Human empathy

2006 ? 3 ? 25 ?
11
? Special Lecture
Human empathy
Jean Decety
Abstract : The psychological construct of empathy refers to an emotional response that em-
anates from the emotional state of another individual without loosing sight of whose feelings be-
long to whom. This response is contingent on cognitive as well as emotional factors. Empathy
involves not only the affective experience of the other person’s actual or inferred emotional
state but also some minimal recognition and understanding of another’s emotional state. Draw-
ing on cognitive neuroscience and neuropsychological data, I propose that empathy involves
parallel and distributed processing in a number of dissociable computational mechanisms.
Shared neural representations, self-awareness, mental flexibility and emotion regulation consti-
tute the basic macro-components of empathy, which are underpinned by specific and interact-
ing neural systems. Furthermore, this model of empathy is consistent with the view that social
cognition draws on both domain-general mechanisms and domain-specific embodied represen-
tations.
Japanese Journal of Neuropsychology 22 ; 11-33, 2006
Key words ?social cognition, shared representations, emotion regulation, self-awareness, empathy,
cognitive neuroscience
The ability to perceive and process the infor-
ability may be used for both helpful and hurtful
mation about the emotional states of others is
purposes). Moreover, the social and emotional sit-
considered a cornerstone of social sensibility and
uations eliciting empathy can become quite com-
adaptation. Among the various forms of emotion-
plex depending on the feelings experienced by
al connections with others, empathy has re-
the observed and the relationship of the target
ceived lots of attention from philosophers and
to the observer (Feshbach, 1997). This capacity
psychologists, and more recently from social
to understand others and experience their feel-
neuroscientists. The construct of empathy de-
ings in relation to oneself illustrates the social na-
notes, at a phenomenological level of description,
ture of the self, its inherent intersubjectivity.
a sense of similarity between the feelings one ex-
The goal of this paper is to propose a new
periences and those expressed by others. It can
model of empathy that articulates data from so-
be conceived of as an interaction between any
cial psychology and cognitive neuroscience (in-
two individuals, with one experiencing and shar-
cluding neuropsychology). Bridging social psy-
ing the feeling of the other. This sharing of feel-
chology and cognitive neuroscience provides im-
ings does not necessarily imply that one will act
portant guidelines for investigating the neural
or even feel impelled to act in a supportive or
processes underlying empathy. On the other
sympathetic way (empathy’s paradox is that this
hand, cognitive neuroscience may help disam-
Prof. Jean Decety, Social Cognitive Neuroscience, The University of Chicago
????
11

---

12
??????? 22 ?? 1 ?
biguate competing social theories. To that end,
contagion (see de Waal & Thompson, 2005),
instead of addressing each of these research do-
which mediate emotional sharing between differ-
mains separately, theoretical notions and find-
ent conspecifics. But humans may be unique in
ings from these different approaches will be
their ability to intentionally “feel for” and act
marshaled with the guidance of an overarching
on behalf of other people whose experiences
conceptual framework. This framework consid-
may differ greatly from their own (Batson, 1991 ;
ers that empathy involves parallel and distrib-
Decety & Hodges 2006 ; Davis, 1994). In addition,
uted processing in a number of dissociable com-
social neuroscience research has shown that em-
putational mechanisms. Shared neural circuits,
pathic concern is often associated with prosocial
self-awareness, mental flexibility and emotion
behaviors such as helping a kin, and has been
regulation constitute the basic macro-compo-
considered as a chief enabling process for altru-
nents of empathy, which are mediated by specif-
ism (Batson, 1991). Note that Wilson (1998) sug-
ic neural systems, including aspects of the pre-
gested that empathic helping behavior has
frontal cortex, the anterior insula and fronto-
evolved because of its contribution to genetic fit-
parietal networks. Consequently, damage to each
ness (kin selection). In humans and other mam-
of these components may lead to an alteration of
mals, an impulse to care for offspring is almost
empathic behavior, and produce selective social
certainly genetically hard-wired. It is however
disorders depending on which aspect is disrupt-
far less clear that an impulse to care for siblings,
ed.
more remote kin, and similar non-kin is geneti-
cally hard-wired (Batson, 2006). The emergence
Evolutionary origins of empathy
of altruism, of empathizing with and caring for
Much of our social behavior is similar to that
those who are not kin is thus not easily ex-
of other primates and other mammals. However,
plained within the framework of neo-Darwinian
it seems evident from the descriptions of com-
theories of natural selection. Social learning ex-
parative psychologists and ethologists that some
planations of kinship patterns in human helping
behaviors homologous to empathy can be ob-
behavior are thus highly plausible. However, one
served in other species (e.g., Plutchik, 1987). For
of the most striking aspects of human empathy
de Waal (1996) empathy is not an all-or-nothing
is that it can be felt for virtually any target -
phenomenon, and many forms of empathy exist
even targets of a different species. In addition, as
between the extremes of mere agitation at the
emphasized by Harris (2000), humans, unlike
distress of another and full understanding of
other primates, can put their emotions into
their predicament. Many other comparative psy-
words, allowing them not only to express emo-
chologists, however, view empathy as a kind of
tion, but to report on current, as well as past
induction process by which emotions, both posi-
emotions. These reports provide an opportunity
tive and negative, are shared, and by which the
to share, explain and regulate emotional experi-
probabilities of similar behavior are increased in
ence with others to an extent that is not found in
the participants. In my view, this is not a suffi-
any other species. Notably, conversation helps to
cient mechanism to account for the full-blown
develop empathy, for it is often here that one
ability of human empathy. This does not mean
learns of shared experiences and feelings. Impor-
that some aspects of empathy are not present in
tantly, this self-reflexive capability (which in-
other species, like motor mimicry and emotion
cludes emotion regulation processes) may be an
????
12

