998), imagining another’s action (Ruby & Decety,
2001), and imitating actions performed by a model
(Decety, Chaminade, Grèzes, & Meltzoff, 2002; Decety
et al., 1997; Iacoboni et al. 1999). Another strong evi-
dence for the involvement of motor representation dur-
ing observation comes from measurements of the corti-
cospinal excitability by means of transcranial magnetic
stimulation. One study found specific modulation of
motor-evoked potentials (MEP) in participants asked to
observed simple hand gestures (Fadiga, Fogassi, Pavesi,
& Rizzolatti, 1995). In another recent study, partici-
pants were asked to observe, imaging, or imitate hand
actions while magnetic stimulation was delivered (Clark,
Tremblay, & St.-Marie, 2003). Imitation produced the
greatest MEP followed by the observation and imagery
conditions (see Figure 1).
Such a shared motor-representations mechanism
offers an interesting foundation for intersubjectivity
because it provides a functional bridge between first-
person information and third-person information
(Decety & Sommerville, 2003). But as suggested by our
model, this mechanism is necessary but not sufficient for
empathic understanding. Moreover, it should be noted
that the overlap between cortical areas involved in self-
related actions and other-related actions is not com-
plete. There are specific subcircuits within the premotor,
prefrontal, and parietal cortices that account for either
the self or the other (e.g., Ruby & Decety, 2001). A recent
functional magnetic resonance imaging (fMRI) experi-
ment by Ramnani and Miall (2004) has shown that the
motor system is engaged when participants use arbitrary
visual cues to prepare their own actions and also when
they use the same cues to predict the actions of other
people. However, these two tasks activate separate sub-
circuits within the premotor cortex.
Emotional expression and perception are an integral
part of human interactions (Schulkin, 2004). At one
level, emotional expressions are governed by rules and
can be elicited by simple stimuli, as in the example of dis-
gust in the presence of bitter taste. However, humans
and other animals also use bodily expressions to commu-
nicate various type of information to members of their
own species. Understanding other people’s emotional
signals has clear adaptive advantages and is especially
important in the formation and maintenance of social
The phenomenon of emotional contagion, defined
as the tendency to automatically mimic and synchronize
facial expressions, vocalizations, postures, and move-
ments with those of another person and, consequently,
to converge emotionally with the other (Hatfield,
Cacioppo, & Rapson, 1994) is certainly the most simple
expression of emotion sharing that does not need con-
scious awareness. There are a variety of ways in which
happy people add color and life to our world and sullen
people take it away (Chartrand, Maddux, & Lakin, in
The affective component of empathy may be concep-
tualized in its most rudimentary form as the ability to
detect the immediate affective state of another person
(Trevarthen & Aitken, 2001). This emotional arousal
stems from the apprehension or comprehension of an-
other’s affective state. Developmental research indicates
that we are, from birth, not only acting and thinking
selves, but we also express an intuitive need to relate our-
selves to other people. It has been shown that very young
infants express what Trevarthen (1979) terms intersubjec-
tive sympathy—that is, they are innately predisposed to be
sensitive and responsive to the subjective states of other
people. This can be demonstrated through several
means, including spontaneous face-to-face interaction
between infants and their mothers and through more
specialized “still-face procedures” (i.e., when mothers
adopt a neutral face and stop responding to the infant),
which can lead to withdrawal by the infant. Research per-
formed by Field, Woodson, Greenberg, and Cohen
(1982) has shown that neonates (aged 36 hours) can dis-
criminate three facial expressions (happy, sad, and sur-
prised) posed by a live model. This affective communica-
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 77
Figure 1: Comparison ofthe mean changes in MEP log-amplitude in a
group of participants during various experimental condi-
tions including passive observation, observation to imitate,
mental imagery, and imitation of hand actions.
SOURCE: Clark, Tremblay, St.-Marie, 2003 (reprinted with permission).
NOTE: MEP = motor-evoked potentials.
tion bridges the gap between an infant’s grasp of other
people’s behavior and other people’s experiences.
Through the direct perception of feelings in the bodily
expressiveness of others and through progressively more
differentiated experiences of affective commonality
with other people, an infant comes to understand some-
thing of what it means to be a person (Hobson, 1989;
The infant affective state may be similar to or con-
gruent with what the other person is feeling (Eisenberg
& Strayer, 1987). For example, Simner (1971) investi-
gated the phenomenon of neonatal crying in reaction to
the distress cries of other newborns. He found that the
sound of neonatal crying produced significantly more
reactive crying in the newborn than did either white
noise, a cry from a 5-month old, or a synthetic cry. These
findings were later replicated by several groups (e.g.,
Sagi & Hoffman, 1976). Of special interest is the study by
Martin and Clark (1987), who tested 1-day-old babies
with, audiotapes of neonatal crying, the crying of an 11-
month-old, and the newborn’s own crying. Not only did
they replicate Simner’s results, but they also showed that
newborns did not respond to the sound of their own
cries. These latter results suggest that there is some self-
other distinction already functioning right from birth.
Thus, it is clear that from very early on in devel-
opment, infants are capable of emotional resonance,
which is one important precursor of empathy (Hoffman,
2000). Another important foundation of the develop-
ment of empathy is the affective synchrony in mother-
infant play, which begins to occur regularly around 2 to 3
months of age. These play episodes, which each partner
contributes through a repertoire of interactive behav-
iors, entail a sharing of affect between mother and baby
(Stern, 1985; Trevarthen, 1979). Microanalyses of such
social interactions conducted by Malatesta and Haviland
(1982) showed that mothers were highly likely to imitate
infant expressions of enjoyment and interest (which
occurred most frequently), as well as expressions of sur-
prise, sadness, and anger when they occurred. However,
mothers rarely displayed negative emotions to their
baby. Thus, infant-mother dyads exhibit considerable
positive synchrony, partly as a consequence of the moth-
er’s contingent matching of positive infant emotional
expressions. Field, Healy, Goldstein, and Guthertz
(1990) have convincingly documented how depressed
mothers can influence the subjective state of their babies
through these interactions.
Another compelling evidence of infants’ ability to
discriminate among different facial expressions of emo-
tion and to interpret them as emotional communication
is known as “social referencing” (Campos & Stenberg,
1981). This process, which starts at about 10 months of
age, reflects an active effort by infants to obtain emotion
cues from others to assist in their own assessment of an
uncertain or ambiguous situation (Rosen, Adamson, &
Bakeman, 1992). However, it is difficult to interpret
whether the child’s referencing is genuinely empathic as
the purpose of social referencing is to assess one’s own
circumstances rather than another’s (Thompson, 1987).
This process is best conceptualized as an integral, flexi-
ble aspect of the social construction of an infant’s reality
It is around the 2nd year that empathy may be mani-
fested in prosocial behaviors (e.g., helping, sharing, or
comforting) indicative of concern for others. Studies of
children in the 2nd year of life indicate that they have the
requisite cognitive, affective, and behavioral capacities
to display integrated patterns of concern for others in
distress (Bretherton, Fritz, Zahn-Waxler, & Ridgeway,
1986). During this period of development, children
increasingly experience emotional concern “on behalf
of the victim,” comprehend others’ difficulties, and act
constructively by providing comfort and help (Zahn-
Waxler, Radke-Yarrow, Wagner, & Chapman, 1992).
The shared-representations mechanism may also ac-
count (at least partly) to emotion processing (Adolphs,
Damasio, Tranel, Cooper, & Damasio, 2000). In this
model, perception of emotion activates the neural
mechanisms that are responsible for the generation of
emotions (Adolphs, 2002). Such a system prompts the
observer to resonate with the state of another individual,
with the observer activating the motor representations
and associated autonomic and somatic responses that
stem from the observed target—that is, a sort of inverse
mapping. For example, while watching someone smile,
the observer activates the same facial muscles involved in
producing a smile at a subthreshold level, and this would
create the corresponding feeling of happiness in the
observer. There is evidence for this mechanism in the
recognition of emotion from facial expression. For in-
stance, viewing facial expressions triggers expressions on
one’s own face, even in the absence of conscious recogni-
tion of the stimulus (Dimberg, Thunberg, & Elmehed,
2000; Wallbott, 1991).
Making a facial expression generates changes in the
autonomic nervous system and is associated with feeling
the corresponding emotion (Ekman, Levenson, &
Friesen, 1983). In a series of experiments, Levenson,
Ekman, and Friesen (1990) instructed participants to
produce facial configurations for anger, disgust, fear,
happiness, sadness, and surprise while heart rate, skin
conductance, finger temperature, and somatic activity
were monitored. They found that such a voluntary facial
activity produced significant levels of subjective experi-
ence of the associated emotions as well as specific and
reliable autonomic measures. In another study, Ekman
and Davidson (1993) were able to demonstrate similar
78 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
patterns of electroencephalographic activity for sponta-
neous and voluntary forms of smiling. Recently, an fMRI
experiment confirmed these results by showing that
when participants are required to observe or to imitate
facial expressions of various emotions, increased neuro-
dynamic activity is detected in the superior temporal
sulcus, the anterior insula, and the amygdala, as well as
areas of the premotor cortex corresponding to the facial
representation (Carr, Iacoboni, Dubeau, Mazziotta &
One may thus deliberately experience one’s own
emotion or use this mechanism to experience the emo-
tion of the other, something that Edgar Allan Poe knew.
In The Purloined Letter (1845/1990), he wrote:
When I wish to find out how wise, or how stupid, or how
good, or how wicked is any one, or what are his thoughts
at the moment, I fashion the expression of my face, as
accurately as possible, in accordance with the expression
of his, and then wait to see what thoughts or sentiments
arise in my mind or heart, as if to match or correspond
with the expression.
Converging neurophysiological arguments in favor
of this model are also supported by the finding of
paired deficits between emotion production and emo-
tion recognition. A lesion study carried out with a large
number of neurological patients by Adolphs and col-
leagues (2000) found that damage within the right
somatosensory-related cortices (including primary and
secondary somatosensory cortices, insula, and anterior
supramarginal gyrus) impaired the judgment of other
people’s emotional states from viewing their face. The
same authors also reported that there is an association
between the impaired somatic sensation of one’s own
body and the impaired ability to judge other people’s
emotions. It has been reported that patients with Parkin-
son’s disease may be impaired in expressing emotional
faces and perceiving emotional facial affect (Jacobs,
Shuren, Bowers, & Heilman, 1995). A study of brain-
damaged individuals found that recognizing emotions
from prosody draws on the right fronto-parietal cortex
(Adolphs, Damasio, & Tranel, 2002). The authors stated
that their results are consistent with the hypothesis that
the recognition of emotion in others requires the per-
ceiver to reconstruct images of somatic and motoric
components that would normally be associated with pro-
ducing and experiencing the emotion signaled in the
Moreover, there are several dramatic case studies that
support the idea that the same neural systems are in-
volved both in the recognition and in the expression of
specific emotion. Adolphs, Tranel, Damasio, and
Damasio (1995) investigated S.M., a 30-year-old patient,
whose amygdala was bilaterally destructed by a meta-
bolic disorder. Consistent with the prominent role of the
amygdala in mediating certain negatively valenced emo-
tions, such as fear, S.M. was found to be impaired at both
the recognition of fear from facial expressions as well as
in the phenomenological experience of fear. Another
case, N.M, who suffered from bilateral amygdala damage
and left thalamic lesion was found to be impaired when
it came to recognizing fear from facial expressions
and exhibited an equivalent deficit affecting fear recog-
nition from body postures and emotional sounds
(Sprengelmeyer et al., 1999). The patient reported
reduced anger and fear in his everyday experience of
emotion as well.
There is also evidence for paired deficits for the emo-
tion of disgust. Calder, Keane, Manes, Antoun, and
Young (2000) described patient N.K., with left insula and
putamen damage, who was selectively impaired at recog-
nizing social signals of disgust from multiple modalities
(facial expressions, nonverbal sounds, and emotional
prosody) and who was less disgusted than controls by
disgust-provoking scenarios. Further and direct support
for a specific role of the left insula in both the recogni-
tion and the experience of disgust was recently provided
by a neuroimaging study in which participants inhaled
odorants producing a strong feeling of disgust and, in
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 79
Figure 2: Overlap (white) between the brain activation elicited by the
visual observation (blue) and the feeling (red) of disgust in a
group of healthy volunteers.
SOURCE: Wicker, Keyser, et al., 2003 (reprinted with permission).
another condition, watched video clips showing the
facial expression of disgust (Wicker, Keysers, et al.,
2003). It was found that observing such facial expres-
sions and feelings of disgust activated the same sites in
the anterior insula and anterior cingulate cortex (see
By virtue of its aversiveness, pain serves to promote
the organism’s health and integrity (Williams, 2002).
The expression of pain provides a crucial signal, which
can motivate helping behaviors in others. One does not
feel the sensory aspects of another’s pain, but one may
understand her distress. Interestingly, a single-neuron
recording study in neurological patients has shown that
there are pain-related neurons in the anterior cingulate
cortex (ACC) that respond both to actual stimulation
(thermal stimuli) and also to the observation of the same
stimuli delivered to another individual (Hutchison,
Davis, & Lozano, 1999). Jackson, Meltzoff, and Decety
(2004) recently conducted an fMRI study to identify the
whole neural network engaged in the perception of pain
in other individuals. The participants were shown still
photographs depicting right hands and feet in painful or
neutral everyday-life situations and were asked to imag-
ine the level of pain that these situations would produce.
Significant activation in regions involved in the affective
aspects of the pain-processing network—notably, the
anterior cingulate cortex and the anterior insula—was
detected, but there was no activity in the somatosensory
cortex. Moreover, the level of activity within the anterior
cingulate cortex was strongly correlated with subjects’
mean ratings of pain attributed to the different situa-
tions. Another recent fMRI study demonstrated that the
anterior cingulate cortex, the anterior insula, cere-
bellum, and brainstem were activated when subjects
experienced a painful stimulus, as well as when they
observed another person receiving a similar stimulus
(Singer et al., 2004). Again, no activity was detected in
the somatosensory cortex during the observation of pain
in others. Together, these findings support the idea that
part of the neural network mediating pain experience is
shared when empathizing with pain in others.
A positron emission tomography study investigated
the neural response to externally (by watching emo-
tional laden film clips) versus internally (by autobio-
graphical scripts) generated emotions (Reiman et al.,
1997). Both film-generated emotion and recall-
generated emotion were associated with symmetrical
increases in the medial prefrontal cortex and thalamus.
