Content uploaded by Claus Lamm
Author content
All content in this area was uploaded by Claus Lamm
Content may be subject to copyright.
The Neural Substrate of Human Empathy: Effects
of Perspective-taking and Cognitive Appraisal
Claus Lamm
1,3
, C. Daniel Batson
2
, and Jean Decety
1,3
Abstract
&Whether observation of distress in others leads to empathic
concern and altruistic motivation, or to personal distress and
egoistic motivation, seems to depend upon the capacity for
self–other differentiation and cognitive appraisal. In this experi-
ment, behavioral measures and event-related functional mag-
netic resonance imaging were used to investigate the effects
of perspective-taking and cognitive appraisal while participants
observed the facial expression of pain resulting from medical
treatment. Video clips showing the faces of patients were pre-
sented either with the instruction to imagine the feelings of
the patient (‘‘imagine other ’’) or to imagine oneself to be in
the patient’s situation (‘‘imagine self ’’). Cognitive appraisal was
manipulated by providing information that the medical treat-
ment had or had not been successful. Behavioral measures
demonstrated that perspective-taking and treatment effective-
ness instructions affected participants’ affective responses to
the observed pain. Hemodynamic changes were detected in
the insular cortices, anterior medial cingulate cortex (aMCC),
amygdala, and in visual areas including the fusiform gyrus.
Graded responses related to the perspective-taking instructions
were observed in middle insula, aMCC, medial and lateral pre-
motor areas, and selectively in left and right parietal cortices.
Treatment effectiveness resulted in signal changes in the peri-
genual anterior cingulate cortex, in the ventromedial orbito-
frontal cortex, in the right lateral middle frontal gyrus, and in
the cerebellum. These findings support the view that humans’
responses to the pain of others can be modulated by cognitive
and motivational processes, which influence whether observ-
ing a conspecific in need of help will result in empathic con-
cern, an important instigator for helping behavior. &
INTRODUCTION
Empathy refers to the capacity to understand and re-
spond to the unique affective experiences of another
person (Decety & Jackson, 2004; Batson, Fultz, &
Schoenrade, 1987). This psychological construct de-
notes, at a phenomenological level of description, a
sense of similarity between the feelings one experiences
and those expressed by others. Despite the various
definitions of empathy among psychologists, there is
broad agreement on three primary components: (1) an
affective response to another person, which some be-
lieve entails sharing that person’s emotional state; (2) a
cognitive capacity to take the perspective of the other
person; and (3) some monitoring mechanisms that keep
track of the origins (self vs. other) of the experienced
feelings. Depending on how empathy is triggered, the
automatic tendency to mimic the expressions of others
(bottom-up processing) and the capacity for the im-
aginative transposing of oneself into the feeling and
thinking of another (top-down processing) may be
differentially involved. It also seems likely that both
processes rely upon, to some extent, neural mechanisms
that are involved when the self experiences emotion. It
is not plausible, however, that this sharedness is abso-
lute. A complete overlap between self and other repre-
sentations would produce distress and hamper the
ability to toggle between self and other perspectives.
In recent years, there has been a growing interest in
research on the neural mechanisms that mediate empa-
thy, particularly following the target article by Preston
and de Waal (2002), in which they reviewed an impres-
sive array of evidence in support of the perception–action
model and its fundamental role in social interaction. This
model posits that perception of emotion activates the
neural mechanisms that are responsible for the genera-
tion of emotions. Such a system prompts the observer to
resonate with the emotional state of another individual,
with the observer activating the motor representations
and associated autonomic and somatic responses that
stem from the observed target (i.e., a sort of inverse
mapping). For instance, a handful of functional magnetic
resonance imaging (fMRI) studies have shown that the
observation of pain in others is mediated by several brain
areas that are implicated in processing the affective and
motivational aspects of pain (see Jackson, Rainville, &
Decety, 2006, for a review). In one study, participants
received painful stimuli in some trials and, in other trials,
observed a signal that their partner, who was present in
1
INSERM Unit 280, France,
2
University of Kansas,
3
University
of Chicago
D2007 Massachusetts Institute of Technology Journal of Cognitive Neuroscience 19:1, pp. 42–58
the same room, would receive the same stimuli (Singer
et al., 2004). The anterior medial cingulate cortex (aMCC;
Vogt, 2005), the anterior insula, and the cerebellum were
activated during both conditions. Similar results were re-
ported by Morrison, Lloyd, di Pellegrino, and Roberts
(2004), who applied a moderately painful pinprick stim-
ulus to the fingertips of their participants, and—in a
second condition—showed them a video clip showing
another person undergoing similar stimulation. Both con-
ditions resulted in common hemodynamic activity in
pain-related areas of the right cingulate cortex. In con-
trast, the primary somatosensory cortex showed signifi-
cant activations in response to tactile stimuli only, but not
to visual stimuli. The different response patterns in the
two areas are consistent with the role of the aMCC
in coding the motivational-affective dimension of pain,
which is associated with the preparation of behavioral
responses to aversive events (Vogt, 2005; Paus, 2001). In
another study, participants were shown photographs de-
picting right hands and feet in painful or neutral everyday-
life situations, and were asked to imagine the level of pain
that these situations would produce (Jackson, Meltzoff,
& Decety, 2005). Significant activation in regions in-
volved in the network processing the affective aspect of
pain, notably the aMCC and the anterior insula, was de-
tected. Moreover, the level of activity within the aMCC
was strongly correlated with participants’ mean ratings of
pain attributed to the different situations. These results
lend support to the idea that common neural circuits are
involved in representing one’s own and others’ affective
pain-related states. Recently, Singer and colleagues (2006)
demonstrated that the hemodynamic response in this
circuit is modulated by learned social preferences, espe-
cially in male participants.
Imagining how another person feels and how one
would feel in a particular situation requires distinct
forms of perspective-taking that likely carry different
emotional consequences (Batson, Early, & Salvarini,
1997). Research in social psychology (e.g., Batson et al.,
2003; Underwood & Moore, 1982) has documented this
distinction by showing that the former may evoke em-
pathic concern (defined as an other-oriented response
congruent with the perceived distress of the person in
need), whereas the latter induces both empathic con-
cern and personal distress (i.e., a self-oriented aversive
emotional response). In a recent fMRI study, partici-
pants were shown pictures of people with their hands or
feet in painful or nonpainful situations with the instruc-
tion to imagine themselves or to imagine another in-
dividual experiencing these situa tions ( Jackson, Brunet,
Meltzoff, & Decety, 2006). Both the self-perspective and
the other-perspective were associated with activation in
the neural network involved in pain processing, includ-
ing the parietal operculum, the aMCC, and the anterior
insula. These results reveal the similarities in neural
networks representing first-person and third-person in-
formation, which is consistent with the shared represen-
tations account of social interaction (Decety & Gre`zes,
2006; Decety & Sommerville, 2003). In addition, the
self-perspective yielded higher pain ratings and involved
the pain matrix (Derbyshire, 2000) more extensively in
the secondary somatosensory cortex, the posterior part
of the anterior cingulate cortex (ACC), and the middle
insula. These results highlight important differences be-
tween the self- and other-perspectives. For instance,
although the anterior insula is activated both when par-
ticipants imagine their own and when they imagine
another’s pain, nonoverlapping clusters can be identi-
fied within the middle insula. Likewise, both self- and
other-perspectives are associated with a common sub-
area in the aMCC, but the self-perspective selectively
activated another part of this region.
Finally, being aware of one’s own emotions and feel-
ings enables us to reflect on them. Among various
emotion regulation strategies when observing a target in
pain, reappraisal by denial of relevance (i.e., taking a de-
tached observer position), by implicitly or explicitly gen-
erating an image of the observing self which is unaffected
by the target, is known to reduce the subjective experi-
ence of anxiety, sympathetic arousal, and pain reactivity
(Kalisch et al., 2005). Such a strategy is likely to play an
important role in preventing empathic overarousal (think
about a psychotherapist and his/her client). fMRI studies
have identified a limited number of regions in the antero-
lateral prefrontal and medial prefrontal/orbito-frontal
cortices that mediate such function (Kalisch et al., 2005;
Ochsner, Bunge, Gross & Gabrieli, 2002).
The goal of the present experiment was to assess the
respective contribution of the processes that mediate
empathy: affective sharing, perspective-taking, and cog-
nitive appraisal. We exposed participants to video clips
showing the faces of persons who were described as
patients suffering a neurological disease affecting their
audition. As a cover story, participants were told that
patients underwent a sound therapy supposed to im-
prove their medical status; however, as this therapy
involved being stimulated with sounds of a certain
frequency, patients had to suffer great pain during
treatment. Participants were requested to watch the
videos adopting two different perspectives, that is, ei-
ther imagining how they themselves would feel if they
were in the place of the other (imagine self ), or
imagining how the other feels (imagine other). In
addition, participants were told that the video clips
had been shot from two groups of individuals. In one
group, patients got better after treatment, whereas
patients from the other group did not benefit from that
treatment. This manipulation was performed to elicit
different cognitive appraisals by observers watching
identical stimuli, but with knowledge of different impli-
cations. It was anticipated that witnessing another per-
son suffering and knowing that the treatment had not
been effective would increase emotional distress in the
observer (and vice versa). During scanning, participants
Lamm, Batson, and Decety 43
had to rate intensity and unpleasantness of pain imag-
ined when watching the video clips. After the scan-
ning session, additional behavioral data, including an
emotional response scale and two memory tests, were
collected. It was anticipated that the neural network
involved in the processing of the affective and motiva-
tional aspects of pain (MCC/ACC and insula) would not
only be activated by the perception of pain in others,
but also be modulated by the perspective-taking in-
structions as well as by the cognitive appraisal resulting
from knowledge about the current state of the patients.