---

2006 ? 3 ? 25 ?
13
important difference between humans and other
particular terminology that is used, there is
animals (Povinelli, 2001).
broad agreement among scholars on three pri-
Overall, this evolutionary conceptual view is
mary aspects : 1) an affective response to anoth-
compatible with the hypothesis that advanced
er person, which often, but not always, entails
levels of social cognition may have arisen as an
sharing that person’s emotional state, 2) a cogni-
emergent property of powerful executive func-
tive capacity to take the perspective of the other
tioning assisted by the representational proper-
person, and 3) some self-regulatory and monitor-
ties of language (Barrett, Henzi & Dunbar, 2003).
ing mechanisms that modulate inner states (e.g.,
However, these higher levels operate upon pre-
Batson, 1991 ; Davis, 1996 ; Decety & Hodges,
vious levels of organization, and should not be
2006 ; Eisenberg, 2000 ; Hodges & Wegner, 1997 ;
seen as independent or conflicting with one an-
Preston & de Waal, 2002). According to Ickes
other. Evolution has constructed layers of in-
(1997), empathy is a complex form of psychologi-
creasing complexity, from non-representational
cal inference in which observation, memory,
to representational and meta-representational
knowledge, and reasoning are combined to yield
mechanisms, which need to be taken into ac-
insights into the thoughts and feelings of others.
count for a full understanding of human empa-
As such, empathy involves not only the affective
thy. This distributed nature of social processing
experience of the other person’s actual or in-
poses a challenge to understanding the neurolo-
ferred emotional state, but also some minimal
gy of social cognition in general and empathy in
recognition and understanding of another’s emo-
particular. In addition, social psychological con-
tional state (or most likely emotional state) . This
cepts, such as empathy, do not necessarily corre-
definition captures the multidimensional nature
spond to neurological processes. Relations be-
of empathy and makes explicit reference to
tween psychological and biological processes
some minimal mentalizing capacity. This latter
cannot be comprehended fully by investigations
concept refers to the broad social-cognitive abili-
at a single level of organization. Therefore multi-
ty used by humans to explain and predict their
level research is necessary to form bridges
own behavior and that of others by attributing
among disciplines, and ultimately achieve a truly
to them independent mental states, such as be-
interdisciplinary social neuroscience (Cacioppo,
liefs, desires, emotions or intentions (Flavell,
Tassinary & Berntson, 2000). Given the fact that
1999).
empathy is a psychological construct, the first
Of all the sources from which one can draw in-
step is to break it down into its component neu-
sight as to the constituents of human empathy,
rocognitive functions.
psychotherapeutic schools provide the most in-
teresting experience-related knowledge. Indeed,
Breaking down empathy into its constitutive
empathy is appreciated to play a central role in
components
psychotherapies, since almost all of them involve
For many psychologists, empathy implies at
intersubjective communication between individ-
least three different processes : feeling what an-
uals in order for the clinician to understand
other person is feeling ; knowing what another
his/her client sufficiently to proceed along a
person is feeling ; and having the intention to re-
treatment path (Bohart & Greenberg, 1997). Al-
spond compassionately to another person’s dis-
though not a major construct for psychoanalysis,
tress (Thompson, 2001). Yet, regardless of the
Freud (1921) wrote that empathy was indispens-
????
13