The former condition also resulted in activation of the
hypothalamus, the amygdala, the anterior temporal cor-
tex, and the occipito-tempo-parietal junction, whereas
the latter condition was specifically associated with acti-
vation in the anterior insula and orbitofrontal cortex.
There is an overlap between externally and internally
produced emotions, but this overlap is partial. It should
be noted that the films and scripts included three emo-
tions (happiness, sadness, and disgust), which were not
Another recent neuroimaging study has demon-
strated the involvement of shared representations (in
both emotion processing areas, and fronto-parietal net-
works) when subjects feel sympathy for another individ-
ual (Decety & Chaminade, 2003a). In this study, partici-
pants were presented with a series of video clips showing
individuals telling sad and neutral stories, as if they had
personally experienced them. These stories were told
with either congruent or incongruent motor expression
of emotion. At the end of each movie, subjects were
asked to rate the mood of the actor and also how likable
they found that person. Watching sad stories versus neu-
tral stories was associated with increased activity in emo-
tion processing–related structures (including the amyg-
dala and parieto-frontal areas) predominantly in the
right hemisphere. This network was not activated when
subjects watched incongruent social behavior. Indeed,
the condition of mismatch between the narrative con-
tent of the stories and the motor expression of emotion
elicited strong hemodynamic increase in the ventrome-
dial prefrontal cortex and superior frontal gyrus, which
are involved in monitoring conflict between expected
and actual outcomes (e.g., Fink et al., 1999).
Altogether, shared representations between self and
other at the cortical level have been found for action
understanding, pain processing, and emotion recogni-
tion. This mechanism provides the neurophysiological
basis for the operation of social cognition by means of
the automatic activation of motor representations or
emotions. There is no specific cortical site for shared
representations; their neural underpinnings are widely
distributed, and the pattern of activation (and also pre-
sumably deactivation) varies according to the process-
ing domain, the particular emotion, and the stored
The shared-representations mechanism lends cre-
dence to the idea that the same representational form is
used in coding embedded intentional relations, whether
it involves the self as an agent or another agent (Gopnik,
1993). This idea is also consistent with the model stating
that human consciousness is formed in the dynamic
interrelation of the self and the other and therefore is
inherently intersubjective in nature (Thompson, 2001).
Knowledge of the self paves the way for achieving an
inferential knowledge of the mental states of others. Yet,
we do understand that self and other are similar but sep-
80 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
arate, and we usually do not confuse first-person knowl-
edge from third-person knowledge.
As suggested earlier, in our model, empathy presup-
poses self-awareness. Individuals who are self-aware, as
evidenced by being able to become the object of their
own attention, experience a sense of psychological conti-
nuity over time and space (Gallup, 1998). Any organisms
capable of self-recognition would have an introspective
awareness of their own mental states and the ability to
ascribe mental states to others (Humphrey, 1990). It is
an adaptive trait that has evolved by natural selection
because it confers some advantage on those individuals
who possess it (Humphrey, 2002; Mandler, 2002). More-
over, the emergence of a self-representation in psycho-
logical development is vital for the empathic process
(Lewis, 1999). We do not think that self-awareness relies
on a specific brain region. Rather, it arises from the in-
teraction between processes distributed in the brain,
especially the prefrontal cortex and the inferior parietal
lobule, and in which the right hemisphere plays a prom-
inent role (Keenan, Gallup, & Falk, 2003).
The roots of the self begin early in infancy. Indeed,
Gibson (1979) suggested that from birth, infants co-
perceive themselves in acting and perceiving their envi-
ronment. Neisser (1991) proposed this implicit self-
knowledge may take two forms: an ecological self,
formed through interactions with physical objects and
bodily perception, and an interpersonal self, formed
through infants’ interactions with others.
Infants’ representations of self- and other-actions are
both overlapping and distinct (Rochat & Striano, 2000).
Infants readily distinguish their own actions from those
of others early on. For instance, Rochat and Hespos
(1997) tested newborn infants within 24 hours of their
birth to measure the frequency of rooting in response to
either external tactile stimulation (the experimenter
stroking the infant’s cheek) or in response to tactile self-
stimulation when the infants spontaneously brought one
of their hands in contact with their cheek. They found
that newborns tended to manifest rooting behavior
three times more often in response to external com-
pared to self-stimulation, suggesting some level of dis-
crimination between these two sources of stimulation.
The awareness of others develops very early on in con-
junction with an awareness of being the object of others’
attention. The developmental psychologist Vasudevi
Reddy suggested that infants are aware of the directed-
ness of others’ attention before evidence of joint atten-
tion. This is particularly well documented by the mea-
sure of a variety of emotional reaction during the first
months (Reddy, 2003). She further argued that the
development in awareness of attention during the first 2
years can be explained in terms of an expanding aware-
ness of the objects of attention. Interestingly, measure-
ments of cerebral metabolism in children (aged
between 18 days to 12 years) indicate a right hemispheric
predominance, which is mainly because of the neural
activity in the posterior associative areas, and the fact
that its functions develop earlier than the left hemi-
sphere (Chiron et al., 1997). This latter finding led us to
speculate that there is a functional relation (anatomi-
cally wired) between the parietal cortex and the implicit
sense of self that infants manifest from birth. In addition,
a number of studies have demonstrated that young
infants are very sensitive to the contingent relationships
between their motor behaviors and consequent stimulus
events and that this capacity serves to distinguish the self
from the external world (Gergely, 2001; Watson, 1972).
Moreover, infants also form shared representations
of their own and others’ actions. Asendorpf and
Baudonnière (1993) have suggested that self-awareness
and other-awareness develop in close synchrony during
the 2nd year because both types of cognition are based
on one common cognitive capacity: the capacity for sec-
ondary representation. Self-awareness requires a capac-
ity for secondary representation because the self as an
object of knowledge is a secondary representation. Simi-
larly, other-awareness requires a capacity for secondary
representation because other-awareness implies taking
the perspective of another person into account. This
framework would account for the dramatic increase in
children’s social-cognitive competence during the 2nd
Indeed, over the first several years of life, children
acquire knowledge of both objective and subjective
aspects of self and others. By 18 to 24 months of age
infants begin to recognize their own mirror image, to
display self-conscious emotions such as embarrassment
or shame (Lewis, Sullivan, Stanger, & Weiss, 1989), to
communicate with peers through the synchronic imita-
tion of each other’s activity (Asendorpf, Warkentin,
Baudonnière, 1996), to cooperate with peers (Brownell
& Carriger, 1990), and to react with empathic behavior
to victims of distress (Zahn-Waxler, Radke-Yarrow, &
During the preschool years, children simultaneously
develop the capacity to represent and report their own
and others’ mental states (e.g., Meltzoff & Gopnik,
1994). This development entails the ability to recognize
when self- and other-perspectives and experiences are
shared and thus congruent as well as under which cir-
cumstances they differ from one another.
Interestingly, the development of self- and other-
mental-state understanding is functionally linked to that
of executive functions (Russell, 1996)—that is, the pro-
cesses that serve to monitor and control thought and
actions, including self-regulation, planning, cognitive
flexibility, response inhibition, and resistance to inter-
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 81
ference (Eslinger, 1996; Shallice, 1988). There is increas-
ingly clear evidence of a specific developmental link
between theory-of-mind development and improved
self-control at around the age of 4 (Perner & Lang,
1999). Carlson and Moses (2001) have convincingly doc-
umented how executive functions, especially inhibitory
control, play a crucial enabling role in both the emer-
gence and expression of children’s mental-state attri-
bution. Furthermore, it has been demonstrated that the
development of cognitive control is related to the matu-
ration of the prefrontal cortex (Tamm, Menon, & Reiss,
2002). Bunge, Dudukovic, Thomason, Vaidya, and
Gabrieli (2002) investigated interference and response
inhibition with fMRI in children (ages 8 to 12) and
adults. They found that children were more susceptible
to interference and less able to inhibit inappropriate
responses than were adults and that different brain
regions were recruited between the two groups. Notably,
the response inhibition in children was not associated
with the right ventrolateral prefrontal cortex as it was in
This is not to suggest that executive functioning can
be equated with theory of mind. There is some evidence
from neurological case studies indicating that theory of
mind is dissociable from executive functioning (Blair &
Cipolotti, 2000; Fine, Lumsden, & Blair, 2001; Lough,
Gregory, & Hodges, 2001). In addition, executive func-
tion is a much broader construct than theory of mind
(see the next section, Mental Flexibility and Self-
In addition, there is evidence that a region around
the paracingulate sulcus in the medial prefrontal cortex
plays a specific role in attribution of intention (for a
recent review, see Gallagher & Frith, 2003). This region
contains spindle cells, a class of large projection neurons
found only in great apes and humans, which are thought
to be involved in coordinating widely distributed neural
activity involving emotion and cognition (Allman,
Hakeem, Erwin, Nimchinsky, & Hof, 2001). This region
has been found to be reliably activated by mentalizing
tasks of various cognitive difficulty, ranging from judging
the emotion in another person’s gaze (Baron-Cohen et
al., 1999; Wicker, Perret, Baron-Cohen, & Decety, 2003),
detecting the intention in simple dynamic animations
(Castelli, Happé, Frith, & Frith, 2000), attributing inten-
tion to cartoons characters (Brunet, Sarfati, Hardy-
Baylé, & Decety, 2000; Gallagher et al., 2000), compre-
hending stories (Fletcher et al., 1995; Vogeley et al.,
2001), detecting social transgression (Berthoz, Armory,
Blair, & Dolan, 2002), and appreciating humor (Goel &
Dolan, 2001). It is also interesting that this region is not
activated by individuals with mentalizing impairments,
such as schizophrenics (Brunet, Sarfati, Hardy-Baylé, &
Decety, 2003) and high-level, functioning autistic
individuals (Happé et al., 1996).
It is thus possible that executive functions are crucial
for the development of theory of mind and that this lat-
ter process is specific and has a dedicated neural under-
pinning. Future research is needed to elucidate the
functional relation between executive functions and
mentalizing and how each can be fragmented into sub-
components with their respective neural implementa-
tion. For instance, Gallagher and Frith (2003) suggested
that the activity in the medial prefrontal cortex occurs
when cues are used to determine an agent’s mental state
that is decoupled from reality and to simultaneously han-
dle these two perspectives on the world.
Neuropsychological research supports a preeminent
role of the right frontal lobe in self-related processing.
For instance, Keenan and his group (Keenan, Nelson,
O’Connor, & Pascual-Leone, 2001) demonstrated that
patients undergoing a Wada test were temporarily de-
sensitized with regard to the recognition of their own
faces when the right hemisphere was anesthetized. This
was not the case when the left hemisphere was anes-
thetized. Right hemisphere damage is also found to
be linked with impairments in autobiographical mem-
ory and self-evaluation. Personal confabulation (akin to
the creation of fictitious stories about the self) appears
to be associated with damage to the right frontal lobe
(Feinburg, 2001). Finally, severe deficits in personal
autobiographical memory retrieval are also associ-
ated with damage to the right ventral prefrontal region
(Levine et al., 1998).
Using fMRI measurements of subjects who were asked
to categorize pictures of their own faces, Keenan,
McCutcheon, and Pascual-Leone (2001) reported selec-
tive activation of the right inferior frontal gyrus on the
border of the medial frontal gyrus. Fink et al. (1996)
found activation of a right hemispheric network of
temporomesial (including the amygdala and hippocam-
pus), posterior cingulate, prefrontal right insula, and
prefrontal regions during presentation of personal auto-
biographical memories versus impersonal statements.
Two functional imaging studies have reported specific
increase in activity in the medial prefrontal cortex and
posterior cingulate during tasks that involved self-
reflection (Gusnard, Akbudak, Shulman, & Raichle,
2001; Johnson et al., 2002).
Based on these numerous studies (and many others
not reviewed here), Keenan et al. (2003) reasonably
argued that the right hemisphere is a key player in self-
awareness and mental-state attribution. Note that their
original definition of consciousness includes awareness
of one’s own thoughts as well as awareness of others’
thoughts. Similar (but not identical) neural processing
for self and other raises the question of how we distin-
82 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
guish between representations activated by the self and
those activated by other (see Jackson & Decety, 2004).
Yet, we have seen in the work reviewed above that the
neural networks underlying self-processing and other-
processing have some common components and some
An influential cognitive-developmental model pro-
poses that individuals represent their own (first-person
knowledge) and others’ activities (third-person knowl-
edge) via a single conceptual system (Barresi & Moore,
1996). These authors also suggest that whenever an
action is taking place, it activates an intentional schema,
a structure internal to every person involved in that
action. The intentional schema has the capacity of coor-
dinating first- and third-person information; according
to the input signals available, the action is attributed to
the self or to the other person. Jeannerod (1999) sug-
gested that conscious agency judgments are not directly
based on explicit knowledge of the stimuli (external
or internal) that have triggered the action or on the sig-
nals generated during its execution. Rather, several
levels of processing are needed so that the level used
for execution is distinct from the level of the conscious
At the neural level, neuroscience research indicates
that the right inferior parietal cortex in conjunction with
prefrontal areas may be critical in distinguishing the self
from the other and therefore navigating shared repre-
sentations (Figure 2). The inferior parietal cortex is a
heteromodal association area, which receives input from
the lateral and posterior thalamus, as well as visual, audi-
tory, somesthetic, and limbic areas. It has reciprocal con-
nections to the prefrontal cortex and to the temporal
lobes (Eidelberg & Galaburda, 1984). These multiple
connections confer on this region a role in the elabo-
ration of an image of the body in space and in time
(Benton & Silvan, 1993) on which the sense of agency
depends (see Gallagher, 2000, for a theoretical account
about the sense of agency). Interestingly, not only the
prefrontal but also the inferior parietal and temporo-
parietal areas have evolved tremendously in humans as
compared to nonhuman primates (Passingham, 1998).
Accumulating empirical evidence indicates that the
parietal cortex plays a major role in the sense of agency—
that is, in distinguishing between self-produced actions
and actions generated by others (Blakemore & Frith,
2003; Jackson & Decety, 2004, for reviews). For instance,
Farrer and Frith (2002) scanned individuals while watch-
ing a moving dot on a computer screen. In some trials,
the participants were in control of the dot’s movements,
whereas in other trials, someone else controlled the dot.
They found increased activity in the right inferior pari-
etal cortex when the dot was controlled by the other, and
increased activity in the anterior insula when the dot was
controlled by the self.