Notably, different aspects of the aMCC/ACC and insula
were expected to be differentially associated with these
two factors. Further, if imagining self in pain leads to
more personal distress than imagining other, one may
anticipate stronger signal increase in the amygdala for
the former than for the latter. In addition, the process of
self–other differentiation during perspective-taking was
expected to selectively activate left and right temporo-
parietal areas (Decety & Gre`zes, 2006).
METHODS
Participants
Seventeen right-handed healthy volunteers (8 women),
aged between 18 and 31 years (mean = 23.5 years,
SD = 4.4), participated in the main experiment of this
study. They gave informed written consent and were paid
for their participation. No subject had any history of
neurological, major medical, or psychiatric disorder. The
study was approved by the local Ethics Committee and
was conducted in accordance with the Declaration of
Helsinki. From a pretest sample of 64 candidates, we se-
lected those having at least moderately high scores on
the Empathic Concern Scale of the Interpersonal Reac-
tivity Index (IRI; Davis, 1996), and an IRI Perspective
Taking score of at least 11. This was done to exclude par-
ticipants with low empathy and perspective-taking abil-
ities (see Results). In order to reduce social desirability,
study-compliant responding, and priming effects, ques-
tionnaires were completed without information about the
purpose of the study several weeks before the study. In
addition, 111 volunteers participated in behavioral experi-
ments designed for stimulus selection and validation.
Stimulus Preparation and Validation
Two types of stimuli were generated for this study:
aversive sounds and video clips showing persons listen-
ing to these sounds. Sounds and video clips were
validated in two independent behavioral experiments.
Thirty aversive sounds were composed by mixing
three highly dissonant tone pairs in a frequency range
from 1300 to 11,000 Hz (to minimize interference with
MR gradient noise). Sounds were composed using
S_TOOLS-STx (v3.6.1; Acoustics Research Institute of
the Austrian Academy of Sciences, Vienna, Austria) and
were digitally amplified to yield sound pressure levels of
approximately 95 dB(A). Affective reactions to these
sounds were evaluated using the SAM manikin approach
(Bradley & Lang, 2000), and sounds with an average
unpleasantness rating of 8 were selected for further
use (with ratings ranging from 1 = very pleasant to 9 =
very unpleasant). Video clips (without sound) showing
the face of individuals listening to these sounds were re-
corded from 50 healthy individuals (targets), who were
either professional actors or experienced pantomime
players (26 women, age range: 18–37). Video clips were
shot from a frontal view using a digital color camcorder,
were centered on the target’s nose, and showed the
whole head and parts of the shoulders. Targets were
instructed to direct their gaze at a point approximately
50 cm below the camcorder lens to avoid direct eye
contact. In order to imply that videos had been taken in
a hospital environment, videos were taken against a light
blue background curtain (as used in hospitals), and
targets were wearing a white medical blouse and audio-
metric headphones. Targets were instructed to empha-
size their painful response to the sounds in order to
yield facial expressions of strong pain. Video clips were
edited to show the transition from a neutral facial ex-
pression to the painful reaction resulting from sound
presentation (Figure 1).
Video clips in which targets displayed brow lowering,
orbit tightening, and either cursing or pressing of the
Figure 1. Sample frames
extracted from a video clip
used in this study showing
the transition from neutral
to painful facial expression
triggered by the presentation
of an aversive, painful sound.
44 Journal of Cognitive Neuroscience Volume 19, Number 1
lips, or mouth opening or stretching, were selected for
further analysis, as these movements have consistently
been attributed to the facial expression of pain (e.g.,
Craig, Prkachin, & Grunau, 2001). Only video clips
displaying a natural pain response were selected (al-
though recent evidence documents that the deliberate
exaggeration of pain does not yield unrealistic facial
expressions; Prkachin, 2005). This selection procedure
yielded 105 video clips that were shown to a sample of
94 healthy participants (67 women, age range: 18–54),
who rated the pain experienced by targets on a 7-point
Likert-type scale ranging from ‘‘not painful at all’’ to
‘‘extremely painful.’’ The resulting mean ratings ranged
between M= 2.568 and M= 6.274 for the 105 video
clips (mean and standard deviation of all clips: M=
4.593, SD = 0.966). Fifty-five different clips of 24 targets
(12 women) with the highest pain ratings were selected
for the fMRI study. The mean rating of these clips was
M= 5.419 (SD = 0.493, range: M= 4.579 – 6.274).
Experimental Procedure
Using a standardized written and verbal instruction
procedure, participants were informed that they would
watch video clips of patients experiencing painful audi-
tory stimulation due to medical treatment. According to
instructions, the patients were suffering from a neuro-
logical disease (Tinnitus aurium) that had been treated
using a new therapy. The new therapy required repeated
stimulation of patients by sounds of specific frequencies
and amplitudes, resulting in great pain. As this new
therapy was being used for the first time, some of the
patients benefited from it, whereas others did not.
Participants were instructed to watch the video clips
adopting either of two perspectives (imagine self vs.
imagine other), and were told to which treatment group
(effective vs. not-effective treatment) each patient be-
longed. A sample of the sounds was played to partic-
ipants, pointing out that the pain evoked in patients was
considerably stronger due to their neurological illness.
Before scanning, participants performed several practice
trials to familiarize them with the experimental design as
well as with the button box used for responding.
A22 factorial design with factors perspective-taking
(levels: imagine self vs. imagine other) and treatment ef-
fectiveness (levels: treatment effective vs. not-effective)
was implemented. A mixed blocked/event-related presen-
tation mode was used for stimulus presentation. Before
each block, an instruction screen was shown that indi-
cated the perspective to adopt, and whether the patients
to be shown belonged to the effective or to the not-
effective treatment group. Each block consisted of four
video clips (i.e., trials) of four different patients, showing
the transition from a neutral facial expression (0.5 sec)
to the expression of strong pain triggered by auditory
stimulation (3 sec). The last video clip of each block had
to be evaluated in terms of intensity and unpleasantness
of pain (see Behavioral Measurements section). In order
to control for the intensity of visual stimulation, scram-
bled static images with a centered fixation dot were
shown during intertrial intervals (ITIs). Mean ITI duration
was 6 sec, and ITIs were randomly jittered to reduce
stimulus predictability and to allow efficient event-related
signal estimation (Donaldson & Buckner, 2001).
Four consecutive fMRI runs comprising five blocks
each were performed, with the sequence of blocks being
pseudorandomized and counterbalanced across partic-
ipants, and with no condition being repeated more than
once per run. In each condition, different video clips of
three male and female patients were shown. A patient
shown in one condition was never shown in any of the
other conditions. Some of the clips were repeated (not
more than once) to yield a final number of 20 trials per
condition. Video clips of conditions had equal mean
ratings and standard deviations, and identical ITI distri-
butions. Assignment of patients to conditions was coun-
terbalanced across participants. After each run, a short
break was provided to participants, and before starting
the next run, perspective-taking and emotion regulation
instructions were briefly recapitulated verbally.
In addition to the empathy-related fMRI runs, a local-
izer task was performed to identify the sensory and affec-
tive neural network activated by the first-hand experience
of painful auditory stimulation. Aversive sounds were
presented in an ON/OFF block design with 9 ON and
10 OFF epochs (duration 6 and 16 sec, respectively). After
each ON epoch, participants had to evaluate the intens-
ity and unpleasantness of the pain evoked by the sound.
Behavioral Measures
A variety of behavioral measures was employed to in-
vestigate the effects of experimental manipulations and
to assess the relationships between personal traits and
neural activity. In the scanner, intensity and unpleas-
antness of the imagined pain were rated on a 4-point
scale ranging from ‘‘no pain’’ to ‘‘worst imaginable
pain.’’ Mean ratings of conditions were analyzed using
a22 repeated-measures analysis of variance (ANOVA),
with factors perspective taking and treatment effective-
ness. After scanning, participants were submitted to a
recognition memory test, a forced-choice memory test,
and a behavioral experiment assessing self-reported
emotional responses. They were also extensively de-
briefed using a structured semistandardized interview.
In the recognition memory test, 52 static photos of faces
were presented. Half of them depicted patients that had
been shown during MRI scanning, and half of them were
false targets. Participants were asked to decide whether
the person on the photo was one of the patients shown
during MRI scanning. Photos were edited to contain only
the faces and not the entire heads of targets in order to
avoid recognition by nonfacial characteristics such as
hair color, hair cut, characteristic ears, foreheads, or the
Lamm, Batson, and Decety 45
like. In the forced-choice memory test, participants had
todetermineforeachofthe24patientstowhich
treatment group he/she belonged. Data of the two
memory tests were analyzed using correct recognition
or correct classification rates as dependent variables. It
was predicted that the perspective-taking instructions
would lead to a self-referential memory effect (Rogers,
Kuipers & Kirker, 1977), that is, better recognition and
classification rates should be associated with patients
viewed using the self-perspective.
Emotional responses in the four experimental con-
ditions were assessed using a procedure developed by
Batson, Early, et al. (1997). Participants were shown four
video clips per condition, and rated the degree to which
they experienced 14 emotional states (e.g., alarmed,
concerned, compassionate, distressed) while watching
a clip (1 = not at all, 8 = extremely). In addition,
intensity and unpleasantness of pain were evaluated in
the same way as in the MR scanner, but using an 8-point
rating scale. Ratings of emotional states were aggregated
by calculating empathic concern and personal distress
indices (see Batson, Early, et al., 1997, for details).