---

14
??????? 22 ?? 1 ?
able when it came to taking a position regarding
parietal lobule and the anterior insula, espe-
another person’s mental life, and considered it as
cially on the right side.
the process that plays the largest part in our un-
3. Mental flexibility to adopt the subjective
derstanding of what is inherently foreign to our
perspective of the other. This aspect heavily
ego in other people. A number of analysts have
draws on executive resources of the pre-
pointed out that empathy involves resonating
frontal cortex.
with the other’s unconscious affect and experi-
4. Regulatory processes including emotion
encing his/her experience with him/her while
reappraisal, which depend upon interaction
the empathizer maintains the integrity of
between prefrontal and anterior cingulate
his/her self intact. For instance, Basch (1983)
systems and subcortical emotion-generation
speculated that, because their respective auto-
systems.
nomic nervous systems are genetically pro-
In this view, none of these components can ac-
grammed to respond in like fashion, a given af-
count solely for the potential of human empathy.
fective expression by a member of a particular
The four components are intertwined and dy-
species tends to recruit a similar response in
namically interact with one another to produce
other members of that species.“This is done
the subjective experience of human empathy.
through the promotion of an unconscious auto-
For instance, sharing emotion without self-
nomic imitation of the sender’s bodily state and
awareness corresponds to the phenomenon of
facial expression by the receiver. It generates in
emotional contagion, which takes the form of
the receiver the autonomic response associated
‘total identification without discrimination be-
with that bodily state and facial expression,
tween one’s feelings and those of the other’ (de
which is to say, the receiver experiences an af-
Waal, 1996).
fect identical with that of the sender (p. 108).”
Furthermore, this model of empathy involves
Such a view subsequently received empirical
both bottom-up and top-down information pro-
validation by a series of studies conducted by
cessing (see Fig. 1). Furthermore, it combines
Levenson and Ruef (1992). They found evidence
representational aspects, i.e., memories that are
that a perceiver’s accuracy in inferring a target’s
localized in distributed neural networks that en-
negative emotional states was related to the de-
code information and, when temporarily activat-
gree of physiological synchrony between the
ed, enable access to this stored information, as
perceiver and the target. In other words, when
well as processes, i.e., computational procedures
two people feel similar emotions, they more ac-
that are localized and are independent of the na-
curately perceive each other’s intentions and
ture or modality of the stimulus that is being
motivations.
processed.
The model proposed here suggests that four
Like many emotion-related processes, some
major functional components dynamically inter-
components involved in empathy occur implicit-
act to produce the experience of empathy :
ly, without awareness, in a bottom-up fashion.
1. Affective sharing between the self and the
This is the case of the emotion-sharing and
other, based on the shared circuits between
motor mimicry aspect. Other components re-
perception-action, mediated by fronto-pari-
quire explicit top-down processing, such as per-
etal networks.
spective-taking, representing our own thoughts
2. Self-awareness, mediated by the inferior
and feelings as well as those of others, and also
????
14