Another neuroimaging study varied the degree of
concordance of the visual feedback provided to the par-
ticipants about their movements with a joystick (Farrer
et al., 2003). Activation of the right inferior parietal lobe
was found to inversely correlate with the subjective sense
of ownership in action execution; the more discordance
between what the subjects did and what they saw, the
more there was increase in this region. Similarly, activa-
tion in the right inferior parietal lobe was found in a
reciprocal imitation paradigm when participants were
aware (and observed) that their actions were being imi-
tated online by another person (Chaminade & Decety,
2002; Decety et al., 2002; see Figure 3). The right inferior
parietal cortex is also involved when subjects mentally
simulate the actions of another person (Ruby & Decety,
2001). There are new findings suggesting that this mech-
anism is also at play during thinking about others. It has
been demonstrated that when subjects are asked to
adopt another person’s perspective to evaluate their
beliefs or imagine their feelings as compared to their
own perspective, the right inferior parietal cortex is
strongly involved (Ruby & Decety, 2003, 2004). Finally,
an fMRI experiment recently demonstrated that the
neurodynamic activity starts earlier in a number of corti-
cal regions involved in motor control when participants
made judgments about their own actions versus those of
others (Grèzes, Frith, & Passingham, 2004). This latter
finding shows that the dynamics of neural activation
within the shared cortical network is an important aspect
to distinguish one’s own actions from the actions of oth-
ers. It can also be conjectured that the latency difference
between the changes in activity elicited by the percep-
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 83
Figure 3: Right inferior parietallobule activation at thejunction of the
temporal cortex superimposed on a rendered MRI from the
Montreal Neurological Institute. In this study, participants
were scanned during a variety of object-directed actions, in-
cluding self-action (A), imitation of actions demonstrated by
an experimenter (B), and observation of their actions being
imitated by the experimenter (C). Note the dramatic in-
crease in this region in this latter condition.
SOURCE: Adapted from Decety et al., 2002.
tion of self versus others’ actions reflects the calibration
process of shared representations. Furthermore, the fact
that the onset of the hemodynamic signal is earlier for
the self than for the others can be considered as a neural
signature of the privileged (and readily) access of self-
Importantly, it has been reported that some patients
with right parietal lesions exhibit a denial of hemiplegia
that extended to the motor deficits of other patients
(Ramachandran & Rogers-Ramachandran, 1996). This
suggests that availability of an efficient body schema is
necessary not only for recognizing one’s own actions but
also for understanding the actions of others. We suggest
that the inferior parietal cortex in conjunction with the
prefrontal cortex plays a pivotal role in the sense of self
by comparing the source of sensory signals. Such a role is
crucial for empathy to maintain a distinction between
the self and the other and keep track of the origin of the
Empathy may be initiated by a variety of situations—
for instance, when one sees another person in distress or
in discomfort, when one imagines someone else’s behav-
ior, by the reading of a narrative in a fiction book, or
when one sees a moving television report. However, in
these conditions, empathy requires one to adopt more
or less consciously the subjective point of view of the
other. A more obvious instance is when a psychothera-
pist adopts the mental world of his client.
Perspective taking is acknowledged as an important
source of human empathy (Batson, 1991a, 1991b;
Batson et al., 2003; Eisenberg, Shea, Carlo, & Knight,
1991). Comparative studies performed with nonhuman
primates and children seem to show that only the latter
are able to adopt the point of view of another individ-
ual (Reaux, Theall, & Povinelli, 1999). Tomasello (1999)
argues that this ability sets us apart from other pri-
mates and is an integral element in intersubjective
An experiment by the social psychologist Ezra
Stotland (1969) illustrates the effect of perspective tak-
ing to generate empathy. In his experiment, the partici-
pants watched someone else whose hand was strapped in
a machine that they were told generated painful heat.
One group of subjects was told just to watch the target
person carefully, another group of subjects was asked to
imagine the way the target was feeling, and one more
group was told to imagine themselves in the target’s
place. Both physiological (i.e., palm sweating and vaso-
constriction) and verbal measures of empathy showed
that the deliberate acts of imagination produced a
greater response than just watching. Batson and his
group conducted a variety a studies that demonstrate the
effectiveness of perspective-taking instructions in induc-
ing empathy. An important aspect of Batson’s theoretical
framework is that empathy-inducing conditions do not
compromise the distinction between the self and other
(e.g., Batson, Sager, et al., 1997, but see Cialdini et al.,
1997, for a different account of empathy and self-other
There is plenty of evidence from various disciplines to
suggest that the mental flexibility to adopt someone
else’s point of view is an effortful and controlled process.
In addition, the ability to take the conceptual perspec-
tive of the other is considered an indispensable element
in the fully developed, mature theory of mind. Develop-
mental research also indicates that perspective-taking
ability develops gradually. In the affective domain, it is
around 18 months that children demonstrate an emerg-
ing awareness of the subjectivity of other people’s emo-
tions. By that age, infants even seem to understand, for
instance, that they should give an experimenter a piece
of food that the experimenter reacts to with apparent
happiness (e.g., broccoli) rather than one toward which
the experimenter acts disgusted (e.g., cookies), even
when they prefer the latter food; in contrast, 14-month-
olds do not show this understanding (Repacholi &
Gopnik, 1997). This finding appears to be the first em-
pirical evidence that infants of this age have at least some
limited ability to reason nonegocentrically about peo-
ple’s desires (Flavell, 1999).
This does not mean that adults reliably (and sponta-
neously) use the ability to adopt the perspective of others
when reasoning about them. Indeed, even adults fre-
quently make a less sharp distinction between what they
know, or believe they know, and what they assume others
do. Realizing that another can have a perspective that
differs from one’s own does not necessarily entail being
able to adopt that perspective. A series of experiments
performed by Keysar, Lin, and Barr (2003) revealed that
adult subjects exhibit a tendency to infer that others
have the same knowledge (and beliefs) as they do, even
when they are aware that the others have a different
point of view.
Several social and developmental psychologists have
suggested, and documented through empirical work,
that our default mode to reasoning about others is bi-
ased toward self-perspective, and this is a general feature
of human cognition. Stated in other words, people are
fundamentally egocentric and have difficulty getting
beyond their own perspective when anticipating what
others are thinking or feeling (Royzman, Cassidy, &
Baron, 2003). For instance, humans have the tendency
to believe that their actions and appearance are more
likely to be noticed, judged, and remembered by others
84 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
than is actually the case (Gilovich, Kruger, & Medvec,
2002). Humans are also inclined to impute their own
knowledge to others and overestimate what they know
(Nickerson, 1999). Recent research indicates that peo-
ple’s predictions of the feelings of others who are in a sit-
uation that arouses drive states (i.e., motivations caused
by bodily needs such as exhaustion, hunger, and thirst)
are based largely on their predictions of how they would
feel in that situation. Van Boven and Loewenstein (2003)
showed that people project their current drive states
(i.e., motivations caused by physiological needs such as
exhaustion, hunger, and thirst) when predicting how
they and how other people would feel in a situation that
arouses drive states (see Figure 4). It has been proposed
that errors in such appraisal are rooted in a lack of
suppression of the self-perspective (Hodges & Wegner,
1997; Vorauer & Ross, 1999).
This egocentric bias is also well documented in chil-
(1994) taught 4- and 5-years-olds a novel fact (e.g., that
cats use their whiskers to determine whether they can fit
in tight places) or told them about a novel process (e.g.,
how to make litmus paper change color on its own). Chil-
dren were then asked whether they thought another
child who was similar to them would know this fact. The
vast majority of the children said that a similar child
would also know that fact. Interestingly, this bias can be
demonstrated with both children and adults and is
found be relatively stable over age (Bernstein, Atance,
Loftus, & Meltzoff, 2004).
We argue that the egocentric bias is compatible with
the simulation theory, which states that we ordinarily
understand and predict the behavior and mental states
of others by simulation—that is, by using our own mental
resources off-line as if we were in the situation of the
other (Goldman, 1989; Gordon, 1986; Harris, 1991,
2000). For Harris (1991), the simulation of the experi-
ence of other persons is not a straightforward process,
and the child often has to judge the simulation with his
or her knowledge of the real situation from a first-person
perspective. Consistent with this idea, Perner and Lopez
(1997) found that children were better at predicting
what another person would see in a particular situation if
they had actually been in that situation first themselves.
Furthermore, we speculate that this view is coherent
with the shared representations mechanism. One sees
others through one’s own embodied cognition. One
uses one’s own knowledge (including beliefs, opinions,
attitudes, feelings) as the primary basis for understand-
ing others. As suggested earlier, self-perspective may be
considered as the default mode of the human mind
because one experiences one’s own point of view more
directly. It is a very parsimonious and advantageous
mechanism to understand and predict the behavior of
others. Yet, there is a cost of making inappropriate psy-
chological inferences about others. Many social misun-
derstandings are rooted in people’s failure to recognize
and take into account the degree to which their under-
standing of a situation may differ from those of others
(Griffin, Dunning, & Ross, 1990). Flavell (1977) specu-
lated that all people may be at risk for egocentric think-
ing throughout their life:
We experience our own point of view more of less di-
rectly, whereas we must always attain the other person’s
in a more indirect manner. Furthermore, we are usually
unable to turn our own viewpoint off completely when
trying to infer the other’s, and it usually continues to ring
in our ears while we try to decode the other’s. It may
take considerable skill and effort to represent another’s
point of view accurately through this kind of noise, and
the possibility of egocentric distortion is ever present.
For successful social interaction, then, and empathic
understanding in particular, an adjustment must oper-
ate on these shared representations. Indeed, a complete
merging or confusion of self- and other-feelings is not
the goal of empathy (Batson 1987; Batson 1997; Ickes,
1997, 2003). An essential aspect of empathy is to recog-
nize the other person as like the self while maintaining a
clear separation between self and other. Hence, mental
flexibility and self-regulation are important components
of empathy. One needs to regulate one’s own perspec-
tive that has been activated by the interaction with the
others or even the mere imagination of such an inter-
action. Such a regulation is also important to modu-
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 85
Thirst unpleasant Regret water Thirst unpleasant Regret water
Self-prediction Hiker prediction
Figure 4: In this study, participants were asked to predict the feelings
of people in a situation that aroused drive states (three hik-
ers lost in the woods with neither food nor water for several
days). Participants made these predictions either immedi-
ately before or immediately after engaging in vigorous car-
diovascular exercise, which made them thirsty and warm.
The graph shows the percentage of participants before and
after exercising who indicated that they and the lost hikers
would be more bothered by thirst than hunger and would re-
gret not bringing water more than food.
SOURCE: VanBoven & Loewenstein,2003 (reprinted with permission).
late one’s own vicarious emotion so that it is not expe-
rienced as aversive. Previous research has shown that
emotion regulation is positively related to feelings of
concern for the other person (Derryberry & Rothbart,
1988; Eisenberg et al., 1994). In contrast, people who
experience their emotions intensely, especially negative
emotions, are prone to person distress—that is, an
aversive emotional reaction, such as anxiety or discom-
fort based on the recognition of another’s emotional
state or condition (Davis, 1983; Eisenberg et al., 1991).
We argue that empathy requires some form of active
inhibitory mechanism (i.e., the deliberate suppression
of a cognition or response to achieve an internally repre-
sented goal [Nigg, 2001]) and the contribution of the
prefrontal cortex has an essential role in this regula-
tion process (Fuster, 1989). A “self-regulatory disorder”
has been coined by Levine and colleagues (Levine,
Freedman, Dawson, Black, & Stuss, 1999) for the syn-
drome exhibited by these patients with ventromedial
prefrontal cortex damage (particularly on the right).
This syndrome is defined as the inability to regulate
behavior according to internal goals and constraints. It
arises from the inability to hold a mental representation
of the self on-line and to use this self-related information
to inhibit inappropriate responses. This may conse-
quently lead to deficit of empathy (see also section titled
The Lack of Empathy).
An association between executive functions and
social competence has been shown by using self-report
inventory in neurological patients with prefrontal cortex
lesions (Grattan & Eslinger, 1989). Moreover, Rowe,
Bullock, Polkey, and Morris (2001) tested neurological
patients with unilateral frontal lobe lesion on theory of
mind tasks as well as executive function tasks. They
found that both patient groups exhibited significantly
impaired performance on their ability to infer first- and
second-order beliefs. Both frontal lobe groups also
exhibited a range of deficits in tests of executive func-
tions, but analyses revealed that these seemed to be inde-
pendent of theory-of-mind impairments. A major study
with frontal lobe patients with limited focal lesions tested
for visual perspective taking and detecting deception
(Stuss, Gallup, & Alexander, 2001). The authors
reported dissociation of performance within the frontal
lobes. Right frontal lobe lesions were associated with
impaired visual perspective taking, and medial frontal
lesions, particularly right ventral, with impaired detec-
tion of deception.
A series of three neuroimaging studies performed by
our group investigated in healthy volunteers the neural
underpinning of perspective taking in three different
modalities (i.e., motoric, conceptual, and emotional) of
self-other representations. In a first study, participants
were scanned while they were asked to either imagine
themselves performing a variety of everyday actions
(e.g., winding a watch up) or imaging the experimenter
doing similar actions (Ruby & Decety, 2001). Both condi-
tions were associated with common activation in the sup-
plementary motor area (SMA), premotor cortex, and
the occipito-temporal region. This network corresponds
to the shared motor representations between the self
and the other. Taking the perspective of the other to sim-
ulate his or her behavior resulted in selective activa-
tion of the frontopolar cortex and right inferior parietal
In a second study, medical students were shown a
series of affirmative health-related sentences (e.g., tak-
ing antibiotic drugs causes general fatigue) and were
asked to judge their truthfulness either according to
their own perspective (i.e., as experts in medical knowl-
edge) or according to the perspective of a layperson
(Ruby & Decety, 2003). Although not statistically signifi-
cant, there was a tendency for response times to be
slightly greater when the participants answered the ques-
tion with the perspective of another person. The set of
activated regions recruited when the participants put
themselves in the shoes of a layperson included the
medial prefrontal cortex, the frontopolar, and right
inferior parietal lobule.
In a third study, the participants were presented with
short written sentences that depicted real-life situations
(e.g., someone opens the toilet door that you have for-
gotten to lock), which are likely to induce social emo-
tions (e.g., shame, guilt, pride), or other situations that
are emotionally neutral (Ruby & Decety, 2004). They
were asked to imagine how they would feel if they were in
those situations, and how their mother would feel in
those situations. The mother was chosen as the target of
empathy because she was the participants’ best known
person. Like in the previous experiment, a slight
increase in the response times was observed when partic-
ipants imagined the reaction of their mother elicited by
neutral situations as compared to their own reactions.