Indices were analyzed using repeated-measures ANOVAs
with the factors index, perspective, and treatment effec-
tiveness. We predicted that imagining how oneself
would feel in the place of the patient would lead to
higher personal distress, whereas imagining how the
other felt would trigger more empathic concern. Behav-
ioral data (including pretest data) were analyzed using
SPSS 12.0.1 (SPSS, Chicago, IL, USA), and significance
was defined as p.05.
Finally, participants completed four dispositional
measures: the IRI (Davis, 1996), Empathy Quotient
[EQ] (Baron-Cohen & Wheelwright, 2004), Emotional
Contagion Scale [ECS] (Doherty, 1997), and Emotion
Regulation Scale [ERS] (Gross & John, 2003). The IRI is
probably the most widely used self-report measure of
dispositional empathy. Its four subscales (empathic con-
cern, perspective taking, fantasy scale, and personal
distress) assess different aspects of empathic responses.
The EQ is a recently developed and well-validated
questionnaire tapping cognitive empathy, emotional
reactivity to others, and social skills. It was used as an
alternative assessment of dispositional empathy. The
ECS assesses the susceptibility to other’s emotions from
afferent feedback generated by mimicry. We expected
such mimicry during watching the facial expression of
pain. Finally, two different strategies of emotion regula-
tion—emotion suppression and emotion reappraisal—
were assessed using the ERS.
fMRI Data Acquisition and Analysis
MRI was performed using a whole-body 1.5-T Siemens
Sonata scanner (Siemens, Erlangen, Germany). Func-
tional images were acquired using an echo-planar imag-
ing (EPI) sequence (echo time TE = 60 msec, repetition
time TR = 1990 msec, flip angle = 908, 21 axial slices
with 4.5 mm slice thickness and 0.45 mm gap, in-plane
resolution = 3.6 3.6 mm
2
,6464 matrix, FOV =
230 230 mm
2
). Images were acquired using an as-
cending interleaved sequence with no temporal gap
between consecutive image acquisitions. The inf luence
of in-plane susceptibility gradients in orbito-frontal re-
gions was reduced by orienting image slices according
to recommendations by Deichmann et al. (2003). Four
fMRI runs with 162 image acquisitions were performed
to investigate hemodynamic responses related to empa-
thy, and one run with 158 images was performed for
the localizer task. The first nine scans of each run served
to achieve steady-state magnetization conditions and
were discarded from analyses.
Stimulus presentation and response collection were
performed using the Presentation software (Neurobeha-
vioural Systems, Albany, CA, USA), with block onsets being
temporally synchronized with functional magnetic reso-
nance image acquisition. Visual stimuli were presented
using a back-projection system, with video clips sub-
tending a visual angle of 9.4787.568. MR-compatible
headphones (CONFON HP-SI01; MR Confon GmbH, Mag-
deburg, Germany) were used to present auditory stimuli
at sound pressure levels of approximately 95 dB(A). A
button box consisting of four buttons pressed using the
dominant right hand recorded the responses of subjects.
Image processing was carried out using SPM2 ( Well-
come Department of Imaging Neuroscience, London,
UK), implemented in MATLAB 6.5 (Mathworks, Sher-
born, MA, USA). Preprocessing included slice-timing cor-
rection, correction for head motion (realignment to first
image volume), normalization to the EPI template pro-
vided in SPM2, and smoothing using a 6-mm full-width
half-maximum isotropic Gaussian kernel. Event-related
responses were assessed by setting up fixed-effects
general linear models for each subject. Regressors of
interest modeling the four experimental conditions,
the instruction display, and the evaluation epochs
were set up, and regressors were convolved with a
canonical hemodynamic response function (hrf ) and
their temporal and dispersion derivatives. The latter
were incorporated into the model to account for poten-
tial timing differences in the (neural and hemodynamic)
response to the video stimuli (Friston et al., 1998).
Hemodynamic responses in the localizer task were
modeled using the canonical hrf only. Fixed-effects
models incorporated a high-pass filter with a frequency
cutoff at 128 sec. Following model estimation, contrasts
were calculated for each subject to assess differences
between factor levels (Self > Other, Other > Self, Effec-
tive > Not-effective, Not-effective > Effective, positive
and negative interaction). In addition, signal changes in
relationship to the inherently modeled baseline were
assessed. The resulting contrast images, containing pa-
rameter estimates for each of the three basis functions,
were entered into second-level random effects repeated
46 Journal of Cognitive Neuroscience Volume 19, Number 1
measures ANOVAs. Nonsphericities of ANOVAs were
accounted for by using Greenhouse–Geisser correction,
as implemented in SPM2. F-contrasts incorporating all
three basis parameters as well as T-contrasts assessing
the parameter estimates for the canonical hrf only were
computed. As analyses assessing derivatives did not yield
relevant effects, only the results of the T-contrasts will be
reported here. Activity common to the observation of
pain in others and the first-hand experience of pain were
analyzed using a masking analysis. This analysis con-
sisted of a random-effects ttest for the contrast Watch-
ing Pain > Baseline that was masked (inclusively) by the
contrast Sound > Baseline. For analyses of activity
differences between factor levels, a voxel-level threshold
of p= .001 (uncorrected) and a spatial extent threshold
of k= 5 was chosen. The contrast Watching Pain >
Baseline was thresholded at p= .00001 (uncorrected),
k= 20, and a threshold of p= .0001 (uncorrected),
k= 20 was used for the sound localizer task, as the latter
had lower power due to fewer stimulus repetitions and
fewer image acquisitions. Choice of thresholds was
based upon former studies of our group using similar
task manipulations ( Jackson et al., 2005, 2006), as well
as upon exploratory data analyses. Note also that similar
thresholds were used in the region-of-interest (ROI)
analyses of Botvinick et al. (2005) and Singer et al.
(2004). In addition, for analyses focusing on the more
subtle differences between factor levels, the threshold
was lowered to p= .005 to assess whether there was
below threshold activation in a priori defined regions in-
volved in the perception of pain and in emotion regula-
tion. Anatomic and Brodmann’s area labeling of activity
clusters was performed using the Anatomy Toolbox
(v1.0; Eickhoff et al., 2005), Anatomic Automatic Label-
ing [AAL] (Tzourio-Mazoyer, Landeau, Papathanassiou,
Crivello, Etard et al., 2002), and the Talairach Demon
database (http://ric.uthscsa.edu/projects/tdc). Nomen-
clature for activations in the cingulate cortex is based
on a recent review by Vogt (2005). In addition to the
whole-brain analyses, an ROI analysis of amygdala ac-
tivity was performed using the MarsBaR toolbox, v0.38
(www.sourceforge.net/projects/marsbar). This analysis
extracted parameter estimates of activity in the left and
right amygdala (structurally defined with ROIs provided
in the MarsBaR toolbox) to analyze them using repeated-
measures ANOVA.
In order to assess the relationship between behavioral
data and brain activity, random-effects correlation anal-
yses were performed. Scores on the Empathic Concern
subscale of the IRI, the EQ questionnaire, and the
normalized values of the empathic concern index were
correlated with parameter estimates of the contrast
Other > Baseline. In addition, Self > Baseline was
correlated with the IRI Personal Distress subscale and
normalized personal distress index values, and ECS
scores were correlated with Watching Pain > Baseline.
A rather liberal significance threshold of p= .001 (un-
corrected) and k= 5 was selected for these analyses. In
order to avoid an abundance of false positives associated
with the multitude of analyses, significant correlations
were only interpreted if they were located in a priori
defined regions of the pain matrix (Derbyshire, 2000).
RESULTS
Dispositional Measures
Table 1 compares responses to the dispositional mea-
sures with published normative data. This comparison
shows that IRI and EQ scores were slightly lower, but
clearly within the range of the norm values—despite the
fact that we had preselected subjects to exclude those
with low empathic concern and perspective-taking abil-
ities. Thus, albeit empathic concern in our final sample
was not above average, the truncated range should be
kept in mind when comparing our results to other
studies. Emotional contagion was noticeably lower than
in the norm population. Emotion regulation by means of
reappraisal was higher than in the norm population,
whereas emotion suppression was slightly lower. Note
though that we are comparing Anglo-American norm
populations with a French sample filling out French
translations of the questionnaires.
Behavioral Data
Due to equipment failure, the behavioral data of one
participant were not available. Ratings acquired during
MR scanning revealed significant effects of the perspec-
tive-taking and the treatment-effectiveness factors. Pain
intensity was significantly affected by treatment effective-
ness [main effect of the effectiveness factor, F(1,15) =
6.059, p= .026, h
2
= 0.288], with pain intensity ratings
being higher when the treatment was not effective
[M(Effective) = 3.116, M(Not-effective) = 3.398]. Per-
spective-taking did not significantly affect the ratings
(no main effect of factor perspective taking, p= .350),
and the interaction term also was not significant
(p= .869). A similar result was obtained with the
unpleasantness ratings [main effect of treatment effec-
tiveness, F(1,15) = 37.31, p< .001, h
2
= 0.713; M(Ef-
fective) = 2.814, M(Not-effective) = 3.567; other effects,
p> .381]. Evaluations collected during the postscanning
behavioral experiment indicated that watching patients
undergoing ineffective treatment resulted in higher
unpleasantness [main effect of treatment effectiveness,
F(1,15) = 6.534, p= .022, h
2
= 0.303; other effects,
p> .611]. No significant results were obtained for pain
intensity ratings ( p> .455 for all effects). Ratings of the
aversive sounds during the localizer task yielded a mean
intensity rating of 2.931 (SD =0.563)andamean
unpleasantness rating of 2.872 (SD = 0.666).