---

2006 ? 3 ? 25 ?
15
Fig. 1
Schematic representation of bottom-up (i.e., direct matching between
perception and action), and top-down (i.e., regulation and control) infor-
mation processing involved in human empathy. These two levels of pro-
cessing are interrelated. The low level, which is automatically activated
(unless inhibited) by perceptual input, accounts for the implicit recogni-
tion that others are like us. The metalevel is continuously updated by
bottom-up information, and in return controls the lower level by provid-
ing top-down input. Thus metacognitive regulation, through executive
functions modulates low levels and adds flexibility, making the individual
less dependent on external cues.
some aspects of emotion regulation.
maintenance of social relationships.
Social psychological research shows that hu-
Affective sharing between self and other
mans mimic unintentionally and unconsciously a
In addition to their survival values for the or-
wide range of behaviors, such as accents, tone of
ganism, emotional expression and perception are
voice, rate of speech, posture and mannerisms,
an integral part of social interaction (Schulkin,
as well as moods (e.g., Chatrand & Bargh, 1999) .
2004). Bodily expressions constitute an external,
This tendency to automatically mimic and syn-
perceivable indication of people’s intentions and
chronize one’s own emotional behavior with oth-
emotions. At one level, emotional expressions are
ers, also known as the phenomenon of emotion
governed by rules and can be elicited by simple
contagion, facilitate the smoothness of social in-
stimuli, as in the example of disgust in the pres-
teraction and may even foster empathy (Hatfield,
ence of bitter taste. However, humans and other
Cacioppo & Rapson, 1994). For instance, a study
animals also use bodily expressions to communi-
demonstrated that participants who had been
cate various type of information to members of
mimicked by the experimenter were more help-
their own species. Understanding other people’s
ful and generous toward other people than non-
emotional signals has clear adaptive advantages
mimicked participants (Van Baaren et al., 2004) .
and is especially important in the formation and
They also found that these beneficial conse-
????
15

---

16
??????? 22 ?? 1 ?
quences of mimicry were not restricted to be-
lobule, the supplementary motor area and the
havior directed toward the mimicker, but includ-
cerebellum. In addition, a number of neuroimag-
ed behavior directed toward people not directly
ing studies have shown that similar brain areas,
involved in the mimicry situation.
pertaining to the same network are reliably acti-
This automatic mapping between self and
vated during imagining one’s own action, imagin-
other is supported by considerable empirical lit-
ing another’s action, and imitating actions per-
erature in the domain of perception and action,
formed by a model (Decety & Chaminade, 2003a
which has been marshaled under the common-
; Decety & Grèzes, 2006). Such shared circuits re-
coding theory (Prinz, 1997). Its core assumption
flect an automatic transformation of what anoth-
is that actions are coded in terms of the perceiv-
er conspecifics do into the neural representation
able effects (i.e., the distal perceptual events)
of one’s own actions.
they should generate. This theory also states
The perception-action mechanism accounts (at
that perception of an action activates action rep-
least partly) for emotion sharing and empathy, as
resentations to the degree that the perceived
suggested by Preston and de Waal (2002). This
and the represented action are similar (Wilson &
model posits that perception of emotion activates
Knoblich, 2005). Furthermore, these representa-
in the observer the neural mechanisms that are
tions may be shared between individuals. In-
responsible for the generation of similar emotion.
deed, the meaning of a given object, action, or so-
This mechanism was also proposed to account
cial situation may be common to several people
for emotion contagion. Indeed, Hatfield, Cacioppo
and activate corresponding distributed patterns
and Rapson (1994) suggested that people may
of neural activation in their respective brains
catch the emotions of others as a result of affer-
(Decety & Sommerville, 2003 ; Jeannerod, 1999).
ent feedback generated by elementary motor
This sharing explains how we come to under-
mimicry of others’ expressive behavior, which
stand each other, that is the isomorphism be-
produces a simultaneous matching emotional ex-
tween action representations allows the individ-
perience. For example, while watching someone
ual to implicitly know the goal of others with the
smile, the observer activates the same facial
use of her/his own action representation system.
muscles involved in producing a smile at a sub-
In neuroscience, evidence for this percep-
threshold level and this would create the corre-
tion/action coupling ranges from electrophysio-
sponding feeling of happiness in the observer. In-
logical recordings in monkeys in which mirror
deed, viewing facial expressions triggers expres-
neurons in the ventral premotor and posterior
sions on one’s own face, even in the absence of
parietal cortices fire both during goal-directed
conscious recognition of the stimulus (e.g., Dim-
actions and observation of the same actions per-
berg, Thunberg, & Elmehed, 2000 ; Wallbott,
formed by another individual (Rizzolatti, Fogassi,
1991). Interestingly, de Gelder and colleagues
& Gallese, 2001), to functional neuroimaging ex-
demonstrated that observing fearful body ex-
periments in humans which indicate that the
pressions not only produces increased activity in
neural circuits involved in action execution over-
brain areas associated with emotional processes
lap with those activated when actions are ob-
but also in areas linked with representation of
served (Blakemore & Decety, 2001). This shared
action and movement. These results demon-
neural network for action production and obser-
strate that the mechanism of fear contagion au-
vation includes the premotor cortex, the parietal
tomatically prepares the brain for action.
????
16