Reaction times were statistically greater when the sub-
jects imagined emotional-laden situations, both from
their own perspective and the perspective of their moth-
ers. Activation was detected in the frontopolar cortex,
the ventromedial prefrontal cortex, the medial pre-
frontal cortex, and the right inferior parietal lobule
when the participants adopted the perspective of their
mother, regardless of the affective content of the situa-
tions depicted. Cortical regions that are involved in emo-
tional processing were found activated in the conditions
that integrated emotional-laden situations, including
the amygdala and the temporal poles. The amygdala is
acknowledged to be critical for normal judgments about
the internal states of others (Adolphs, 2003). It is thus
86 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
really interesting to detect its activation for both self and
other imagined emotional reactions.
In a recent fMRI study, Seger, Stone, and Keenan
(2004) asked participants to make food preference judg-
ments about themselves or about someone else (a per-
son who they fairly knew). Self-judgments were associ-
ated with increases in the medial prefrontal cortex, the
anterior insula, and secondary somatosensory areas.
Other-judgments resulted in activation of the medial
prefrontal cortex, the frontopolar cortex, and the poste-
One of the most striking findings of this series of stud-
ies that investigated self- versus other-perspective is
the systematic involvement of the frontopolar cortex,
medial prefrontal cortex, and posterior cingulate when
the participants adopt the perspective of another person
(see Figure 5).
Converging evidence from clinical neuropsychology
(e.g., De Renzi, Cavalleri, & Facchini, 1996; Verfaellie &
Heilman, 1987) and neuroscience (e.g., Brass, Zysset, &
von Cramon, 2001; Fuster, 1989) points to the fronto-
polar cortex as being chiefly involved in inhibitory or
regulating processing. Frontal damage may result in
impaired perspective-taking ability (Price, Daffner,
Stowe, & Mesulam, 1990) and a lack of cognitive flexibil-
ity (Eslinger, 1998). Interestingly, Anderson, Bechara,
Damasio, Tranel, and Damasio (1999) reported the
cases of two patients with early damage to the anterior
prefrontal cortex (encompassing the frontopolar cortex
but not the gyrus rectus), who, when tested on moral
dilemmas, exhibited an excessively egocentric perspec-
tive. The behavior of those patients reveals a lack of inhi-
bition (or modulation) of self-perspective at the concep-
tual level. Hence, the study of Anderson et al. (1999)
provides evidence for the role of the frontopolar cortex
in inhibition at both conceptual and social levels. The
results of the three neuroimaging studies of perspective
taking support the hypothesis of such an inhibitory role
of the frontopolar cortex for adopting the subjective
viewpoint of others, whether the shared activated rep-
resentations are motor, conceptual, or emotional in
nature. Further support for this claim is provided by two
recent fMRI studies in which involvement of the right lat-
eral prefrontal cortex was detected when participants
inhibited a prepotent response in a sensory motor task
(Bunge et al., 2002) and also in a deductive-reasoning
task (Goel & Dolan, 2003).
We argue that this inhibitory component is required
to regulate and tone down the self-perspective to allow
the evaluation of the other-perspective. This is necessary
because the prepotent self-perspective, driven by the
automatic link between perception and action, is the
default mode, and this regulation allows cognitive and
affective flexibility. Such a view is compatible with the
role of the prefrontal cortex in top-down control of be-
havior (Miller & Cohen, 2001). It is also congruent with
the empathy-altruism hypothesis, which claims that a dis-
tinction between self and other, rather than a merging
between them, is required (Batson, 1991b).
An alternative interpretation for the role of the
frontopolar cortex in adopting the perspective of
another individual is based on the distinction between
different psychological operations mediated by distinct
subregions of the prefrontal cortex. There is evidence
that the frontopolar cortex is involved in the process of
evaluation of self-generated responses and is recruited
when the task requires monitoring and manipulation of
information that has been internally represented
(Christoff & Gabrieli, 2000). Adopting the subjective
perspective of another individual to understand his or
her feelings is a self-generated process that operates on
internally represented information fed by the internal
activation of shared representations.
Empathy, as presented in our model, necessitates
some level of emotion regulation to manage and opti-
mize intersubjective transactions between self and other.
Indeed, the emotional state generated by the perception
of the other’s state or situation needs regulation and
control for the experience of empathy. Without such
control, the mere activation of the shared representa-
tion, including the associated autonomic and somatic
responses, would lead to emotional contagion or emo-
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 87
Figure 5: Brain regions (frontopolar, medial prefrontal/anterior
paracingulate, and posterior cingulate cortices) found acti-
vated when subjects overtly adopt the perspective of an-
other individual versus self-perspective. The activated
clusters, represented by yellow circles, are superimposed
onto an MRI sagittal section. Numbers correspond to condi-
tion in which participants imagined actions (1: Ruby &
Decety, 2001), knowledge (2: Ruby & Decety, 2003) or feel-
ings (3: Ruby & Decety, 2004). Hemodynamic changes in
these areas are more pronounced in the right hemisphere.
Key structures in the circuitry underlying emotion
regulation play an important role in empathy (Figure 6).
Among these regions, the orbitofrontal-ventromedial
and dorsolateral cortices have been reported in the neu-
rological literature to be implicated in empathy. These
two regions have separate anatomical pathways to com-
municate with subcortical regions and seem to medi-
ate very different processes (Masterman & Cummings,
1997). The orbitofrontal is essential for regulating emo-
tion, and its damage is associated with a wide range of
social emotional deficits, including impaired social judg-
ment and disinhibited behavior. The ventromedial pre-
frontal cortex with its reciprocal connections with brain
regions involved in emotional processing (amygdala),
memory (hippocampus), and executive functions
(dorsolateral prefrontal cortex) plays a special role in
emotion regulation (Davidson, Putnam, & Larson,
2000). Damasio (1994) has proposed that somatic mak-
ers (i.e., stored memories of somatic states that are asso-
ciated with particular experiences or outcomes) are
stored in the ventromedial prefrontal cortex. It is likely
that the ventromedial region is a crucial component in
the neural network underpinning empathy (Shamay-
Tsoory, Tomer, Berger, & Aharon-Peretz, 2003). Finally,
the anterior cingulate cortex is part of a circuit involved
in a form of attention that serves to regulate both cogni-
tive and emotional processing (Bush, Luu, & Posner,
2000). Its lesion produces a host of symptoms, including
apathy, inattention, dysregulation of autonomic func-
tions, and emotional instability.
Finally, studies that investigate the cognitive and neu-
ral mechanisms involved in affective reappraisal (i.e., the
cognitive regulation of social perception and emotional
experience; Ochsner, 2004) are relevant to the under-
standing of empathy. An fMRI experiment has shown
neural correlates of emotion reappraisal in the lateral
prefrontal and medial prefrontal cortices and decreased
activity in the medial orbitofrontal cortex and the amyg-
dala (Ochsner, Bunge, Gross, & Gabrieli, 2002). Similar
neural network was identified for voluntary suppression
of sadness (Lévesque et al., 2003).
THE LACK OF EMPATHY
Although the loss of empathy has mainly been
described after lesion of the frontal lobe, more specifi-
cally the prefrontal cortex, our model suggests that there
may be distinct disorders related to empathy rather than
a unique deficit. Furthermore, since our model assumes
that empathy relies on dissociable components, it pre-
dicts a variety of structural or functional dysfunctions
depending on which aspect is disturbed. We believe that
this view is more coherent with the broad range of disor-
ders that are related to empathy and with the multidi-
mensional nature of this behavior. Indeed, we do not
think it is reasonable to assume a single source of empa-
thy deficit in very different conditions, such as socio-
pathy, conduct disorders, narcissistic personality dis-
order, Asperger’s syndrome, stroke, or traumatic brain
It is well accepted that empathic processing may be
impaired after focal lesions of the prefrontal cortex
(Eslinger, 1998). Patients with bilateral lesions of the
orbitofrontal cortex were found to be impaired in
the “faux pas” task (Stone, Baron-Cohen, & Knight,
1998). This task requires both an understanding of
false or mistaken belief and an appreciation of the emo-
tional effect of a statement on the listener (Baron-
Cohen, O’Riordan, Stone, Jones, & Plaisted, 1999). A
study conducted by Stuss and colleagues (2001) ex-
tended this finding by showing that only lesions in the
right orbitofrontal produce such a deficit. In addition,
several other patient studies reported a relationship
between the deficit in empathy and performance of cog-
nitive flexibility tasks among patients with lesions in the
dorsolateral lesions, whereas those with orbitofrontal
cortex lesions were more impaired in empathy but not in
cognitive flexibility (Grattan, Bloomer, Archambault, &
Eslinger, 1994; Shamay-Tsoory et al., 2003). Further-
more, the study by Shamay-Tsoory et al. (2003) demon-
strated that among patients with posterior lesions, only
those with damage to the right hemisphere (parietal cor-
tex) were impaired in empathy. Another recent study by
the same group tested patients with lesions of the
88 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
Figure 6: Ke y str uct ure s in v olv e d in emot i ona l reg u lat ion :
orbitofrontal (purple) and ventromedial prefrontal cortex
(green) (A), dorsolateral prefrontal cortex (violet) (B),
amygdala (orange) (C), and rostral anterior cingulate cortex
SOURCE: Davidson, Putnam, & Larson, 2000 (reprinted with permis-
ventromedial prefrontal cortex or dorsolateral pre-
frontal cortex with three theory-of-mind tasks (second
beliefs and faux pas) differing in the level of emotional
processing involved (Shamay-Tsoory, Tomer, Berger,
Goldsher, & Aharon-Peretz, in press). The authors
found that patients with ventromedial lesions were most
impaired in the faux pas task but presented normal per-
formance in the second-belief tasks. They further
argued that to detect faux pas, one is required not only to
understand the knowledge of the other but also to have
empathic understanding of their feelings. Unfortu-
nately, there is no clear indication of the lesion sites in
these studies. Also the ventromedial prefrontal region of
interest encompasses both the orbitofrontal, the ventro-
medial, as well as medial aspects (including the
paracingulate gyrus). Nevertheless, this study shows that
it is possible to distinguish within the same group of
patients quite specific deficits of social cognition.
From these patient studies aforementioned, it can be
tentatively concluded that different parts of the right
prefrontal cortex are involved in the capacity to reason
about the feelings of others, including the ability to
adopt the perspective of others. It is not yet clear what
specific process each subregion subserves. Moreover, as
described in Box 1, one of the most challenging limits to
comparison between studies of empathy stems from the
use of different tools and methods.
Some researchers have theorized that there is a rela-
tion between aggressive behavior and a lack of empathy.
The tendency to have low concern for the needs of other
and the consequences of one’s own actions seems to be a
common characteristic of disruptive behavior disorders
(Zahn-Waxler, Cole, Welsh, & Fox, 1995). A recent study
by Gill and Calkins (2003) that examined situational
empathy behaviors, including physiological measures in
2-year-olds identified as either high or low in aggressive/
destructive behaviors, produced mixed results that do
not fully support prior research.
Empathy deficit in antisocial personality disorder has
been suggested to come from a reduced ability to feel
other people’s emotional state and more so for sadness
and fear (Blair, 1995). This deficit has been ascribed to a
dysfunction in the amygdala of developmental origin
(Blair, 2001). This is also compatible with the fact that
individuals with this disorder have generally intact exec-
utive functions and can successfully complete theory-of-
mind tasks (Hare, 1993). Therefore, their lack of empa-
thy would be related to disrupted affective processing
rather than an inability, for instance, to adopt the per-
spective of others. In fact, people with antisocial person-
ality disorders are probably good at perceiving others’
intentions while disregarding the emotional content
and may thus take advantage of it. This is precisely what
the research of Mealey (1995) suggests. The psychopath
cannot simulate emotions he cannot experience and
must rely exclusively on cognitive inputs to his theory-of-
A very interesting single case of acquired sociopathy
has been investigated by Blair and Cipolotti (2000).
The authors investigated an individual, J.S., with orbito-
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 89
Box 1. Measurements Used inthe Evaluationof EmpathyAbility AcrossDifferentStudies (note that thislist is notintended to beexhaustive).
•Emotional attribution task: The participant is presented with short stories describing emotional situations and asked what
the protagonists might feel in those situations (e.g., Blair & Cipolotti, 2000).
•Faux pas task: The participant is read a story with the occurrence of a faux pas and asked if she or he detected the faux
pas—that is, a socially awkward situation (e.g., Stone, Baron-Cohen, & Knight, 1998).
•Interpersonal Reactivity Index (Davis, 1983): The participant is asked to fill in self-report scales that contain a numberof
items to assess various psychological aspects of cognitive and emotional empathy.
•Balance Emotional Empathy Scale (Mehrabian & Epstein, 1972): A 33-item questionnaire targeting emotional empathy.
•Hogan’s (1969) Empathy Scale: Measures an individual’s cognitive ability to understand another’s viewpoint using a 64 true-
•Measures of autonomic nervous system responses: Skin conductance, heart rate, heart period, respiration rate are used to moni-
tor the changes during a given task—for instance, watching emotion-laden pictures or listening to short stories (e.g.,
Blair, 1999; Decety & Chaminade, 2003b).
•Dyadic interaction (Ickes, Stinson, Bissonnette, & Garcia, 1990): For instance, after two people have been seated in a “waiting
room,” their interaction is unobtrusively videotaped. Then, in a second phase, the participants are shown the videotape
and asked to report their own thoughts and feelings during the interaction and infer each of their interaction partner’s
reported thoughts and feelings.
•Behavioral measures in empathy-eliciting situations, such as latency to respond to the feigned distress of another (e.g.,
Zahn-Waxler, Radke-Yarrow, Wagner, & Chapman, 1992).
•Participants are assigned to groups that differ in terms of perspective-taking instructions (e.g., imagine how the person
feels vs. imagine how you would feel in this situation) before watching videotape or listening to an emotionally laden story.
Then, they complete reaction questionnaires using Likert-type scales to measure their emotional response to the stimuli
(e.g., Batson, 1997).
frontal cortex and left amygdala damage, with an impres-
sive battery of measures including skin conductance
response (SRC), tests of executive functions, emotion
recognition, and social cognition tasks. While J.S.
showed executive impairments but no reversal learning
impairment, he was significantly impaired on most of the
social cognition tasks. Notably, he was both impaired in
the recognition of emotional expressions (happiness,
anger, disgust, and sadness) and in the attribution of
emotional states to others (fear, anger, and embarrass-
ment). His ability to attribute mental states to others was
preserved. His SRC responses to negative emotional
expressions were reduced. Blair and Cipolotti argued
that the distinctive features of the acquired sociopathy
of J.S. were the result of impairment of a system that
responds to angry expressions/expectations of others’
anger and that this system is particularly involved in the
suppression of socially aberrant behavior.