The recognition memory test revealed that patients
viewed with the self-perspective were remembered better
Lamm, Batson, and Decety 47
[main effect of perspective taking, F(1,15) = 4.623, p=
.048, partial h
2
=0.236;M(Self ) = 34.375, SD = 15.478;
M(Other) = 27.083, SD = 13.088]. Treatment effec-
tiveness did not have a significant effect on recogni-
tion rates ( p= .485), nor was the interaction term
significant ( p= .736). In addition, patients displaying
stronger pain were more likely to be recognized (as indi-
cated by a significant correlation between the perceived
intensity of pain determined in the pretests and the
percentage of correct hits, r= .429, p= .004).
The forced-choice memory test showed that the self-
perspective led to a higher percentage of correct classi-
fications [main effect of perspective taking, F(1,14) =
4.421, p=.054,h
2
=0.24;M(Self ) = 34.444, SD =
15.387; M(Other)= 24.444, SD =17.385].Neitherthe
treatment-effectiveness main effect ( p=.815)northe
interaction term was significant ( p=.526).Themain
effect of perspective-taking missed the chosen threshold
because one participant showing stereotyped response
behavior had to be excluded from the analysis, resulting in
a reduction of degrees of freedom; note though that
estimated effect size (partial h
2
)isslightlyhigherthanin
the recognition memory test. Correlation with perceived
pain intensity was not significant (r=.191,p=.220).
Analysis of the behavioral experiment performed after
scanning confirmed our predictions concerning the
effects of perspective-taking on empathic concern and
personal distress. Empathic concern was considerably
stronger when participants focused on the feelings of
the other, whereas adopting the self-perspective led to
stronger personal distress [interaction between indices
and perspective factor, F(1,15) = 16.715, p= .001,
partial h
2
= 0.527; Figure 2A]. In addition, personal
distress was generally more pronounced if the treatment
was not effective [main effect of treatment effectiveness,
F(1,15) = 10.103, p= .006, partial h
2
= 0.402]. These
effects were additionally modulated by the treatment
effectiveness manipulation: Whereas the treatment out-
come had almost no modulating effect on empathic
concern for the self-perspective, adopting the other-
perspective strongly increased personal distress when
the treatment was not effective [three-way interaction,
F(1,15) = 5.884, p= .028, partial h
2
= 0.282; Figure 2B].
The semistandardized interviews performed during
experimental debriefing revealed that participants were
able to differentiate the four different conditions, and that
there were no suspicions concerning the authenticity of
the cover story. The majority of participants reported
reacting to both patient groups in similar ways, but
that they used reappraisal strategies when watching pa-
tients undergoing effective treatment. For example, self-
reassuring statements such as ‘‘the patient is in pain, but
he/she will be OK soon’’ were used. During the other-
condition, participants adopted a more other-oriented
perspective and focused more on the facial expressions
of the patients than during the self-perspective. Notably,
14 subjects reported overt facial mimicry while watching
Table 1. Mean Scores and Standard Deviations for the Dispositional Measures
Interpersonal Reactivity Index (IRI) Emotion Regulation Questionnaire
PT EC PD FS
Empathy
Quotient
Emotional
Contagion Scale Reappraisal Suppression
Sample (n= 17) Mean (SD)16.35 (2.91) 19.76 (1.95) 9.94 (6.15) 17.12 (5.07) 38.13 (9.98) 43.56 (3.37) 31.35 (4.23) 13.59 (6.26)
Normative Data
a
Mean (SD) 17.37 (4.79) 20.36 (4.02) 10.87 (4.78) Not available 41.8
b
/47.2
c
(11.2
b
/10.2
c
)
54.3 (8.1) 27.6
b
/27.66
c
(5.64
b
/6.12
c
)
14.56
b
/12.56
c
(4.44
b
/4.72
c
)
The table provides results for the sample investigated in our study, and published normative values.
PT = perspective taking; EC = empathic concern; PD = personal distress; FS = fantasy. Maximum Scores: IRI Subscales = 28; Empathy Quotient = 80; Emotional Contagion = 60; Reappraisal = 42;
Suppression = 28.
a
Normative data derived and transformed to sum scores from: IRI (Bellini, Baime, & Shea, 2002); EQ (Baron-Cohen & Wheelwright, 2004); ECS (Doherty, 1997); ERS (Gross & John, 2003).
b
Male sample.
c
Female sample.
48 Journal of Cognitive Neuroscience Volume 19, Number 1
the videos, with reports of mimicry being stronger in the
self-perspective in eight of these subjects.
Network of Areas Involved in the Observation
of Pain
Observation of pain expressed by the patients activated
a widely distributed network of brain regions, ref lecting
the sensory, cognitive–motor, and affective processing
of the stimuli (Figure 3). Clusters comprising (bilater-
ally) medial and lateral occipital cortex, including the
fusiform gyrus, indicate the visual processing of the
stimuli. Activity in bilateral anterior and in the left
middle insula, aMCC, thalamus, basal ganglia (pallidum
and caudate nucleus), and bilateral periamygdalar re-
gion reveals the affective response to the observation
of pain. Additional significant clusters were detected in
motor control-related regions, such as the cingulate and
supplementary motor area (CMA/SMA) and the lateral
precentral gyrus, as well as in temporo-parietal and
lateral prefrontal areas (Brodmann’s areas 8 and 9).
Analysis of the localizer task resulted in bilateral
hemodynamic changes in the posterior superior tempo-
ral gyrus, comprising the Heschl gyrus and overlapping
with the probabilistic cytoarchitectonic maps of the
primary auditory cortices provided in the Anatomy
Toolbox (Morosan et al., 2001). Also, activity was de-
tected bilaterally in the anterior insula, in the left middle
Figure 2. Mean values for
the empathic concern and
personal distress indices.
(A) Note that adopting
the self-perspective elicits
higher personal distress,
whereas the other-perspective
triggers higher empathic
concern. (B) This effect is
modulated by the treatment
effectiveness factor (effective
vs. not-effective treatment).
See text for further details.
Figure 3. Significant
hemodynamic response to
the observation of patients
expressing pain (Watching
Pain > Baseline). Activation
was detected in the neural
network involved in the
sensory (FFG = fusiform
gyrus, MOG = middle
occipital gyrus) and the
affective processing of pain
(insula, aMCC). Results are
superimposed on axial (z=
16, z= 6), coronal ( y=
70), and sagittal sections
(x= 0) of the single-subject
structural MNI MRI template
(used in all figures and
displayed in neurological
convention). Threshold p=
.00001 (uncorrected), k= 20.
Lamm, Batson, and Decety 49
insula, in several subclusters in the aMCC, in the rostral
thalamus, in a mesencephalic cluster containing the
corpus geniculatum mediale and the periaqueductal
gray, and in areas involved in motor control (pallidum
and caudate nucleus, SMA, CMA, and right precentral
gyrus; Table 2 and Figure 4).
The masking analysis revealed a vastly overlapping
neural network, including several subclusters in bilateral
anterior and left middle insula, a cluster in the aMCC and
the right amygdala, as well as in the SMA, CMA, and right
precentral gyrus (see Table 2 and Figure 5).
Responses Related to Perspective-taking
and Treatment Effectiveness
Contrasting the self with the other conditions revealed
different responses in a number of brain regions involv-
ed in pain processing, perspective-taking, and agency
(see Table 3, Figures 6 and 7). The contrast Self > Other
revealed stronger responses with the self-perspective
in bilateral insula, left supramarginal gyrus (BA 40), left
middle frontal gyrus (Brodmann’s area 9), and in sev-
eral areas involved in motor control such as the SMA,
the right dorsal premotor cortex (lateral BA 6), the pu-
tamen, and the caudate nucleus. Note that clusters in
the insula are located in an area classified as the mid-
dle (dysgranular) insula (Mesulam & Mufson, 1982) and
Table 2. Common Hemodynamic Responses during the
Observation of Patients Expressing Pain and the First-hand
Experience of Pain (Masking Analysis, Voxel Threshold:
p= .00001, Uncorrected; Cluster Size Threshold: k= 20;
Masking Threshold: p= .001, Uncorrected)
MNI
Coordinates
Brain Region L/R/M t Value x y z
Middle Temporal Gyrus L 9.45 52 216
x Temporal Pole of STG L 8.49 38 220
x STG L 6.47 48 812
SMA M 8.84 0 6 62
x SMA/CMA M 7.09 0 14 50
Insula R 7.17 42 218
Anterior Insula R9.6634186
Anterior Insula L 8.66 38 18 4
x Anterior Insula L 7.87 36 26 4
xIFG L 7.86 32 26 10
STG R 8.18 50 44 18
x STG R 6.90 40 42 4
aMCC L 7.98 10 12 40
Orbital part of IFG R 7.35 44 20 16
x Temporal Pole of STG R6.3850614
Olfactory Bulb/Gyrus Rectus R 7.09 22 10 16
x Amygdala R 6.98 26 0 22
Stereotactic coordinates and tvalues are provided for the local voxel
maxima in the respective cluster. x = subpeaks of a cluster; L = left
hemisphere; R = right hemisphere; M = medial activation; IFG =
inferior frontal gyrus; STG = superior temporal gyrus; aMCC =
anterior medial cingulate cortex; SMA = supplementary motor area;
CMA = cingulate motor area.