---

2006 ? 3 ? 25 ?
17
Making a facial expression generates changes
encing the emotion signaled in the stimulus
in the autonomic nervous system and is associat-
(Adolphs, 2002).
ed with feeling the corresponding emotion. In a
Moreover, there are several dramatic case
series of experiments, Levenson, Ekman and
studies that support the idea that the same neur-
Friesen (1990) instructed participants to produce
al systems are involved both in the recognition
facial configurations for anger, disgust, fear, hap-
and in the expression of specific emotion. For in-
piness, sadness, and surprise while heart rate,
stance, Adolphs and collaborators (1995) investi-
skin conductance, finger temperature, and so-
gated S.M., a 30-year old patient, whose amyg-
matic activity were monitored. They found that
dala was bilaterally destructed by a metabolic
such a voluntary facial activity produce signifi-
disorder. Consistent with the prominent role of
cant levels of subjective experience of the associ-
the amygdala in mediating certain negatively va-
ated emotions as well as specific and reliable au-
lenced emotions such as fear, S.M. was found to
tonomic measures. Recently an fMRI experiment
be impaired at both the recognition of fear from
confirmed and extended these findings by show-
facial expressions as well as in the phenomeno-
ing that when participants are required to ob-
logical experience of fear. Another case, N.M,
serve or to imitate facial expressions of various
who suffered from bilateral amygdala damage
emotions, increased neurodynamic activity is de-
and left thalamic lesion was found to be impaired
tected in the superior temporal sulcus, the ante-
in at recognizing fear from facial expressions
rior insula and the amygdala, as well as areas of
and exhibited an equivalent impairment of fear
the premotor cortex corresponding to the facial
recognition from body postures and emotional
representation (Carr et al., 2003).
sounds (Sprengelmeyer et al., 1999). The patient
The finding of paired deficits between emotion
also reported reduced anger and fear in his
production and emotion recognition also pro-
everyday experience of emotion. There is also
vides strong arguments in favor of the percep-
evidence for paired deficits for the emotion of
tion-action matching model. A lesion study car-
disgust. Calder and colleagues (2000) described
ried out with a large number of neurological pa-
patient N.K., with left insula and putamen dam-
tients by Adolphs and colleagues (2000) found
age who was selectively impaired at recognizing
that damage within the right somatosensory re-
social signals of disgust from multiple modalities
lated cortices (including primary and secondary
(facial expressions, non-verbal sounds, and emo-
somatosensory cortices, insula and anterior
tional prosody), and who was less disgusted than
supramarginal gyrus) impaired the judgment of
controls by disgust-provoking scenarios. Further
other people’s emotional states from viewing
and direct support for a specific role of the left
their face. Another study with brain-damaged
insula in both the recognition and the experience
patients indicated that recognizing emotions
of disgust was recently provided by an fMRI
from prosody draws on the right fronto-parietal
study in which participants inhaled odorants
cortex (Adolphs, Damasio, & Tranel, 2002).
producing a strong feeling of disgust, and in an-
These findings strongly support the hypothesis
other condition, watched video clips showing the
that the recognition of emotion in others re-
facial expression of disgust. It was found that ob-
quires the perceiver to reconstruct images of so-
serving such facial expressions and feelings of
matic and motoric components that would nor-
disgust activated the same sites in the anterior
mally be associated with producing and experi-
insula and anterior cingulate cortex (Wicker et
????
17