Antisocial personalities are often reported to per-
form poorly on neuropsychological tests of execu-
tive functioning (e.g., Séguin, Pihl, Harden, Tremblay,
& Boulerice, 1995). Executive functions are consid-
ered necessary for socially appropriate conduct, and in
our model, they contribute to empathy through self-
regulation. Morgan and Lilienfeld (2000) conducted
a meta-analysis of 39 studies (yielding a total of 4,589
participants) to clarify the relation between antisocial
behavior and executive functions. The results of this
meta-analysis indicate that there is a robust and statisti-
cally significant relation between executive functions
and antisocial behavior. The authors were unable to sub-
divide executive function measure in terms of their asso-
ciations with different brain regions (e.g., dorsolateral,
orbitofrontal) because of the lack of knowledge con-
cerning the neuroanatomical substrates of most execu-
tive functions tasks. Interestingly, Blair (1995) proposed
that people with antisocial personalities have a disrup-
tion of a violence inhibition mechanism that is normally
triggered by distress cues of others, and this aspect be-
longs to executive functioning.
Clinical and forensic research distinguish “affective”
or “reactive” aggression, which is a response to physical
or verbal aggression initiated by others with violence
that is relatively uncontrolled and emotionally charged,
from a “predatory” or “instrumental” cold-blooded
aggression, which is a controlled, purposeful aggression
lacking in emotion that is used to achieve a desired goal
(Blair, 2001; Dodge, Lochman, Harnish, Bates, & Petit,
1997). Our model of empathy predicts that the former
type of personality would lack executive control (partic-
ularly self-control), whereas the latter personality would
have some dysfunctions in sharing feelings with others.
Interestingly, measurements of glucose metabolism in
two groups of affective and predatory murderers have
shown that the first group has lower prefrontal activity,
and the second group has similar prefrontal activity as
compared to controls but lower activity at the subcortical
level, including the amygdala (Raine et al., 1998).
Children with autism, a neurodevelopmental disor-
der, display a broad range of social communication defi-
cits, and most scholars agree that a lack of empathy
prominently figures amongst them (Frith, 2001). The
underlying cause of the empathy deficit is, however,
more controversial. Baron-Cohen, Leslie, and Frith
(1985) proposed the hypothesis that the social impair-
ment in autism arises from a failure of a mentalizing
mechanism (a theory-of-mind module). Other authors
believe that children with autism have a hard time feel-
ing and expressing emotion and that this basic deficit
prevents them from engaging in social interactions (e.g.,
Hobson, 1989). Others still, such as Russell (1996),
argue that deficits in executive functions are the major
cause for the social/communicative disorders observed
in autism. Rogers and Pennington (1991) suggested a
cascade model of autism in which the lack of certain
aspects of interpersonal development at every previous
stage disrupts certain developments in the following
stage. These authors view early imitation skills, emotion
sharing, and theory of mind as increasingly complex
expressions of the ability to form and coordinate certain
representations of self and other. These representations
are then used to guide the planning and execution of
one’s own behavior. Finally, Dawson (1991) proposed
that autism involves impairment in attentional function-
ing for social stimuli (e.g., facial expressions, speech,
gestures). She hypothesized that because social stim-
uli are complex, variable, and unpredictable, chil-
dren with autism have difficulty processing and repre-
senting them, and therefore, their attention is not
naturally drawn to such stimuli (Dawson, Meltzoff,
Osterling, Rinaldi, & Brown, 1998). These different
views (imitation/emotions sharing vs. executive func-
tions) are not incompatible with our model of empathy,
as it remains possible that empathy deficits in autism are
related to disruption of either emotion sharing or
mental flexibility/self-regulation components or even
When compared with developmentally delayed chil-
dren, 20-month-old infants with autism were found to be
specifically impaired on empathy tasks, joint attention,
and imitation (Charman et al., 1997). Imitation deficits
have been proposed to explain the difficultly of autistic
children in establishing social relationships and identify-
ing with others (Meltzoff & Gopnik, 1993; Rogers, 1999).
For instance, a study by Hobson and Lee (1999) demon-
strated that autistic children can imitate the goal of
actions displayed by an experimenter but that they failed
to imitate the affective style with which the actions were
90 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
carried out. This suggests that these children cannot
readily adopt the experimenter’s perspective entirely. It
has also been demonstrated that long before children
with autism show theory-of-mind deficits, they show
deficits in joint attention and attention monitoring
(Osterling & Dawson, 1994). A study examined 30- to 70-
month-old autistic and normal children’s social behav-
ior, affect, and use of gaze during naturalistic inter-
actions with their mothers (Dawson, Hill, Spencer,
Galpert, & Watson, 1990). Both autistic and normal chil-
dren responded with smiles more frequently to social
events than to nonsocial events. However, when autistic
children’s responses to the mother’s smiles were exam-
ined, the authors found that they never smiled in
response to the mother’s smile. In other words, they do
not exhibit the biologically based ability to respond
empathically to others (Hobson, 1989).
Several studies have examined behavioral and auto-
nomic responses of children with autism who look at
adults depicting facial emotional expressions (see Blair,
2003, for a critical review). Although most studies report
that children with autism look less frequently at the adult
faces than control subjects in empathy-eliciting situa-
tions, the remaining findings are equivocal. For
instance, one study did not find a reduction of heartbeat
rate during the observation of someone in distress
(Corona, Dissanayake, Arbelle, Wellington, & Sigman,
1998). Another study has shown that the autonomic
responses of these children change according to the dis-
tress of the target, if the emotions displayed are not
ambiguous and if they are presented under conditions
with reduced distraction (Blair, 1999). Moreover, and
contrary to what is often claimed, children with autism
can make moral/conventional distinction (Blair, 1996).
It is likely, however, that these children present a diffi-
culty in taking the perspective of others, which requires
executive resources, but they seem to have the physiolog-
ical substrate to display affective sharing abilities. Alto-
gether, both impairment in executive functions and
emotion sharing may account for the empathy deficit in
Given the complexity of human empathy, many areas
of research and theory are necessary for its understand-
ing, including evolutionary psychology, comparative psy-
chology, developmental science, social psychology,
neuropsychology, and cognitive neuroscience. Each dis-
cipline contributes valuable information, especially if
the data can be integrated into a functional model.
Is it possible to identify at the computational level
primitive constructs that are recruited in empathic
understanding? Psychological models use hypothetical
representations and processes that operate on those
representations. What is the psychological reality of
the constructs people employ to account for empathy?
A construct is composed of process and representa-
tion working together to serve a particular function
(Willingham & Dunn, 2003). Self-awareness, executive
functions, theory of mind, affective reappraisal, and per-
spective taking are all psychological constructs, but none
of these constructs can be described as a primitive one.
They further need to be reduced, and each is likely to be
implemented by a complex neural machinery. Thus, as
stated by Cacioppo et al. (2003), the most powerful tool
in achieving sound social neuroscience research
remains the expertise, intelligence, and creativity of the
investigators and not a higher field strength magnet.
One important theoretical aspect is whether there is
partial overlap between theory-of-mind and empathic-
understanding processes. Should these two psychologi-
cal constructs be separated—one for emotion and affect
processing and the other for other mental-states attri-
bution? Most neuroimaging studies using theory-of-
mind tasks have detected specific activation in the
medial prefrontal cortex around the paracingulate
sulcus (Gallagher & Frith, 2003; Happé, 2003, for recent
reviews). Our model of empathy includes explicit pro-
cessing of the mental states of self and others and thus
requires anterior cingulate computing resources similar
to that of mentalizing tasks. Conversely, Farrow and col-
leagues (2001) argued from the results of an fMRI study
in which subjects were required to predict and experi-
ence the emotions of others, the neural basis of empathy
is distinct from that subserving inference of other’s
intentions. It is therefore an important task in the future
to explore the respective computational role of every key
region of the prefrontal cortex (including the medial
prefrontal cortex, the anterior paracingulate sulcus,
and the ventromedial prefrontal cortex) in mental,
affective state attribution, as well as in executive func-
tions, in relation to how humans navigate the social
world. Passingham, Stephan, and Kötter (2002) have
argued that each cytoarchitectonic area has unique pat-
terns of cortico-cortical connections that determine its
function, and differences in neural activity during dis-
tinct tasks are produced by distributed subsystems of
brain regions. Even though there is massive parallel pro-
cessing, the temporal dynamics of activation in these
regions are also an important aspect to be investigated
Another interesting issue is whether there are gender
differences in empathy. If so, are they learned or related
to hormonal and innate differences in the way our brain
is shaped? The work in social psychology, although not
entirely conclusive, has seriously questioned the alleged
female-superiority in empathic understanding, suggest-
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 91
ing motivational differences between the genders
instead (Ickes, 2003). But perhaps certain, more specific
differences are biologically based, as suggested by the
results of a recent fMRI study that investigated neural
response in men and women to infant crying and laugh-
ing and showed significant differences between the two
groups (Seifritz et al., 2003). Women but not men, inde-
pendent of their parental status, showed neural deacti-
vation in the anterior cingulate cortex in response to
infant crying and laughing. In addition, the response
pattern in the amygdala and interconnected limbic
structures changed fundamentally with parental expe-
rience in both men and women. Nonparents showed
stronger activation from laughing, whereas parents
showed stronger activation for crying. These results
seem to demonstrate that the emotion-sharing compo-
nent may be subjected to personal experience and/or
emotion regulation is prepared biologically different in
men and women.
Some people have greater empathic ability than oth-
ers, as demonstrated by the work of Marangoni, Garcia,
and Ickes (1995) and Ickes et al. (2000). Could these dif-
ferences be related to individual differences in personal-
ity traits, and how do they fit into this multicomponent
model of empathy? Can social neuroscience help to
better understand the origins of these individual differ-
ences? An interesting and testable personality difference
related to emotion regulation is that of temperament. It
is a moderately stable psychological profile in quality
and intensity of emotional reaction, attention, and self-
regulation under some genetic constraints that emerges
during childhood (Kagan, 1998; Rothbart & Bates,
1998). In addition, research with adults shows that there
are a number of different strategies people use to regu-
late emotion. For example, reappraisal and suppression
are both effective ways to regulate emotion, but they
have quite distinct mechanisms and side effects. In gen-
eral, these strategies can be classified as being either
response focused or antecedent focused (Gross, 2001).
How do these different strategies modulate empathy,
and how are they expressed in the nervous system?
Lesion studies are needed to further elucidate a
causal role of any given structure in the neural systems
involved in empathy. Nonetheless, it is not easy to evalu-
ate empathic disorder in clinical settings, and often
patients with different etiologies are pooled into sepa-
rate groups provided that their lesions fall into a prede-
fined category. If possible, more discrete subdivision of
the prefrontal cortex is necessary because each subre-
gion is likely to play a specific role in empathy behavior.
For instance, ventromedial, orbitofrontal, and medial
prefrontal cortices subserve distinct functions in social-
emotional cognition. Also, too often, empathy abilities
in patient populations are measured only with self-
report and rating inventories. A combination of physio-
logical measures and empathy-eliciting tasks would
provide deeper knowledge into the mechanisms of this
Finally, of special interest will be the understanding of
what motivates us to feel empathy in the sense of caring
for the other (see Batson et al., 2003). Indeed, empathy
is a motivated behavior and does not so often get auto-
matically triggered. Most of the time, this behavior is reg-
ulated by top-down processing involving cultural values,
concepts, and the like. For instance, although a large
number of people were apparently devastated by the
Challenger disaster and expressed a lot of sympathy
(especially for the children who witnessed the dramatic
death of their school teacher), it seems that fewer people
have expressed concern for the 800,000 individuals,
including newborns, that have been slaughtered in the
1994 Rwanda genocide. Is it a matter of psychological
identification or just a difference as the result of a dispar-
ity in media coverage? It is certainly easier to identify
with one individual than with many. This idea is sup-
ported by the findings that judgments in response to per-
sonal moral dilemmas compared with impersonal ones
involve greater activity in brain areas associated with
emotion and social cognition (Greene, Sommerville,
Nystrom, Darley, & Cohen, 2001). But is this the sole ex-
planation of such an asymmetry between the reactions to
those dramatic events, knowing that beliefs, opinions,
suppositions, attitudes, cultural norms, and related
states of mind have a top-down modulatory influence
on empathic processing? For instance, Nelson and
Baumgarte (2004) have shown that individuals experi-
ence less emotional and cognitive empathy for a target
experiencing distress stemming from an incident re-
flecting unfamiliar cultural norms and that this reduc-
tion of empathy is mediated by a lack of perspective tak-
ing on the part of the observer. These findings suggest
that modulatory mechanisms incorporate internal
representations of prior experience as well as similarity
between self and other.
EMPATHY BEYOND MENTAL SIMULATION
OF THE SUBJECTIVITY OF THE OTHER
The way our nervous system is organized and tai-
lored by evolution provides the basic mechanism for
resonating with others, as well as the capacity to simu-
late our own actions, their consequences, and also the
actions of others (Jackson & Decety, 2004). This shared-
representations mechanism (i.e., distributed neural pat-
terns temporarily activated by actual perception or
evoked from memory), driven by the common coding
between perception and action, provides the default
92 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
mode of self-processing (or tendency) to relate implic-
itly to others. Humans come prepared with the innate
motivation to seek engagements of conspecifics and
implicitly learn that others are similar to themselves
(Meltzoff, 2002). It is through bodily activity that one
first grasps the presence of others and then gains insight
into their subjective lives (Gallagher, 2001). It is thus not
productive to disentangle cognition and affects from
We argued that the shared-representations mecha-
nism is responsible for the human projective tendency,
which needs to be regulated (or calibrated) when shar-
ing emotions or when adopting the perspective of others
to understand their feelings. This requires additional
processing mechanisms, including monitoring and
manipulation of internal information generated by the
activation of the shared representations.