Figure 4. Significant hemodynamic response to painful auditory
stimulation (Sound > Baseline). Results are superimposed on a
sagittal section (x= 3). PAG = periaqueductal gray. Threshold
p= .0001 (uncorrected), k= 20.
Figure 5. Brain areas
commonly activated by the
observation and the first-
hand experience of pain
(masking analysis). Results are
superimposed on sagittal (x=
6, x= 0) and axial sections
(z= 4), showing shared neural
activation in areas coding the
motivational–affective aspects
of pain. Threshold p= .00001
(uncorrected), k= 20.
Threshold for the mask:
p= .001 (uncorrected).
50 Journal of Cognitive Neuroscience Volume 19, Number 1
are clearly distinct from the more rostral clusters of
the contrast Watching Pain > Baseline. When lowering
the threshold to p= .005, two additional clusters were
identified in the aMCC (MNI coordinates of cluster
maxima: x=0,y= 16, z= 24; t= 3.28; and x=6,
y=4,z= 40; t= 3.25). Also, although the size of the
clusters in the middle insula increased considerably, no
additional clusters were identified in the anterior or
posterior parts of the insular cortex. The reverse con-
trast (Other > Self ) revealed significant clusters in the
right superior and right inferior parietal lobe (Table 3).
Observing patients undergoing ineffective treatment
(Not-effective > Effective) evoked a stronger response
in the perigenual ACC (pgACC; x=4, y= 38, z= 20;
t= 3.76). This cluster was clearly rostral to the clusters
detected with the contrast Self > Other and with the
localizer task. Lowering the threshold yielded two addi-
tional clusters in the ventromedial part of the orbito-
frontal cortex (OFC; x= 16, y= 22, z=14; t=
3.39; x=20,y= 46, z=14; t= 3.06; Figure 7). The
reverse contrast (Effective > Not-effective) revealed a
significant cluster in the cerebellum (x= 16, y= 22,
z=14; t= 4.64). Note also that watching patients
undergoing effective treatment (Effective > Baseline)
clearly activated the aMCC, the insula, and the amygdala
(see Discussion).
Contrasts assessing the interaction between the per-
spective-taking and the treatment-effectiveness manipu-
lations resulted in a significant positive interaction in
the pgACC (x=0,y= 48, z=0;t= 3.37) and the right
middle frontal gyrus (x= 36, y= 28, z=50;t=
4.02), showing that the treatment-effectiveness manipu-
lation had a stronger effect in these regions when
participants adopted the other-perspective (i.e., (Other
Effective Other Not-effective) > (Self Effective Self
Not-effective). Note that the difference in the pgACC
Table 3. Differences in Hemodynamic Responses When
Adopting Different Perspectives during the Observation of
Patients Expressing Pain (Imagine Self vs. Imagine Other);
Voxel Threshold: p= .001 (Uncorrected); Cluster Size
Threshold: k=5
MNI
Coordinates
Brain Region L/R/M t Value x y z
Self
>
Other
Supramarginal Gyrus L 4.31 52 26 26
x Supramarginal Gyrus L 3.63 60 34 24
Insula R3.794282
Caudate Nucleus L 4.03 16 414
Putamen/Insula L 4.08 30 28
x Insula L 4.07 36 84
SMA M/R 4.01 8 0 54
Precentral Gyrus R 3.97 46 14 56
Precentral Gyrus R 3.57 22 32 66
STG L 3.64 38 10 12
Middle Frontal Gyrus L 4.10 36 36 34
Rolandic Operculum L 3.47 60 82
Other
>
Self
Superior Parietal Lobe R 4.10 26 76 56
x Inferior Parietal Lobe R 3.98 38 68 56
x Superior Parietal Lobe R 3.91 24 66 64
Angular Gyrus R 3.49 48 64 44
Refer to Table 2 for abbreviations.
Figure 6. Brain areas showing
stronger hemodynamic
responses when adopting
the self-perspective (contrast
Self > Other). Activity is
stronger in areas coding
the motivational–affective
aspects of pain (middle
insula, clusters 1 and 3 in
the aMCC), and in the left
temporo-parietal junction
(left TPJ ), ref lecting self–other
distinction. Results are
superimposed on sagittal
(x=6, x=0,x=55)
and coronal ( y=8)
sections. Threshold p= .005
(uncorrected), k=5.
Lamm, Batson, and Decety 51
reflects a difference in relative deactivation that was more
pronounced with the other-perspective [mean param-
eter estimates: M(Other Effective) = 4.15, M(Other
Not-effective) = 0.4, M(Self Effective) = 0.48, M(Self
Not-effective) = 3.4].
The ROI analysis of amygdala activity revealed a
stronger response in the amygdala with the self-perspec-
tive [trend-like main effect for the perspective factor,
F(1,16) = 4.38, p= .053, partial h
2
= 0.215], with this
difference being slightly more pronounced in the left
hemisphere [trend-like interaction Perspective Hemi-
sphere, F(1,16) = 3.077, p= .099, partial h
2
= 0.161;
Figure 8].
Correlation of Brain Activity with Behavioral
Data and Dispositional Measures
Analysis of the EQ scores revealed significant correla-
tions in the right putamen, the left posterior/middle
insula, the aMCC, and the left cerebellum. Scores of the
empathic concern index correlated with a similar cluster
in the aMCC. No significant correlation was observed for
the Empathic Concern scale of the IRI, the personal
distress index, and the Personal Distress scale of the IRI.
Scores of the Emotional Contagion scale correlated with
activity in brain regions involved in affective pain pro-
cessing (insula and aMCC) and movement control (SMA,
lateral premotor area; Table 4, Figure 9). In addition,
two clusters in the left and right parietal cortex close
to the ones revealed by the Self–Other contrasts indi-
cated stronger hemodynamic responses with higher
Emotional Contagion scores in these regions.
DISCUSSION
There are good reasons to posit that witnessed pain in
others may result in anxiety, promoting at least cautious
approach behavior or general threat-defense mecha-
nisms (MacDonald & Leary, 2005). The affective experi-
ence of pain signals an aversive state and motivates
behavior to terminate, reduce, or escape exposure to
the source of the noxious stimulation (Price, 1999).
Indeed, negative feelings triggered by pain usually mo-
tivate organisms to avoid dangerous stimuli and move
away from danger. However, the observation of pain in
others may also instigate an altruistic motivation to help
the other, which is quite different from the egoistic
motivation to reduce personal distress.
The aim of our study was to investigate the effects of
perspective-taking and cognitive appraisal on the behav-
ioral and neural correlates of pain observation. To this
end, participants watched video clips of patients display-
ing an aversive emotional response due to painful
auditory stimulation under different conditions. Using
a number of both state and trait behavioral measures
and event-related fMRI, we were able to demonstrate
distinct behavioral and neural responses that are in
good agreement with both empirical findings and theo-
retical concepts concerning the promotion of empathic
emotion and, ultimately, altruistic motivation (Batson
Figure 7. Brain areas showing
stronger hemodynamic
responses when observing
patients who did not benefit
from the painful sound
treatment (Not-effective >
Effective). Results are
superimposed on sagittal
sections. Threshold p= .005
(uncorrected), k=5.
Figure 8. Stronger
hemodynamic responses in
left and right amygdala when
subjects adopted the self-
perspective (contrast Self >
Other). Left: tvalues from the
contrast Self > Other, overlaid
on a coronal section of the
single-subject structural MNI
MRI template. Right: mean
(±SE) parameter estimates
for signal changes in the
whole amygdala.
52 Journal of Cognitive Neuroscience Volume 19, Number 1
et al., 2003; Batson, Early, et al., 1997; Batson, Sager,
et al., 1997).
Behavioral Data
Data of the emotional response indices clearly confirm
that perspective-taking and treatment-effectiveness ma-
nipulations were effective. As predicted by previous
social psychology experiments (Batson et al., 2003;
Batson, Early, et al., 1997), adopting the perspective of
the other evoked stronger empathic concern, whereas
personal distress was higher when imagining oneself to
be in the painful situation. In addition, ineffective treat-
ment triggered higher personal distress when patients
were watched with the other-perspective. This might
indicate that participants did not focus on the sensory
aspects of the observed pain, but rather its ultimate
unpleasantness or ‘‘badness’’ by taking into account the
long-term consequences for the patient. Pain ratings
support this interpretation, as they were also modulated
by treatment effectiveness—although no difference in
patients’ actual expression of pain existed because pa-
tients were naı¨ve to the treatment effectiveness. Further,
memory tests demonstrated significant effects of per-
spective-taking, as participants showed better recogni-
tion and classification rates for patients watched using
the self-perspective. This finding is in line with studies
indicating that events that are more relevant for the self
are more likely to be remembered (‘‘self-referential
bias’’; Rogers et al., 1977).
Shared Neural Circuits during Observation
and First-hand Experience of Pain
Watching the video clips was associated with hemody-
namic changes in the medial and lateral occipital cortex
(BA 18) and in the fusiform gyrus. These changes ref lect
the sensory processing of stimuli, as already reported in
previous neuroimaging studies using static and dynamic
face stimuli (e.g., Botvinick et al., 2005; Grosbras & Paus,
2006; Haxby, Hoffman & Gobbini, 2000). In addition, as
predicted, the video clips elicited increased activity in a
number of areas associated with the first-hand experi-
ence of pain, such as the insular cortex, dorsal and ven-
tral areas of the cingulate cortex, the thalamus, and areas
involved in motor control (basal ganglia, medial and
lateral premotor areas). The existence of shared neural
circuits between the experience and the observation
Figure 9. Correlation
between hemodynamic activity
in the SMA and Emotional
Contagion scores.