---

18
??????? 22 ?? 1 ?
al., 2003).
visual stimulation was restricted to the right in-
The expression of pain provides a crucial sig-
ferior Brodmann’s area 24b. In contrast, the pri-
nal, which can motivate caring behaviors in oth-
mary somatosensory cortex showed significant
ers. It is thus an ecologically valid way to inves-
activations in response to noxious tactile, but not
tigate the neural systems involved in empathy
visual, stimuli. The different response patterns in
and evaluate to what extent there is an overlap
the two areas are consistent with the ACC’s role
between the response to self-experienced pain
in coding the motivational-affective dimension of
and pain perceived in others. It is already well
pain, which is associated with the preparation of
known that a restricted number of neural re-
behavioral responses to aversive events. These
gions are involved in the processing of painful
findings are supported by an fMRI study con-
stimuli, including the anterior cingulate cortex,
ducted by Jackson, Meltzoff and Decety (2005a)
the insula, the somatosensory cortex, the periac-
in which participants were shown still pho-
queductal gray, the thalamus, and the ventral
tographs depicting right hands and feet in
prefrontal cortex. Further, these regions are dif-
painful or neutral everyday-life situations, and
ferentially involved in the sensory, and affective
asked to imagine the level of pain that these sit-
and motivational aspects of pain processing. In-
uations would produce. Significant activation in
terestingly, a micro-electrode exploration in neu-
regions involved in the affective aspect of pain
rological patients has documented pain-related
processing, notably the ACC, the thalamus and
neurons in the anterior cingulate cortex (ACC)
the anterior insula was detected, but no activity
that respond both to actual stimulation (thermal
in the somatosensory cortex (see Fig. 2). More-
stimuli) and also to the observation of the same
over, the level of activity within the ACC was
stimuli delivered to another individual (Hutchi-
strongly correlated with subjects’ mean ratings
son, Davis, & Lozano, 1999). In one of the first
of pain attributed to the different situations.
fMRI study of empathy for pain, it demonstrated
In a follow up fMRI study, Jackson, Brunet,
that the ACC, the anterior insula, cerebellum,
Meltzoff and Decety (2005b), again using pictures
and brainstem were activated when healthy par-
of hands and feet in painful scenarios, instructed
ticipants experienced a painful stimulus, as well
the participants to imagine and rate the level of
as when they observed another person receiving
pain perceived from two different perspectives
a similar stimulus, but only the actual experience
(self versus other). Results indicated that both
of pain resulted in activation in the somatosenso-
the self and the other perspectives are associat-
ry cortices and in subcalosal cingulate cortex
ed with activation in the neural network in-
(Singer et al., 2004). Similar results were also re-
volved in the processing of the affective aspect
ported by Morrison et al. (2004) from a study in
of pain, including the ACC and the insula. How-
which participants were scanned during a condi-
ever, the self-perspective yielded higher pain rat-
tion of feeling a moderately painful pinprick
ings and recruited the pain matrix more exten-
stimulus to the fingertips and another condition
sively, including the secondary somatosensory
in which they witnessed another person’s hand
cortex, the mid-insula, and the posterior part of
undergo similar stimulation. Both conditions re-
the subcalosal ACC. Adopting the perspective of
sulted in common hemodynamic activity in a
the other was associated with increase in the
pain-related area in the right dorsal ACC. Com-
right temporo-parietal junction. In addition, dis-
mon activity in response to noxious tactile and to
tinct subregions were activated within the insu-
????
18

---

2006 ? 3 ? 25 ?
19
Fig. 2
Sagittal (on the left) and coronal (on the right) views of activated clusters in the anterior
cingulate and insular cortex elicited by the perception of pain in others, superimposed
on an averaged structural MR image. Physiological research in pain processing demon-
strates that the ACC has a role in the affective dimension of pain, particularly those re-
lated to behavioral responses associated with avoiding or escaping the nociceptive stim-
ulus. This region interrelates attentional and evaluative functions with that of establish-
ing emotional valence and response priorities. The insular cortex is involved in monitor-
ing the physiological state of the body. It receives direct input from the spinothalamic
pathways via the medial thalamic nuclei (the major nociceptive pathway). Interestingly
both the ACC and the anterior insula are found activated by the mere sight of pain in
others (adapted from Jackson, Meltzoff and Decety, 2005).
lar cortex for the two perspectives (anterior as-
externally (by watching emotional laden film
pect for others and more posterior for self).
clips) versus internally (by autobiographical
These neuroimaging data highlight both the sim-
scripts) generated emotions (Reiman et al., 1997).
ilarities and self-other distinctiveness as impor-
Both film-generated emotion and recall-generat-
tant aspects of human empathy. The experience
ed emotion were associated with symmetrical in-
of pain in self is associated with more caudal ac-
creases in the medial prefrontal cortex and thal-
tivations (within area 24), consistent with the
amus. The film condition also resulted in activa-
spino-thalamic nociceptive projections, while the
tion of the hypothalamus, the amygdala, the an-
perception of pain in others is represented in
terior temporal cortex and the occipito -temporo-
more rostral (and dorsal) regions (within area 32).
parietal junction, while the recall condition was
A similar rostro-caudal organization is observed
specifically associated with activation in the an-
in the insula, which is coherent with its anatomi-
terior insula and orbitofrontal cortex. Thus,
cal connectivity and electrophysiological proper-
there is an overlap between externally and in-
ties. For instance, painful sensations are evoked
ternally produced emotions, but this overlap is
in the posterior part of the insula (and not in the
partial. It should be noted that the films and re-
anterior part) by direct electrical stimulation of
call scripts included three emotions (happiness,
the insular cortex in neurological patients (Os-
sadness, and disgust), which were not analyzed
trowsky et al., 2002).
separately.
Shared circuits between emotion generation
A more recent neuroimaging study demon-
and emotion perception in others have also been
strated the involvement of shared representa-
documented in a positron emission tomography
tions (in both emotion-processing areas, and fron-
study which compared the neural response to
to-parietal networks) when subjects feel sympa-
????
19