One of the main components of empathy is based on
a mental simulation of the subjectivity of others, which
can be initiated in two ways: automatically or intention-
ally. The idea of an unconscious and automatic simu-
lation is far from new (see Gallese, 2001; Gallese &
Goldman, 1998; Goldman, 1993). For instance, Lipps
(1903) suggested that an involuntary, instinctual, “kines-
thetic” imitation of the observed vital activity of another
occurs in empathy. When empathy produces this “physi-
cal mimicry” in the spectator, the intentional focus does
not remain on the spectator’s body but is projected into
the other. Later, Ax (1964) suggested that empathy
might be thought of “as an autonomic nervous system
state, which tends to simulate that of another person.” In
psychoanalysis, Basch (1983) speculated that because
their respective autonomic nervous systems are geneti-
cally programmed to respond in like fashion, a given
affective expression by a member of a particular species
tends to recruit a similar response in other members of
that species. This is done through the promotion of an
unconscious autonomic imitation of the sender’s bodily
state and facial expression by the receiver. This generates
in the receiver the autonomic response associated with
that bodily state and facial expression, which is to say, the
receiver experiences an affect identical with that of the
sender (p. 108). This view was further developed by
Levenson and Ruef (1992), who found evidence that a
perceiver’s accuracy in inferring a target’s negative emo-
tional states was related to the degree of physiological
synchrony between the perceiver and the target. In other
words, when two people feel similar emotions, they more
accurately perceive each other’s intentions and
The idea that the knowledge about emotions ex-
pressed by others relies on a simulation of how the emo-
tion would feel in the perceiver was also proposed by
Damasio (1994, 2003). The presumed mechanism for
such a simulation involves an internal brain simulation
that consists of rapid modification of ongoing body
maps. The discovery that the somatosensory cortex is
involved in the recognition of emotions provided the
first direct evidence in favor of such an unconscious
simulation process (Adolphs et al., 2000). Recently,
Goldman and Sripada (in press) have provided several
detailed cognitive models for a simulational approach to
face-based emotion recognition marshaling neurologi-
cal evidence for paired deficits between emotion pro-
duction and emotion perception.
However, this simulation is not exclusively under
automatic management, and it falls, at least in humans,
under conscious control. This makes empathy, as
described here, an intentional capacity. In many cases,
the outcome of the simulation mechanism is not em-
pathic feeling. In addition, without self-awareness and
emotion-regulation processing, there is no true empa-
thy. Indeed, the activation of shared representations
would lead to anxiety or discomfort. This formulation is
consistent with the observation that prosocial behaviors,
which stem from empathy, emerge in parallel with self-
conscious emotions. These emotions require self-
evaluation and comparison with other selves, as well as
some form of regulation.
Forming an explicit representation of another per-
son’s feeling as an intentional agent therefore necessi-
tates additional computational mechanisms beyond the
shared representation level. This requires that second-
order representations of the other are available to the
consciousness (a decoupling mechanism between first-
person information and second-person information),
for which the anterior paracingulate cortex plays a
unique function (Frith & Frith, 2003). Thus, empathy is
not a simple resonance of affect between the self and
other. It involves an explicit representation of the subjec-
tivity of the other. It is a consciously experienced phe-
nomenon. Recent neuroimaging investigations of the
perception of pain in others support such a view (Jack-
son et al., 2004; Morrison, Lloyd, di Pellegrino, & Rob-
erts, in press; Singer et al., 2004). All these studies have
shown that only part of the network mediating pain
experiences (including the anterior cingulate cortex
and the insula) is shared when empathizing or evaluat-
ing the pain in others. Most importantly, empathy also
necessitates emotion regulation for which the ventral
prefrontal cortex, with its strong connections with the
limbic system, dorsolateral, and medial prefrontal areas,
plays an important role. Once again, we do not assume
that there is a unitary empathy system (or module) in the
brain. Rather, we consider multiple dissociable systems
to be involved in the experience of empathy.
Finally, as suggested earlier, empathy is a motivated
process that more often than commonly believed is trig-
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 93
gered voluntarily. This makes empathy a flexible human
capacity as well as a method of gaining knowledge of
understanding another, and it is susceptible to social-
cognitive intervention, such as through training or en-
hancement programs for targeting various goals (e.g.,
reeducation of antisocial personalities, training of psy-
chotherapists or physicians, and training early at-risk-
Adolphs, R. (2002). Recognizing emotion from facial expressions:
Psychological and neurological mechanisms.B ehavioral and Cogni-
tive Neuroscience Reviews,1, 21-62.
Adolphs, R. (2003). Is the human amygdala specialized for process-
ing social information? Annals of the New York Academy of Sciences,
Adolphs, R., Damasio, H., & Tranel, D. (2002). Neural systems for
recognition of emotional prosody: A 3-D lesion study. Emotion,2,
Adolphs, R., Damasio, H., Tranel, D., Cooper, G., & Damasio, A.
(2000). A role for the somatosensory cortices in the visual recogni-
tion of emotion as revealed by three dimensional lesion mapping.
Journal of Neuroscience,20, 2683-2690.
Adolphs, R., Tranel, D., Damasio, H., & Damasio, A. (1995). Fear and
the human amygdala. Journal of Neuroscience,15, 5879-5891.
(2001). The anterior cingulate cortex: the evolution of an inter-
face between cognition and emotion. Annals of the New York Acad-
emy of Science,935, 107-117.
Anderson, S. W., Bechara, A., Damasio, H., Tranel, D., & Damasio,
A. R. (1999). Impairment of social and moral behavior related to
early damage in human prefrontal cortex. Nature Neuroscience,2,
Asendorpf, J. B., & Baudonnière, P. M. (1993). Self-awareness and
other-awareness: Mirror self-recognition and synchronic imita-
tion among unfamiliar peers. Developmental Psychology,29, 88-95.
Asendorpf, J. B., Warkentin, V., & Baudonnière, P. M. (1996). Self-
awareness and other-awareness II: Mirror self-recognition, social
contingency awareness, and synchronic imitation. Developmental
Ax, A. A. (1964). Goals and methods of psychophysiology.
Baron-Cohen, S., Leslie, A., & Frith, U. (1985). Does the autistic child
have a theory of mind? Cognition,21, 37-46.
(1999). Recognition of faux pas by normally developing children
and children with Asperger syndrome or high-functioning
autism. Journal of Autism and Developmental Disorders,29, 407-418.
Baron-Cohen, S., Ring, H. A., Wheelwright, S., Bullmore, E. T.,
Brammer, M. J., Simmons, A., et al. (1999). Social intelligence in
the normal and autistic brain: an fMRI study. European Journal of
Barresi, J., & Moore, C. (1996). Intentional relations and social
understanding. Behavioral and Brain Sciences,19, 107-154.
Barrett, L., Henzi, P., & Dunbar, R. I. M. (2003). Primate cognition:
From what now to what if. Trends in Cognitive Science,7, 494-497.
Basch, M. F.(1983). Empat hicunderst anding: a reviewof the concept
and some theoretical considerations. Journal of the American Psycho-
analytic Association,31, 101-126.
Batson, C. D. (1987). Prosocial motivation: Is it ever truly altruistic? In
L. Berkowitz (Ed.), Advances in experimental social psychology (Vol.
20, pp. 65-122). New York: Academic Press.
Batson, C. D. (1990). How social an animal? The human capacity for
caring. American Psychologist,45, 336-346.
Batson, C. D. (1991a). The altruism question: Toward a social-psychologi-
cal answer. Hillsdale, NJ: Lawrence Erlbaum.
Batson, C. D. (1991b). Empathic joy and the empathy-altruism
hypothesis. Journal of Personality and Social Psychology,61, 413-426.
Batson, C. D. (1997). Self-other merging and the empathy-altruism
hypothesis: Reply to Neuberg et al. Journal of Personality and Social
ing social psychology.
Gale, S., et al. (2003). As you would have them do unto you: Does
imagining yourself in the other’s place stimulate moral action?
Personality and Social Psychology Bulletin,29, 1190-1201.
Dawson, K. (1997). Is empathy-induced helping due to self-other
merging?. Personality and Social Psychology Bulletin,73, 495-509.
Benton, A., & Silvan, A. B. (1993). Disturbance of body schema. In
K. M. Heilman & E. Valenstein (Eds.), Clinical Neuropsychology
(pp. 123-140). Oxford: Oxford University Press.
Beres, D., & Arlow, J.A. (1974). Fantasy and identification in empathy.
Psychoanalytic Quarterly,43, 26-50.
Bernstein, D. M., Atance, C., Loftus, G., & Meltzoff, A. N. (2004). We
saw it all along: Visual hindsight bias in children and adults. Psy-
chological Science,15, 264-267.
Berthoz, S., Armory, J. L., Blair, R. J. R., & Dolan, R. J. (2002). An fMRI
study of intentional and unintentional (embarrassing) violation
of social norms. Brain,125, 1696-1708.
Blair, R. J. R. (1995). A cognitive developmental approach to moral-
ity: Investigating the psychopath. Cognition,57, 1-29.
Blair, R. J. R. (1996). Morality in the autistic child. Journal of Autism
and Developmental Disorders,26, 571-579.
Blair, R. J. R. (1999). Psychophysiological responsiveness to the dis-
tress of others in children with autism. Personality and Individual
Blair,R. J. R. (2001) . Neurocognitive modelsof aggression, t heant iso-
cial personality disorders, and psychopathy. Journal of Neurology,
Neurosurgery and Psychiatry,71, 727-731.
Blair, R. J. R. (2003). Did Cain fail to represent the thoughts of Abel
before he killed him? The relationship between theory of mind
differences in theory of mind (pp. 143-169). Hove, UK: Psychology
Blair, R. J. R., & Cipolotti, L. (2000). Impaired social response rever-
sal: A case of acquired sociopathy. Brain,123, 1122-1141.
Blakemore, S.-J., & Decety, J. (2001). From the perception of action to
the understanding of intention. Nature Reviews Neuroscience,2, 561-
Blakemore, S.-J., & Frith, C. D. (2003). Self-awareness and action. Cur-
rent Opinion in Neurobiology,13, 219-224.
Bloch, H. (1997). Sensory-motor behavioral organization and
changes in infancy. Encyclopaedia of Human Biology,7, 765-773.
Bohart, A., & Greenberg, L. S. (1997). Empathy reconsidered. Washing-
ton, DC: American Psychological Association.
Boyer, P. (2001). Religion explained: The human instincts that fashion
gods, spirits and ancestors. New York: William Heinnemann.
Boyer, P., & Barrett, H. C (2004). Evolved intuitive ontology: Integrat-
ing neural, behavioural and developmental aspects of domain-
specificity. In D. Buss (Ed.), Handbook of evolutionary psychology.
Cambridge, MA: MIT Press.
Brass, M., Zysset, S., & von Cramon, D. Y. (2001). The inhibition of
imitative response tendencies. NeuroImage,14, 1416-1423.
Bretherton, I., Fritz, J., Zahn-Waxler, C., & Ridgeway, D. (1986). The
acquisition and development of emotion language: A functional-
ist perspective. Child Development,57, 529-548.
Brothers, L. (1989). A biological perspective on empathy. The Ameri-
can Journal of Psychiatry,146, 10-19.
Brownell, C. A., & Carriger,M. S. (1990) . Changes in cooperation and
self-other differentiation during the second year. Child Develop-
Brüne, M. (2001). Social cognition and psychopathology in an evolu-
tionary perspective. Psychopathology,34, 85-94.
Brunet, E., Sarfati, Y., Hardy-Baylé, M. C., & Decety, J. (2000). A PET
investigation of the attribution of intentions with a nonverbal
task. NeuroImage,11, 157-166.
94 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
Brunet, E., Sarfati, Y., Hardy-Baylé, M. C., & Decety, J. (2003). A PET
study of the attribution of intentions to others in schizophrenia.
Buck, R., & Ginsburg (1997). Communicative genes and the evolu-
tion of empathy. In W. Ickes (Ed.), Empathic accuracy (pp. 17-43).
New York: Guilford.
Bunge, S. A., Dudukovic, N. M., Thomason, M. E., Vaidya, C. J., &
Gabrieli, J. D. E. (2002). Immature frontal lobe contributions to
cognitive control in children: Evidence from fMRI. Neuron,33,
Bush, G., Luu, P., & Posner, M. I. (2000). Cognitive and emotional
influences in anterior cingulate cortex. Trends in Cognitive Science,
& Nusbaum, H. (2003). Just because you’re imaging the brain
doesn’t mean you can stop using your head: A primer set of first
principles. Journal of Personality and Social Psychology,85, 650-661.
Calder, A. J., Keane, J., Manes, F., Antoun, N., & Young, A. W. (2000).
Impaired recognition and experience of disgust following brain
injury. Nature Neuroscience,3, 1077-1078.
Campos, J. J., & Stenberg, C. R. (1981). Perception, appraisal, and
emotion: The onset of social referencing. In M. E. Lamb & L. R.
Sherrod (Eds.), Infant social cognition: Empirical and theoretical con-
siderations (pp. 273-314). Hillsdale, NJ: Lawrence Erlbaum.
Carlozzi, A. F., Bull, K. S., Stein, L. B., Ray, K., & Barnes, L. (2002).
Empathy theory and practice: A survey of psychologists and coun-
selors. The Journal of Psychology,136, 161-170.
Carlson, S. M., & Moses, L. J. (2001). Individual differences in inhibi-
tory control and children’s theory of mind. Child Development,72,
Carr, L., Iacoboni, M., Dubeau, M. C., Mazziotta, J. C., & Lenzi, G. L.
(2003). Neural mechanisms of empathy in humans: A relay from
neural systems for imitation to limbic areas. Proceedings of National
Academy of Science USA,100, 5497-5502.
Castelli, F., Happé, F., Frith, U., & Frith, C. D. (2000). Movement in
mind: A functional imaging study of perception and interpreta-
tion of complex intentional movement patterns. NeuroImage,12,
Chaminade, T., & Decety, J. (2002). Leader or follower? Involvement
of the inferior parietal lobule in agency. NeuroReport,13, 1975-
Drew, A. (1997). Infants with autism: An investigation of empathy,
pretend play, joint attention, and imitation. Developmental Psychol-
Chartrand, T. L., & Bargh, J. A. (1999). The chameleon effect: The
perception-behavior link and social interaction. Journal of Person-
ality and Social Psychology,71, 464-478.
Chartrand, T.L. , Maddux, W. W., &Lakin, J. L. ( inpress) . Beyondt he
perception-behavior link: The ubiquitous utility and motivational
moderators of nonconscious mimicry. In R. Hassin, J. Uleman, &
J. A. Bargh (Eds.), Unintended thoughts 2: The new unconscious.
Chiron, C., Jambaque, J., Nabbout, R., Lounes, R., Syrota, A., &
Dulac, O. (1997). The right brain hemisphere is dominant in
human infants. Brain,120, 1057-1065.
Christoff, K., & Gabrieli, J. D. E. (2000). The frontopolar cortex and
human cognition: Evidence for a rostrocaudal hierarchical orga-
nization within the human prefrontal cortex. Psychobiology,28,
Cialdini, R., Brown, S., Lewis, B., Luce, C., & Neuberg, S. (1997).