Table 4. Areas Showing Significant ( p< .001) Correlations
between Hemodynamic Response, Behavioral Data, and
Dispositional Measures
MNI Coordinates
Analysis and
Brain Region L/ R/M r x y z
Empathy Quotient—Other
Putamen R .837 26 0 10
Insula L .795 36 212
aMCC M .826 0 442
Cerebellum L .833 22 46 24
Empathic Concern Index—Other
aMCC R .802 14 16 28
Emotional Contagion—Watching Pain
Insula R .867 30 16 12
x IFG/Operculum R .847 44 18 8
Insula L .848 32 10 10
aMCC/CCMA R .737 16 26 42
SMA M .789 6 452
Precentral Gyrus R .740 50 14 44
Supramarginal Gyrus R .752 64 38 34
Inferior Parietal Lobe L .749 50 42 36
Inferior Parietal Lobe L .742 56 44 56
r= Pearson’s correlation coefficient, other abbreviations as in Table 2.
Lamm, Batson, and Decety 53
of pain is specifically corroborated by the masking
analysis. This analysis relies upon a localizer task that
induces affective responses similar to the ones experi-
enced by the patients shown. Masking revealed that the
observation of pain in others, and its first-hand experi-
ence, activated a largely overlapping neural network. It is
worth noting that this overlap did not include areas
coding the sensory aspects of pain, as neither the
auditory cortex nor the primary or secondary somato-
sensory cortices were activated. Common activation was
confined to areas involved in the motivational–affective
dimension of pain processing, such as the anterior
insula, the aMCC, and the amygdala, as well as to areas
involved in motor control. This confirms several recent
fMRI studies ( Jackson et al., 2005, 2006; Singer et al.,
2004, 2006; Botvinick et al., 2005; Morrison et al., 2004),
which indicate that ACC and anterior insula activity
during the observation of pain is related to the affective
aspects of pain processing rather than to its sensory-
discriminative aspects. This interpretation gets addition-
al support by neurophysiological evidence suggesting
that the anterior dysgranular part of the insula plays
a central role in mediating subjective feeling states, pos-
sibly conveyed via mechanisms of interoceptive aware-
ness(Critchley,Wiens,Rotshtein,O
¨hman, & Dolan,
2005; Craig, 2002). It is also worth noting that electrical
stimulation of the posterior part of the insula, but not of
the anterior part, evokes painful sensations (Ostrowsky
et al., 2002). Altogether, these findings are in agreement
with the proposal that indirect pain representations
(as elicited by the observation or imagination of pain
in others) partially overlap with, but are nevertheless
qualitatively different from, first-hand experiences of
pain (Craig, 1968).
Correlation of Brain Activity with Behavioral
Data and Dispositional Measures
Correlation of hemodynamic activity with behavioral
data and dispositional measures further supports the
hypothesis that the affective network of pain processing
is specifically involved in the perception of pain in
others. Subjects scoring higher on EQ showed stronger
activity in the left middle insula, the cingulate cortex,
and the striatum, and higher empathic concern indices
were associated with a stronger hemodynamic response
in the aMCC. A similar result was reported by Singer et al.
(2004) for the bilateral anterior insula and the aMCC
using the Empathic Concern scale of the IRI. However,
no significant correlations with empathic concern were
found in our study. Note though that empathic concern
correlated with EQ scores only weakly (r= .351) in our
sample, and that our range of IRI scores was consider-
ably smaller due to the preselection of participants with
high EC scores (17–24 as opposed to 12–24 in Singer
et al., 2004).
Further, emotional contagion scores indicated brain–
behavior correlations in similar areas, as well as in areas
involved in motor control (i.e., SMA/CMA, dorsal and
ventral precentral gyrus, and posterior parietal cortex).
These areas belong to a circuit involved in the prepara-
tion and planning of self-generated motor action. Cor-
relation of emotion contagion scores with activity in
these motor areas might thus ref lect the ‘‘inverse map-
ping’’ mechanism posited by the perception–action
account of empathy, which assigns a primary role to
motor mimicry and emotional contagion (Preston & de
Waal, 2002). Such a mechanism may be triggered by
overtly or covertly mirroring in the self the facial pain
expressions displayed by the target. Indeed, comments
in debriefing indicated that some subjects used overt
mimicry, especially in the self condition. Another possi-
bility is that witnessing pain in others automatically
prompts motor responses to withdraw oneself from
pain. These two responses tap similar neural mecha-
nisms and are difficult to separate experimentally.
Effect of Perspective-taking
Imagining the self and imagining the other in pain ac-
tivate similar neural mechanisms. However, a complete
blurring of self and other would be detrimental and is
not the purpose of empathy. Therefore, activation of ad-
ditional neural mechanisms is needed to distinguish the
self from other. This distinction has been associated with
the sense of agency (i.e., the feeling of being causally
involved in an action), which relies on the comparison
between self-generated and externally produced signals.
Neuroscience research has provided clues to the exis-
tence of a cerebral network specifically devoted to this
distinction. Attribution of an action to another agent has
been associated with increased activity in the right pari-
etal cortex (e.g., Farrer et al., 2003). The inferior parietal
cortex is a multisensory integration area that is ideally
suited to detect distinctions between self-generated
and external signals. Interestingly, perspective-taking in-
structions in our study resulted in the activation of dis-
tinct subregions of the left and right parietal cortex.
The self-perspective elicited higher activity in the left
parietal cortex, whereas the right parietal cortex was se-
lectively involved when the other-perspective was adopt-
ed. This pattern of activity is consistent with the major
role of inferior posterior parietal areas in self-agency and
perspective-taking (Decety & Gre`zes, in press; Blanke &
Arzy, 2005; Decety & Sommerville, 2003; Ruby & Decety,
2003). Accumulating evidence from neuroimaging stud-
ies and lesion studies in neurological patients indicates
that the right inferior parietal cortex has a critical func-
tion in the distinction between self-produced actions
and actions generated by others ( Jackson & Decety,
2004; Blakemore & Frith, 2003). Importantly, this is also
true when behavior is merely mentally simulated. For
54 Journal of Cognitive Neuroscience Volume 19, Number 1
example, imagining somebody else performing an action
(Ruby & Decety, 2001) or experiencing an emotion
(Ruby & Decety, 2004), as opposed to imagining
performing the action or experiencing the emotion
oneself, revealed a very similar modulation of left- versus
right-hemispheric parietal hemodynamic activities. A
recent fMRI study of social perception and empathy
demonstrated that activity in the inferior parietal cortex
was negatively associated with the degree of overlap
between self and other, and that less self–other over-
lap led to increased accuracy during social perception
(Lawrence et al., 2006).
Further, the self-perspective led to higher activity in
brain areas involved in the affective response to threat or
pain, such as the amygdala, the insula, and the aMCC.
This is consistent with the idea that personal distress can
be elicited by imagine-self instructions, as first demon-
strated in social psychology studies (e.g., Batson, Early,
et al., 1997). Moreover, such a finding is in good
agreement with another fMRI study that used a similar
perspective-taking manipulation and demonstrated that
the first-person perspective taps into affective processes
to a greater extent than the more detached third-person
perspective ( Jackson et al., 2006). The amygdala plays a
critical role in fear-related behaviors, such as the evalu-
ation of actual or potential threats (LeDoux, 2000).
Interestingly, the amygdala receives nociceptive infor-
mation from the spino-parabrachial pain system and the
insula, and its activity appears closely tied to the context
and level of aversiveness of the stimuli (Zald, 2003).
Imagining oneself to be in a painful and potentially
dangerous situation thus might trigger a stronger fearful
and/or aversive response than imagining someone else
to be in the same situation. Higher activity in the middle
insula may reflect the sensory aspects evoked by the
imagination of pain. A meta-analysis of imaging studies
reporting insular activations (Wager & Feldman Barrett,
2004) suggests that the middle part of the insula plays a
role in coding the sensory–motor aspects of painful
stimulation. Importantly, this region has strong con-
nections with the basal ganglia (Chikama, McFarland,
Amaral & Haber, 1997), in which activity was also higher
when adopting the self-perspective (Table 3). Taken
together, activity in this part of the insula possibly
reflects the simulation of the sensory aspects of the
painful experience. This simulation might both lead to
the mobilization of motor areas (including the SMA) in
order to prepare defensive or withdrawal behavior, and
to interoceptive monitoring associated with autonomic
changes evoked by this simulation process (Critchley
et al., 2005).
When lowering the threshold, two additional clusters
were detected in the aMCC, a region involved in pro-
cessing the affective, evaluative, and attentional aspects
of pain perception (Peyron, Laurent, & Garcia-Larrea,
2000). In addition, activity in this region has been related
to the monitoring of autonomic function (Critchley et al.,
2003). We thus suggest that the self-perspective results
in the evaluation of the affective, autonomic, and moti-
vational consequences obtained from the imagination of
a painful experience, in line with the evocation of
personal distress. This interpretation is also in line with
a review of Bush, Posner, and Luu (2000), labeling this
area as the ‘‘cognitive division ’’ of the ACC.