---

20
??????? 22 ?? 1 ?
thy for another individual (Decety & Chaminade,
entangle themselves from others. Therefore, self-
2003b). In this study, participants were present-
agency is a crucial aspect to successfully navi-
ed with a series of video-clips showing individu-
gate shared representations between self and
als telling sad and neutral stories, as if they had
other (Decety & Sommerville, 2003 ; Decety,
personally experienced them. At the end of each
2005) .
movie, subjects were asked to rate the mood of
Self-awareness and mentalizing
the actor and also how likable they found that
person. Watching sad stories versus neutral sto-
Individuals who are self-aware, as evidenced
ries was associated with increased activity in
by being able to become the object of their own
emotion processing-related structures (including
attention, experience a sense of psychological
the amygdala and parieto-frontal areas) predomi-
continuity over time and space (Gallup, 1998). It
nantly in the right hemisphere.
has been speculated that any organisms capable
Altogether, shared neural circuits between
of self-recognition would have an introspective
self and other at the cortical level have been doc-
awareness of their own mental states and the
umented for action understanding, emotion
ability to ascribe mental states to others
recognition, and pain processing. This mecha-
(Humphrey, 1990). Having a clear sense of self
nism offers an interesting foundation for inter-
may have evolved to solve at least two kinds of
subjectivity because it provides a functional
adaptive problems : 1) the self is the repository
bridge (i.e., shared representations) between
of the social feedback one receives from others
first-person information and third-person infor-
and, 2) it allows to model and understand the in-
mation (Decety & Sommerville, 2003), which al-
ternal, subjective worlds of others, making easier
lows an implicit connection between the self and
to infer intentions and causes that lay behind ob-
the other. There is no specific cortical site for
served behaviors, thus improving interaction ef-
shared representations : their neural underpin-
ficacy (Forgas & Williams, 2002). Interestingly,
nings are widely distributed and the pattern of
the development of self and other mental state
activation (and also presumably deactivation)
understanding is functionally linked to that of
varies according to the processing domain, the
executive functions, i.e., the processes that serve
particular emotion, and the stored information.
to monitor and control thought and actions, in-
The affective-sharing mechanism is necessary
cluding self-regulation, planning, cognitive flexi-
but not sufficient for empathic understanding.
bility, response inhibition, and resistance to in-
Given commensurability between self and other
terference (Russell, 1996). There is increasingly
action representations, a number of key addition-
clear evidence of a specific developmental link
al processes are necessary for successful social
between the development of mentalizing and im-
interactions including the capacity to imagine
proved self-control at around the age of four
others’ behavior. However, in the case of emotion
(Carlson & Moses, 2001). The development of
sharing, a complete overlap between self and
cognitive control is related to the maturation of
other representations would likely induce emo-
the prefrontal cortex (Tamm, Menon & Reiss,
tional distress or empathic over arousal (a self-
2002). In addition, there is hard evidence that a
oriented aversive emotional response), which is
region around the paracingulate sulcus in the
not the goal of empathy. In fact, when experi-
medial prefrontal cortex plays a specific role in
encing empathy, individuals must be able to dis-
mentalizing. This region contains spindle cells, a
????
20

---