Reinterpreting the empathy-altruism relationship:When one into
one equals oneness. Journal of Personality and Social Psychology,73,
Clark, S., Tremblay, F., & St.-Marie, D. (2003). Differential modula-
tion of the corticospinal excitability during observation, mental
imagery and imitation of hand actions. Neuropsychologia,42, 105-
Corona, C., Dissanayake, C., Arbelle, A., Wellington, P., & Sigman, M.
(1998). Is affect aversive to young children with autism? Behav-
ioral and cardiac responses to experimenter distress. Chil d Develop-
Cummins, D. D., & Cummins, R. (1999). Biological preparedness and
evolutionary explanation. Cognition,73, B37-B53.
Damasio, A. R. (1994). Descartes’ error: Emotion, reason, and the human
brain. New York: Oxford University Press.
Damasio, A. R. (2003). Looking for Spinoza: Joy, sorrow, and the feeling
brain. Orlando, FL: Harcourt.
Davidson, R. J., Pizzagalli, D., Nitschke, J. B., & Kalin, N. H. (2003).
Parsing the subcomponents of emotion and disorders of emotion:
Perspective from affective neuroscience. In R. J. Davidson, K. R.
Scherer,&H.H.Goldsmith(Eds.),Handbook of affective sciences
(pp. 8-24). New York: Oxford University Press.
Davidson, R. J., Putnam, K. M., & Larson, C. L. (2000). Dysfunction in
the neural circuitry of emotion regulation: A possible prelude to
violence. Science,289, 591-594.
Davis, M. H. (1983). Measuring individual differences in empathy:
Evidence for multidimensional approach. Journal of Personality and
Social Psychology,44, 113-126.
Davis, M. H. (1996). Empathy a social psychological approach.Boulder,
Dawson, G. (1991). A psychobiological perspective on the early
socioemotional development of children with autism. In S. Toth &
D. Cicchetti (Eds.), Rochester symposium on developmental psychology
(Vol. 3, pp. 207-234). Hillsdale, NJ: Lawrence Erlbaum.
Dawson, G., Hill, D., Spencer, A., Galpert, L., & Watson, L. (1990).
Affective exchanges between young autistic children and their
mothers. Journal of Abnormal Child Psychology,18, 335-345.
(1998). Children with autism fail to orient to naturally occurring
social stimuli. Journal of Autism and Developmental Disorders,28, 479-
Decety, J. (2002a). Le sens des autres ou les fondements naturels de la
sympathie [The sense of others: natural foundations of sympa-
thy].InY.Michaud(Ed.),Qu’est-ce que la Vie Psychique? (pp. 71-
101). Paris: Editions Odile Jacob.
Decety, J. (2002b). Naturaliser l’empathie [Empathy naturalized].
Decety, J., & Chaminade, T. (2003a). Neural correlates of feeling sym-
pathy. Neuropsychologia,41, 127-138.
Decety, J., & Chaminade, T. (2003b). When the self represents the
other: A new cognitive neuroscience view of psychological identi-
fication. Consciousness and Cognition,12, 577-596.
Decety, J., Chaminade, T., Grèzes, J., & Meltzoff, A. N. (2002). A PET
exploration of the neural mechanisms involved in reciprocal imi-
tation. NeuroImage,15, 265-272.
et al. (1997). Brain activity during observation of action: Influ-
ence of action content and subject’s strategy. Brain,120, 1763-
Decety, J., & Hodges, S. D. (2004). The social neuroscience of empa-
thy. In P. A. M. La nge (Ed.), Bridging social psychology. Hillsdale, NJ:
Mazziotta, J. C., et al. (1994). Mapping motor representations
with positron emission tomography. Nature,371, 600-602.
Decety, J., & Sommerville, J. A. (2003). Shared representations
between self and others: A social cognitive neuroscience view.
Trends in Cognitive Science,7, 527-533.
De Renzi, E., Cavalleri, F., & Facchini, S. (1996). Imitation and utiliza-
tion behavior. Journal of Neurology Neurosurgery and Psychiatry,61,
Derryberry, D., & Rothbart, M. K. (1988). Arousal, affect, and atten-
tion as components of temperament. Journal of Personality and
Social Psychology,55, 958-966.
De Waal, F. B. M. (1996). Good natured: The origins of right and wrong in
humans and other animals.Cambridge,MA:HarvardUniversity
Dijksterhuis, A., & Bargh, J. A. (2001). The perception-behavior
expressway: Automatic effects of social perception on social be-
havior. Advances in Experimental Social Psychology,33, 1-40.
Dimberg, U., Thunberg, M., & Elmehed, K. (2000). Unconscious
facial reactions to emotional facial expressions. Psychological Sci-
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 95
G. S. (1997). Reactive and proactive aggression in school children
and psychiatrically impaired assaultive youth. Journal of Abnormal
Dunbar, R. I. M. (1998). The social brain hypothesis. Evolutionary
Eidelberg, D., & Galaburda, A. M. (1984) Inferior parietal lobule.
Archives of Neurology,41, 843-852.
Eisenberg, N. (2000). Emotion, regulation, and moral development.
Annual Review in Psychology,51, 665-697.
Smith, M., et al. (1994). The relations of emotionality and regula-
tion to dispositional and situational empathy-related responding.
Journal of Personality and Social Psychology, 66, 776-797.
Eisenberg, N., Shea, C. L., Carlo, G., & Knight, G. (1991). Empathy
related responding and cognition: A “chicken and the egg”
dilemma.InW.Kurtines&J.Gewirtz(Eds.),Handbook of Moral
Behavior and Development, Vol. 2: Research (pp. 63-68). Hillsdale, NJ:
Eisenberg, N., & Strayer,J. (1987). Critical issues in the study of empa-
thy. In N. Eisenberg, & J. Strayer (Eds.), Empathy and its Develop-
ment (pp. 3-15). New York: Cambridge University Press.
Eisler, R., & Levine, D. S. (2002). Nurture, nature, and caring: We are
not prisoners of our genes. Brain and Mind,3, 9-52.
Ekman, P., & Davidson, R. J. (1993). Voluntary smiling changes
regional brain activity. Psychological Science,4, 342-345.
Ekman, P., Levenson, R. W., & Friesen, W. V. (1983). Autonomic ner-
vous system activity distinguishes among emotions. Science,221,
Eslinger, P. J. (1996). Conceptualizing, describing, and measuring
components of executive function: A summary. In G. R. Lyon &
N. A. Krasnegor (Eds.), Attention, memory and executive function
(pp. 367-395). Baltimore: Brookes.
Eslinger, P. J. (1998). Neurological and neuropsychological bases of
empathy. European Neurology,39, 193-199.
Fadiga, L., Fogassi, L., Pavesi, G., & Rizzolatti, G. (1995). Motor facili-
tation during action observation: A magnetic stimulation study.
Journal of Neurophysiology,73, 2608-2611.
Farrer,C., & Frith, C. D. (2002) . Experiencingoneself vs. anoth erper-
son as being the cause of an action: The neural correlates of the
experience of agency. NeuroImage,15, 596-603.
Jeannerod, M. (2003). Modulating the experience of agency: A
positron emission tomography study. NeuroImage,18, 324-333.
Farrow, T. F., Zheng, Y., Wilkinson, I. D., Spence, S. A., Deakin, J. F.,
Tarrier, N., et al. (2001). Investigating the functional anatomy of
empathy and forgiveness. NeuroReport,12, 2433-2438.
Feinburg, T. E. (2001). Altered egos: How the brain creates the self.New
York: Oxford University Press.
Field, T., Healy, B., Goldstein, S., & Guthertz, M. (1990). Behavior-
state matching and synchrony in mother-infant interactions of
nondepressed versus depressed dyads. Developmental Psychology,
Field, T. M., Woodson, R., Greenberg, R., & Cohen, D. (1982). Dis-
crimination and imitation of facial expressions by neonates. Sci-
Fine, C., Lumsden, J., & Blair, R. J. R. (2001). Dissociation between
theory of mind and executive functions in a patient with early left
amygdala damage. Brain,124, 287-298.
Fink, G. R., Markowitsch, H. J., Reinkemeier, M., Bruckbauer, T.,
Kessler,J., & Heiss, W.-D. (1996). Cerebral representation of one’s
own past: Neural networks involved in autobiographical memory.
The Journal of Neuroscience,16, 4275-4282.
Fink, G. R., Marshall, J. C., Halligan, P. W., Frith, C. D., Driver, J.,
Frackowiak, R. S. J., et al. (1999). The neural consequences of con-
flict between intention and the senses. Brain,122, 497-512.
Flavell, J. H. (1977). Cognitive development. Englewood Cliffs, NJ:
Flavell, J. H. (1999). Cognitive development: Children’s knowledge
about the mind. Annual Review of Psychology,50, 21-45.
Frackowiak, R. S., et al. (1995). Other minds in the brain: A func-
tional imaging study of theory of mind in story comprehension.
Freud, S. (1905). Jokes and their relation to the unconscious (Standard edi-
tion, 8). London: Hogarth.
Freud, S. (1921). Group psychology and the analysis of the ego (Standard
edition, 18). London: Hogarth.
Frith, U. (2001). Mind blindness and the brain in autism. Neuron,32,
Frith, U., & Frith, C. D. (2003). Development and neurophysiology of
mentalizing. Philosophical Transactions of the Royal Society, London B,
Fuster, J. M. (1989). The prefrontal cortex.NewYork:Raven.
Gallagher, H. L., & Frith, C. D. (2003). Functional imaging of theory
of mind. Trends in Cognitive Sciences,7, 77-83.
Frith, C. D. (2000). Reading the mind in cartoons and stories: An
fMRI study of theory of mind in verbal and nonverbal tasks.
Gallagher, S. (2000). Philosophical conceptions of the self: Implica-
tions for cognitive Science. Trends in Cognitive Science,4, 14-21.
Gallagher, S. (2001). The practice of mind: Theory, simulation, or
interaction? Journal of Consciousness Studies,8(5-7), 83-107.
Gallagher, S., & Meltzoff, A. N. (1996). The earliest sense of self and
others: Merleau-Ponty and recent developmental studies. Philo-
sophical Psychology,9, 211-233.
Gallese, V. (2001). The “shared manifold” hypothesis: From mirror
neurons to empathy. Journal of Consciousness Studies,8, 33-50.
Gallese, V., & Goldman, A.I. (1998). Mirror neurons and the simula-
tion theory of mind-reading. Trends in Cognitive Sciences,2, 493-
Gallup, G. G. (1998). Self-awareness and the evolution of social intel-
ligence. Behavioural Processes,42, 239-247.
Gergely, G. (2001). The obscure object of desire: Nearly, but clearly
not, like me: Contingency preference in normal children versus
children with autism. Bulletin of the Menninger Clinic,65, 411-426.
Gibson, J. J. (1966). The senses considered as perceptual systems.Boston:
Gibson, J. J. (1979). The ecological approach to visual perception.Boston:
Gill, K. L., & Calkins, S. D. (2003). Do aggressive/destructive toddlers
lack of concern for others? Behavioral and physiological indica-
tors of empathic responding in 2-year-old children. Development
and Psychopathology,15, 55-71
Gilovich, T., Kruger, J., & Medvec, V. H. (2002). The spotlight effect
revisited: Overestimating the manifest variability of our actions
and appearance. Journal of Experimental Social Psychology,38, 93-99.
Goel, V., & Dolan, R. (2001). The functional anatomy of humor: seg-
regating cognitive and affective components. Nature Neuroscience,
Goel, V., & Dolan, R. (2003). Explaining modulation of reasoning by
belief. Cognition,87, B11-B22.
Goldman, A. I. (1989). Interpretation psychologized. Mind and Lan-
Goldman, A. I. (1993). Ethics and cognitive science. Ethics,103, 337-
Goldman, A. I., & Sripada, C. S. (in press). Simulationist models of
face-based emotion recognition. Cognition.
Gopnik, A. (1993). How we know our minds: The illusion of first-
person knowledge of intentionality. Behavioral and Brain Sciences,
16, 1-15, 90-101.
Gordon, R. M. (1986). Folk psychology as simulation. Mind and Lan-
(1994). Cognitive flexibility and empathy after front al lobe lesion.
Neuropsychiatry, Neuropsychology, and Behavioral Neurology,7, 251-
96 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
Grattan, L. M., & Eslinger, P. J. (1989). Higher cognition and social
behavior: Changes in cognitive flexibility and empathy after cere-
bral lesions. Neuropsychology,3, 175-185.
Greene, J. D., Sommerville, R. B., Nystrom, L. E., Darley, J. M., &
Cohen, J. D. (2001). An fMRI investigation of emotional engage-
ment in moral judgment. Science,293, 2105-2108.
Grèzes, J., & Decety, J. (2001). Functional anatomy of execution, men-
tal simulation, observation, and verb generation of actions: A
meta-analysis. Human Brain Mapping,12, 1-19.
Grèzes, J., Frith, C. D., & Passingham, R. E. (2004). Infering false
beliefs from the actions of oneself and others: An fMRI study.
Griffin, D. W., Dunning, D., & Ross, L. (1990). The role of construal
processes in overconfident predictions about the self and others.
Journal of Personality and Social Psychology,59, 1128-1139.
Gross, J. J. (2001). Emotion regulation in adulthood: Timing is every-
thing. Current Directions in Psychological Science,10, 214-219.
Gusnard, D. A., Akbudak, E., Shulman, G. L., & Raichle, M. E. (2001).
Medial prefrontal cortex and self-referential mental activity: Rela-
tion to a default mode of brain function. Proceedings of the National
Academy of Sciences, USA,98, 4259-4264.
Hamilton, W. D. (1964). The genetical evolution of social behavior.
Journal of Theoretical Biology,7, 17-52.
Happé, F. (2003). Theory of mind and the self. Annals of New York
Academy of Science,1001, 134-144.
Happé, F., Ehlers, S., Fletcher, P., Frith, U., Johansson, M., Gillberg,
C., et al. (1996). Theory of mind in the brain. Evidence from a
PET scan study of Asperger syndrome. NeuroReport,8, 197-201.
Hare, R. D. (1993). Without conscience: The disturbing world of the psycho-
paths among us.NewYork:Simon&Schuster.
Rizzolatti, G. (1998). Activation of human primary motor cortex
during action observation: A neuromagnetic study. Proceedings
National Academy of Science, USA,95, 15061-15065.