Effect of Cognitive Appraisal
Humans have the striking capacity to regulate their
emotions (Beer & Heerey, 2003; Tice, Bratslavsky &
Baumeister, 2001). This capacity involves the initiation
of new or the alteration of ongoing emotional responses
through the action of regulatory processes (Ochsner &
Gross, 2005). We suggest that such regulatory processes
play an important role when observing distress in
others, as they enable us to show supportive behavior
even in potentially dangerous or harmful situations. The
majority of neuroimaging studies on emotion regulation
have provided explicit instructions as to how partici-
pants should reappraise or suppress elicited emotions
(see Ochsner & Gross, 2005, for review). In contrast, we
presented information designed to affect cognitive ap-
praisal, as our aim was to create a situation as close as
possible to an everyday context. It was anticipated that
witnessing another person suffering and knowing that
his/her treatment had not been effective would increase
the emotional response in the observer. Conversely,
knowing that a treatment had been beneficial for the
patient was expected to elicit down-regulation of the
perceptually triggered affective response. Both the be-
havioral and hemodynamic data supported this hypoth-
esis. Behavioral data showed higher pain intensity and
unpleasantness ratings when the treatment had not
been effective. This finding was paralleled by activity
differences in a number of brain regions involved in
affective coding and emotion regulation, such as rostral
and perigenual ACC and the ventromedial OFC. Note
also that observing ineffectively treated patients trig-
gered strong activation in the aMCC, the insula, and
the amygdala, indicating an affective response. Based on
self-report and behavioral data, we suggest that this
response was regulated (reappraised) via top-down
mechanisms such as focusing on the long-term conse-
quences of the treatment.
A recent review of cingulate cortex functions in pain
and emotion indicates that a subregion in the rostral
ACC is involved in coding fearful responses ( Vogt, 2005).
Similarly, Bush et al. (2000) labeled this part of the ACC
as its ‘‘affective division,’’ in contrast to the more
posterior-dorsal ‘‘cognitive division’’ (see above). Activ-
ity in this area may hence be related to a stronger
defensive response in cases where the treatment had
no benefit, and thus, the patient’s overall situation was
perceived as being more unpleasant or distressing (as
indicated by the behavioral data). Alternatively and more
Lamm, Batson, and Decety 55
speculatively, ACC activity may have resulted from an
anger-related reaction triggered by the fact that the pa-
tient had to suffer pain without benefiting from it. That
participants experienced more anger when watching
patients undergoing ineffective treatment is—at least
indirectly—indicated by significantly higher scores for
the adjective ‘‘upset’’ on the emotional response scale
when the therapy was not effective [t(15) = 2.695, p=
.027, h
2
= .326]. Note also that self-generation of anger
activates a similar part of the ACC (Damasio et al., 2000).
While watching the videos, the participants in our
study had to consider whether the overall effect of the
painful treatment was positive or negative. This manip-
ulation of the context in which pain occurred differen-
tially elicited OFC activity, which plays an important role
in the evaluation of positive and negative reinforce-
ments, and in the motivational and emotional aspects
of social behavior (Rolls, 2004). The OFC is also involved
in emotion reappraisal, as attending to a negatively
valenced picture evokes stronger activity in the ventro-
medial OFC than reappraising that picture in a way that
it no longer elicited a negative response (Ochsner et al.,
2002). Similarly, watching patients undergoing ineffec-
tive treatment was associated with higher activity in this
region, whereas effective treatment was associated with
a decreased OFC response. Note that the OFC was active
in both conditions (data not shown). Activity in the OFC
thus might reflect differing requirement to evaluate the
overall positive and negative aspects of the presented
stimuli. This top-down process might operate upon the
visually conveyed information about the affective state of
the patients. Interestingly, watching effective versus
ineffective treatment patients did not modulate activity
in either the visual–sensory areas or in the insula (even
when the threshold was lowered to p=.05).This
suggests that the two patient groups were differentiated
after perceiving their emotional reactions (but not
necessarily after participants reacted emotionally them-
selves), and that top-down mechanisms did not operate
on perceptual processing at an early stage. Keep in
mind, however, that this finding might be inf luenced
by the requirement for participants to evaluate the pain
of the patients.
The interaction between perspective-taking and treat-
ment effectiveness yielded increased activity in the
pgACC and the middle frontal gyrus. Involvement of
the pgACC most likely indicates a difference in perspec-
tive-taking requirements, which are more complex when
imagining the feelings of patients who—in contrast to
the observer—do not know the ultimate outcome of the
treatment. In fact, increased activity in the rostral cingu-
late and the paracingulate cortex when performing
mentalizing tasks has reliably been found (see Gallagher
& Frith, 2003, for review). Thus, we speculate that the
stronger deactivation with the other-perspective indi-
cates a stronger requirement to inhibit mentalizing,
which might be counterproductive in a case where
participants have to focus on the long-term consequen-
ces of the treatment and not the immediate affective
reaction of the patient. The middle frontal gyrus, on the
other hand, may be associated with emotion down-
regulation, as indicated by a recent fMRI study on
emotion regulation (Ochsner et al., 2004). However,
studies using explicit emotion regulation instructions
show stronger involvement of lateral prefrontal cortical
regions than our study. This highlights the role of these
regions in deliberately exerting top-down control,
whereas appraisal-based regulation might be supported
by brain structures involved in relatively automatic as-
sessment of reward properties (such as the OFC).
Conclusion
Our findings are consistent with the view that humans’
responses to the pain of others can be modulated both
by cognitive and motivational processes. These process-
es are likely to influence whether observing a conspe-
cific in need of help will result in empathic concern, an
important instigator of helping behavior, or personal
distress. These two types of affective responses are
qualitatively distinct and have different motivational
consequences. Empathic concern may instigate an altru-
istic motivation to help the other; personal distress may
produce an egoistic motivation to reduce personal
distress (Batson et al., 1987). It was thus important to
demonstrate both the similarities in the neural networks
underlying the sharing of pain with others and the
specific mechanisms that permit distinguishing the self
from others, which is critical for the experience of em-
pathy (Decety & Hodges, 2006; Decety & Lamm, 2006;
Decety, 2005). Such an experience cannot be identical
to the actual perception of pain because personal and
vicarious experiences differ neurophysiologically as
demonstrated by our behavioral and fMRI data (as also
demonstrated by Craig, 1968, for autonomic nervous
system measurements). Indeed, in order for the sub-
jective experience to be labeled empathy, the observer
must recognize that the emotion she/ he is experiencing
is a response to the other’s emotional state. Finally, our
results demonstrate that both bottom-up (automatic)
and top-down (controlled) processes interact to pro-
duce the experience of empathy. Knowledge about the
context in which the pain experience occurs provides
important clues to the role of top-down cognitive ap-
praisal in the regulation of the vicarious affective pain
reaction.
Acknowledgments
This study was supported by a fellowship from the Fondation
de Recherche Me´dicale (FRM) to C. L. We thank the Acoustics
Research Institute of the Austrian Academy of Sciences for
providing the software S_TOOLS-STx, v3.6.1, and two anony-
mous reviewers for helpful comments on the manuscript.
56 Journal of Cognitive Neuroscience Volume 19, Number 1
Reprint requests should be sent to Jean Decety, Social Cog-
nitive Neuroscience, Department of Psychology, The University
of Chicago, 5848 S University Avenue, Chicago, IL 60637, USA,
or via e-mail: decety@uchicago.edu.
REFERENCES
Baron-Cohen, S., & Wheelwright, S. (2004). The Empathy
Quotient: An investigation of adults with Asperger syndrome
or high functioning autism, and normal sex differences.
Journal of Autism and Developmental Disorders, 34,
163–175.
Batson, C. D., Early, S., & Salvarini, G. (1997). Perspective
taking: Imagining how another feels versus imagining how
you would feel. Personality & Social Psychology Bulletin,
23, 751–758.
Batson, C. D., Fultz, J., & Schoenrade, P. A. (1987). Distress
and empathy: Two qualitatively distinct vicarious emotions
with different motivational consequences. Journal of
Personality, 55, 19–39.
Batson, C. D., Lishner, D. A., Carpenter, A., Dulin, L.,
Harjusola-Webb, S., Stocks, E. L., et al. (2003). ‘‘... As you
would have them do unto you’’: Does imagining yourself
in the other’s place stimulate moral action? Personality &
Social Psychology Bulletin, 29, 1190–1201.
Batson, C. D., Sager, K., Garst, E., Kang, M., Rubchinsky, K.,
& Dawson, K. (1997). Is empathy-induced helping due to
self–other merging? Journal of Personality and Social
Psychology, 73, 495–509.
Beer, J. S., & Heerey, E. A. (2003). The regulatory function
of self-conscious emotion: Insight from patients with
orbitofrontal lesions. Journal of Personality and Social
Psychology, 85, 594–604.
Bellini, L. M., Baime, M., & Shea, J. A. (2002). Variation of
mood and empathy during internship. Journal of the
American Medical Association, 287, 3143–3146.
Blakemore, S.-J., & Frith, C. D. (2003). Self-awareness and
action. Current Opinion in Neurobiology, 13, 219–224.
Blanke, O., & Arzy, S. (2005). The out-of-body experience:
Disturbed self-processing at the temporo-parietal junction.
The Neuroscientist, 11, 16–24.
Botvinick, M., Jha, A. P., Bylsma, L. M., Fabian, S. A., Solomon,
P. E., & Prkachin, K. M. (2005). Viewing facial expression
of pain engages cortical areas involved in the direct
experience of pain. Neuroimage, 25, 312–319.
Bradley, M. M., & Lang, P. J. (2000). Affective reactions to
acoustic stimuli. Psychophysiology, 37, 204–215.