Harris, P. (1991). The work of imagination. In A. Whiten (Ed.), Natu-
ral Theories of Mind (pp. 283-304). Oxford: Blackwell.
Harris, P. L. (2000). Understanding emotion. In M. Lewi and J. M.
Haviland-Jones (Eds.), Handbook of emotions (pp. 281-292). New
Hatfield, E., Cacioppo, J., & Rapson, R. (1994). Emotional contagion.
New York: Cambridge University Press.
Hobson, R. P. (1989). On sharing experiences. Development and Psycho-
Hobson, R. P. (2002). The cradle of thought. London: Macmillan.
Hobson, R., & Lee, A. (1999). Imitation and identification in autism.
Journal of Child Psychology and Psychiatry,10, 649-659.
Hodges, S. D., & Klein, K. J. K. (2001). Regulating the costs of empa-
thy: The price of being human. Journal of Socio-Economics,30, 437-
Hodges, S. D., & Wegner, D. M. (1997). Automatic and controlled
empathy. In W. Ickes (Ed.), Empathic accuracy (pp. 311-339). New
Hoffman, M. L. (1981). Is altruism part of human nature? Journal of
Personality and Social Psychology,40, 121-137.
Hoffman, M.L. (2000). Empathy and Moral Development.NewYork:
Cambridge University Press.
Hogan, R. (1969). Development of an empathy scale. Journal of Con-
sulting and Clinical Psychology,33, 307-316.
Hommel, B., Müsseler, J., Aschersleben, G., & Prinz, W. (2001). The
theory of event coding: A framework for perception and action.
Behavioral and Brain Sciences,24, 849-878.
Humphrey, N. (1990). The uses of consciousness. In J. Brockman
(Ed.), Speculations: The reality club (pp. 67-84). New York: Prentice
Humphrey, N. (2002). The mind made flesh.NewYork:OxfordUniver
Hutchison, W. D., Davis, K. D., & Lozano, A. M. (1999). Pain-related
neurons in the human cingulate cortex. Nature Neuroscience,2,
Iacoboni, M., Woods, R. P., Brass, M., Bekkering, H., Mazziotta, J. C.,
& Rizzolatti, G. (1999). Cortical mechanisms of human imitation.
Ickes, W. (1997). Empathic accuracy.NewYork:Guilford.
Ickes, W. (2003). Everyday mind reading.NewYork:Prometheus.
M., et al. (2000). On the difficulty of distinguishing “good” and
“poor” perceivers: A social relations analysis of empathic accuracy
data. Personal Relationships,7, 219-234.
Ickes, W., Stinson, L., Bissonnette, V., & Garcia, S. (1990). Imple-
menting and using the dyadic interaction paradigm. In
C. Hendricks & M. Clark (Eds.), Review of personality and social psy-
chology: Research methods in personality and social psychology (Vol. 11,
pp. 16-44). Newbury Park, CA: Sage.
Jackson, P. L., & Decety, J. (2004). Motor cognition: A new paradigm
to study self other interactions. Current Opinion in Neurobiology,14,
Jackson, P. L., Meltzoff, A. N., & Decety, J. (2004). Perceiving others in
painful situations activates the affective pain neural network. Proceed-
ings of the Cognitive Neuroscience Society, A57, April 18-20, 2004,
Jacobs, D. H., Shuren, J., Bowers, D., & Heilman, K. (1995). Emo-
tional facial imagery, perception, and expression in Parkinson’s
disease. Neurology,45, 1696-1702.
Jeannerod, M. (1994). The representing brain. Neural correlates of
motor intention and imagery. Behavioral and Brain Sciences,17,
Jeannerod, M. (1999). To act or not to act: Perspective on the repre-
sentation of actions. Quaterly Journal of Experimental Psycholog y,52A,
& Prigatano, G. P. (2002). Neural correlates of self-reflection.
Kagan, J. (1998). Biology and the child. In W. Damon & N. Eisenberg
(Eds.), Handbook of child personality, Volume 3: Social, emotional, and
personality development (pp. 177-236). New York: Wiley.
Keenan, J. P., Gallup, G. G., & Falk, D. (2003). The face in the mirror: The
search for the origins of consciousness. New York: HarperCollins.
Keenan, J. P., McCutcheon, N. B., & Pascual-Leone, A. (2001). Func-
tional magnetic resonance imaging and event-related potentials
suggest right prefrontal activation for self-related processing.
Brain and Cognition,47, 87-91.
Keenan, J. P., Nelson, A., O’Connor, M., & Pascual-Leone, A. (2001).
Self-recognition and the right hemisphere. Nature,409, 305.
Keysar, B., Lin, S., & Barr, D. J. (2003). Limits on theory of mind in
adults. Cognition,89, 25-41.
Knoblich, G., & Flach, R. (2003). Action identity: Evidence from self-
recognition, prediction, and coordination. Consciousness and Cog-
Kruger, D. J. (2003). Evolution and altruism: Combining psychologi-
cal mediators with naturally selected tendencies. Evolution and
Human Behavior,24, 118-125.
Kugiumutzakis, G. (1999). Genesis and development of early infant
mimesis to facial and vocal models. In Nadel J. & Butterworth G.
(Eds.), Imitation in infancy (pp. 36-59). Cambridge, UK: Cam-
bridge University Press.
Legerstee, M. (1991). The role of person and object in eliciting early
imitation. Journal of Experimental Child Psychology,51, 423-433.
Levenson, R. W., Ekman, P., & Friesen, W. V. (1990). Voluntary facial
action generates emotion-specific autonomic nervous system
activity. Psychophysiology,27, 363-384.
Levenson, R. W., & Ruef, A. M. (1992). Empathy: A physiological sub-
strate. Journal of Personality and Social Psychology,63, 234-246.
Beaudoin, G., et al. (2003). Neural circuitry underlying voluntary
suppression of sadness. Biological Psychiatry,53, 502-510.
J. P., et al. (1998). Episodic memory and the self in a case of iso-
lated retrograde amnesia. Brain,121, 1951-1973.
Decety, Jackson / FUNCTIONAL ARCHITECTURE OF HUMAN EMPATHY 97
(1999). Ventral frontal contribution to self-regulation: Conver-
gence of episodic memory and inhibition. Neurocase,5, 263-275.
Lewis, M. (1999). Social cognition and the self. In P. Rochat (Ed.),
Early social cognition (pp. 81-98). New York: Lawrence Erlbaum.
Lewis, M., Sullivan, M. W., Stanger, C., & Weiss, M. (1989). Self-
development and self-conscious emotions. Child Development,60,
Lipps, T. (1903). Grundlegung der Ästhetik. Hamburg/Leipzig, Ger-
many: Leopold Voss.
Lough, S., Gregory, C., & Hodges, J. R. (2001). Dissociation of social
cognition and executive function in frontal variant frontotem-
poral dementia. Neurocase,7, 123-130.
Malatesta, C. Z., & Haviland, J. M. (1982). Learning display rules: The
socialization of emotional expression in infancy. Child Develop-
Mandler, G. (2002). Consciousness recovered: Psychological functions and
origins of conscious thought.Amsterdam:JohnBenjamins.
Marangoni, G., Garcia, S., & Ickes, W. (1995). Empathic accuracy in
clinically-relevant setting. Journal of Personality and Social Psychol-
Marks, G., & Miller, N. (1987). Ten years of research on the false con-
sensus effect: An empirical and theoretical review. Psychological
Martin, G. B., & Clark, R. D. (1987). Distress crying in neonates: Spe-
cies and peer specificity. Developmental Psychology,18,3-9.
Masterman, D. L., & Cummings, J. L. (1997). Frontal-subcortical cir-
cuits: The anatomic basis of executive, social and motivated
behaviors. Journal of Psychopharmacology,11, 107-14.
McArthur, L. Z., & Baron, R. (1983). Toward an ecological theory of
social perception. Psychological Review,90, 215-238.
Mealey, L. (1995). The sociobiology of sociopathy: An integrated evo-
lutionary model. Behavioral and Brain Sciences,18, 523-599.
Mehrabian, A., & Epstein, N. (1972). A measure of emotional empa-
thy. Journal of Personality,40, 525-543.
Meltzoff, A. N. (1990). Foundations for developing a concept of self.
In D. Cicchetti & M. Beeghly (Eds.), The self in transition: Infancy to
childhood (pp. 139-164). Chicago: Chicago University Press.
Meltzoff, A. N. (2002). Imitation as a mechanism of social cognition:
Origins of empathy, theory of mind, and the representation of
action.InU.Goswami(Ed.),Blackwell handbook of childhood cogni-
tive development (pp. 6-25). Oxford: Blackwell.
Meltzoff, A. N., & Decety, J. (2003). What imitation tells us about
social cognition: a rapprochement between developmental psy-
chology and cognitive neuroscience. Philosophical Transactions of
the Royal Society, London: Biological Sciences,358, 491-500.
Meltzoff, A. N., & Gopnik, A. (1993). The role of imitation in under-
standing persons and developing theories of mind. In S. Baron-
Cohen & H. Tager-Flusberg (Eds.), Understanding other minds: Per-
spectives from autism (pp. 335-366). Oxford: Oxford University
Meltzoff, A. N., & Moore, M. K. (1994). Imitation, memory, and
the representation of persons. Infant Behavior and Development,17,
Meltzoff, A. N., & Moore, M. K. (1997). Explaining facial imitation: A
theoretical model. Early Development and Parenting,6, 179-192.
Miller, E. K., & Cohen, J.D. (2001). An integrative theory of pre-
frontal cortex function. Annual Review of Neurosciences,24, 167-
Morgan, A. B., & Lilienfeld, S. O. (2000). A meta-analytic review of
the relation between antisocial behavior and neuropsychological
measures of executive function. Clinical Psychology Review,20, 113-
Morrison, I., Lloyd, D., di Pellegrino, G., & Roberts, N. (in press).
Vicarious responses to pain in anterior cingulate cortex: Is empa-
thy a multisensory. Cognitive, Affective, and Behavioral Neuroscience.
Nadel, J., & Baudonnière, P. M. (1982). The social function of recip-
rocal imitation in 2-year-old peers. International Journal of Behav-
ioral Development,5, 95-109.
Neisser, U. (1991). Two perceptually given aspects of the self and
their development. Developmental Review,11, 197-209.
Nelson, D. W., & Baumgarte, R. (2004). Cross-cultural misunder-
standings reduce empathic responding. Journal of Applied Social
Nickerson, R. S. (1999). How we know and sometimes misjudge what
others know: Imputing one’s own knowledge to others. Psychologi-
cal Bulletin,126, 737-759.
Nigg, J. T. (2001). Is ADHD a disinhibitory disorder? Psychological Bul-
Ochsner, K. N. (2004). Current directions in social cognitive neuro-
science. Current Opinion in Neurobiology,14, 254-258.
Ochsner, K. N., Bunge, S. A., Gross, J. J., & Gabrieli, J. D. E. (2002).
Rethinking feelings: An fMRI study of the cognitive regulation of
emotion. Journal of Cognitive Neuroscience,14, 1215-1229.
Ochsner, K. N., & Lieberman, M. D. (2001). The emergence of social
cognitive neuroscience. American Psychologist,56, 717-734.
O’Donovan, M. J. (1999). The origin of spontaneous activity in devel-
oping networks of the vertebrate nervous system. Current Opinion
in Neurobiology,9, 94-104.
Osterling, J., & Dawson, G. (1994). Early recognition of children with
autism: A study of first birthday home videotapes. Journal of Autism
and Developmental Disorders,24, 247-257.
Passingham, R. E. (1998). The specializations of the human neocor-
tex. In A. D. Milner (Ed.), Comparative neuropsychology (pp. 271-
298). Oxford: Oxford University Press.
Passingham, R. E., Stephan, K., & Kötter, R. (2002). The anatomical
basis of functional localization in the cortex. Nature Reviews in Neu-
Perner, J., & Lang, B. (1999). Development of theory of mind and
executive control. Trends in Cognitive Science,3, 337-344.
Perner,J., & Lopez, A. (1997). Ch ildren’s understanding of belief and
disconfirming visual evidence. Child Development,12, 367-380.
Plutchik, R. (1987). Evolutionary bases of empathy. In N. Eisenberg,
&J.Strayer(Eds.),Empathy and its development (pp. 38-46). New
York: Cambridge University Press.
Poe, E. A. (1990). Selected works (Unabridged edition, pp. 465-477).
New York: Gramercy. (Original work published in 1845)
Povinelli, D. J. (2001). Folk physics for apes.NewYork:OxfordUniver-
Povinelli, D. J., Bering, J. M., & Giambrone, S. (2000). Toward a sci-
ence of other minds: Escaping the argument by analogy. Cognitive
Preston, S. D., & de Waal, F. B. M. (2002). Empathy: Its ultimate and
proximate bases. Behavioral and Brain Sciences,25, 1-72.
Price, B. H., Daffner, K. R., Stowe, R. M., & Mesulam, M. (1990). The
compartmental learning disabilities of early frontal lobe damage.
Prinz, W. (1997). Perception and action planning. European Journal of
Cognitive Psychology,9, 129-154.
Buchsbaum, M. S. (1998). Reduced prefrontal and increased
subcortical brain functioning assessed using positron emission
tomography in predatory and affective murderers. Behavioral Sci-
Ramachandran, V. S., & Rogers-Ramachandran, D. (1996). Denial of
disabilities in anosognosia. Nature,382, 501.
Ramnani, N., & Miall, R. (2004). A system in the human brain for pre-
dicting the actions of others? Nature Neuroscience,7, 85-90.
Reaux, J. E., Theall, L. A., & Povinelli, D. J. (1999). A longitudinal
investigation of chimpanzees’ understanding of visual percep-
tion. Child Development,70, 275-290.
Reddy, V. (2003). On being the object of attention: Implications for
self-other consciousness. Trends in Cognitive Sciences,7, 397-402.
Reik, T. (1949). Character analysis. New-York: Farrar, Strauss & Giroux.
R. J., Friston, K. J., et al. (1997). Neuroanatomical correlates of
externally and internally generated emotion. American Journal of
98 BEHAVIORAL AND COGNITIVE NEUROSCIENCE REVIEWS
Repacholi, B. M., & Gopnik, A. (1997). Early reasoning about
desires: Evidence from 14-18-months-olds. Developmental Psychol-
Rizzolatti, G., Fogassi, L., & Gallese, V. (2001). Neurophysiological
mechanisms underlying the understanding and the imitation of
action. Nature Review Neuroscience,2, 661-670.
Rochat, P. (2001). Theinfant’