Bush, G., Luu, P., & Posner, M. I. (2000). Cognitive and
emotional influences in anterior cingulate cortex. Trends
in Cognitive Sciences, 4, 215–222.
Chikama, M., McFarland, N. R., Amaral, D. G., & Haber,
S. N. (1997). Insular cortical projections to functional regions
of the striatum correlate with cortical cytoarchitectonic
organization in the primate. Journal of Neuroscience, 15,
9686–9705.
Craig, A. D. (2002). Opinion: How do you feel? Interoception:
The sense of the physiological condition of the body.
Nature Reviews Neuroscience, 3, 655–666.
Craig, K. D. (1968). Physiological arousal as a function of
imagined, vicarious, and direct stress experiences. Journal
of Abnormal Psychology, 73, 513–520.
Craig, K. D., Prkachin, K. M., & Grunau, R. V. E. (2001). The
facial expression of pain. In D. C. Turk & R. Melzack (Eds.),
Handbook of pain assessment (2nd ed., pp. 153–169).
New York: Guilford.
Critchley, H. D., Mathias, C. J., Josephs, O., O’Doherty, J.
Zanini, S., Dewar, B.-K., et al. (2003). Human cingulate
cortex and autonomic control: Converging neuroimaging
and clinical evidence. Brain, 126, 2139–2152.
Critchley, H. D., Wiens, S., Rotshtein, P., O
¨hman, A., & Dolan,
R. D. (2005). Neural systems supporting interoceptive
awareness. Nature Neuroscience, 7, 189–195.
Damasio, A. R., Grabowski, T. J., Bechara, A., Damasio, H.,
Ponto, L. L. B., Parvizi, J., et al. (2000). Subcortical and
cortical brain activity during the feeling of self-generated
emotions. Nature Neuroscience, 3, 1049–1056.
Davis, M. H. (1996). Empathy: A social psychological
approach. Madison, WI: Westview Press.
Decety, J. (2005). Perspective taking as the royal avenue
to empathy. In B. F. Malle & S. D. Hodges (Eds.), Other
minds: How humans bridge the divide between self and
other (pp. 143–157). New York: Guilford Publications.
Decety, J., & Gre`zes, J. (2006). The power of simulation:
Imagining one’s own and other’s behavior. Brain Research,
1079, 4–14.
Decety, J., & Hodges, S. D. (2006). The social neuroscience of
empathy. In P. A. M. Lange (Ed.), Bridging social psychology
benefits of transdisciplinary approaches (pp. 103–109).
Mahwah, NJ: Erlbaum.
Decety, J., & Jackson, P. L. (2004). The functional architecture
of human empathy. Behavioral and Cognitive
Neuroscience Reviews, 3, 71–100.
Decety, J., & Lamm, C. (2006). Human empathy through the
lens of social neuroscience. The Scientific World Journal, 6,
1146–1163.
Decety, J., & Sommerville, J. A. (2003). Shared representations
between self and others: A social cognitive neuroscience
view. Trends in Cognitive Sciences, 7, 527–533.
Deichmann, R., Gottfried, J. A., Hutton, C., & Turner, R.
(2003). Optimized EPI for fMRI studies of the orbitofrontal
cortex. Neuroimage, 19, 430–441.
Derbyshire, S. W. G. (2000). Exploring the pain ‘‘neuromatrix’’.
Current Review of Pain, 4, 467–477.
Doherty, R. W. (1997). The emotional contagion scale: A
measure of individual differences. Journal of Nonverbal
Behavior, 21, 131–154.
Donaldson, D. I., & Buckner, R. L. (2001). Effective
paradigm design. In P. Jezzard, P. M. Mathews, & S. M. Smith
(Eds.), Functional MRI: An introduction to methods
(pp. 177–195). Oxford: Oxford University Press.
Eickhoff, S., Stephan, K. E., Mohlberg, H., Grefkes, C., Fink,
G. R., Amunts, K., et al. (2005). A new SPM toolbox for
combining probabilistic cytoarchitectonic maps and
functional imaging data. Neuroimage, 25, 1325–1335.
Farrer, C., Franck, N., Georgieff, N., Frith, C.D., Decety, J.,
& Jeannerod, M. (2003). Modulating the experience of
agency: A positron emission tomography study.
Neuroimage, 18, 324–333.
Friston, K. J., Fletcher, P., Josephs, O., Holmes, A. P.,
Rugg, M. D., & Turner, R. (1998). Event-related fMRI:
Characterizing differential responses. Neuroimage, 7,
30–40.
Gallagher, H. L., & Frith, C. D. (2003). Functional imaging
of ‘theory of mind’. Trends in Cognitive Sciences, 7,
77–83.
Grosbras, M.-H., & Paus, T. (2006). Brain networks involved
in viewing angry hands or faces. Cerebral Cortex, 16,
1087–1096.
Gross, J. J., & John, O. P. (2003). Individual differences in
two emotion regulation processes: Implications for affect,
relationships, and well-being. Journal of Personality and
Social Psychology, 85, 348–362.
Haxby, J. V., Hoffman, E. A., & Gobbini, M. I. (2000). The
distributed human neural system for face perception.
Trends in Cognitive Sciences, 4, 223–233.
Lamm, Batson, and Decety 57
Jackson, P. L., Brunet, E., Meltzoff, A. N., & Decety, J. (2006).
Empathy examined through the neural mechanisms
involved in imagining how I feel versus how you would
feel pain: An event-related fMRI study. Neuropsychologia,
44, 752–761.
Jackson, P. L., & Decety, J. (2004). Motor cognition: A
new paradigm to study self–other interactions. Current
Opinion in Neurobiology, 14, 1–5.
Jackson, P. L., Meltzoff, A. N., & Decety, J. (2005). How
do we perceive the pain of others: A window into the
neural processes involved in empathy. Neuroimage, 24,
771–779.
Jackson, P. L., Rainville, P., & Decety, J. (2006). To what extent
do we share the pain of others? Insight from the neural
bases of pain empathy. Pain, 125, 5–9.
Kalisch, R., Wiech, K., Critchley, H. D., Seymour, B., O’Doherty,
J. P., Oakley, D. A., et al. (2005). Anxiety reduction through
detachment: Subjective, physiological, and neural effects.
Journal of Cognitive Neuroscience, 17, 874–883.
Lawrence, E. J., Shaw, P., Giampietro, V. P., Surguladze, S.,
Brammer, M. J., & David, A. S. (2006). The role of ‘shared
representations’ in social perception and empathy:
An fMRI study. Neuroimage, 29, 1173–1184.
LeDoux, J. E. (2000). Emotion circuits in the brain. Annual
Reviews of Neuroscience, 23, 155–184.
MacDonald, G., & Leary, M. R. (2005). Why does social
exclusion hurt? The relationship between social pain
and physical pain. Psychological Bulletin, 131, 202–223.
Mesulam, M. M., & Mufson, E. J. (1982). Insula of the old
world monkey: III. Efferent cortical output and comments
on function. Journal of Comparative Neurology, 212,
38–52.
Morosan, P., Rademacher, J., Schleicher, A., Amunts, K.,
Schormann, T., & Zilles, K. (2001). Human primary
auditory cortex: Cytoarchitectonic subdivisions and
mapping into a spatial reference system. Neuroimage,
13, 684–701.
Morrison, I., Lloyd, D., di Pellegrino, G., & Robets, N. (2004).
Vicarious responses to pain in anterior cingulate cortex:
Is empathy a multisensory issue? Cognitive & Affective
Behavioural Neuroscience, 4, 270–278.
Ochsner, K. N., Bunge, S. A., Gross, J. J., & Gabrieli, J. D. E.
(2002). Rethinking feelings: An fMRI study of cognitive
regulation of emotion. Journal of Cognitive Neuroscience,
14, 1215–1229.
Ochsner, K. N., & Gross, J. J. (2005). The cognitive control
of emotion. Trends in Cognitive Sciences, 9, 242–249.
Ochsner, K. N., Ray, R. D., Cooper, J. C., Robertson, E. R.,
Chopra, S., Gabrieli, J. D. E., et al. (2004). For better
or for worse: Neural systems supporting the cognitive
down- and up-regulation of negative emotion. Neuroimage,
23, 483–499.
Ostrowsky, K., Magnin, M., Ryvlin, P., Isnard, J., Guenot, M.,
& Mauguiere, F. (2002). Representation of pain and
somatic sensation in the human insula: A study of responses
to direct electrical cortical stimulation. Cerebral Cortex,
12, 376–385.
Paus, T. (2001). Primate anterior cingulate cortex: Where
motor control, drive and cognition interface. Nature
Reviews Neuroscience, 2, 417–424.
Peyron, R., Laurent, B., & Garcia-Larrea, L. (2000). Functional
imaging of brain responses to pain. A review and
meta-analysis (2000). Neurophysiologie Clinique, 30,
263–288.
Preston, S. D., & de Waal, F. B. M. (2002). Empathy: Its
ultimate and proximate bases. Behavioural and Brain
Sciences, 25, 1–72.
Price, D. D. (1999). Psychological mechanisms of pain and
analgesia. Seattle, WA: IASP Press.
Prkachin, K. M. (2005). Effects of deliberate control on
verbal and facial expressions of pain. Pain, 114, 328–338.
Rogers, T. B., Kuipers, N. A., & Kirker, W. S. (1977).
Self-reference and the encoding of personal information.
Journal of Personality and Social Psychology, 35, 677–688.
Rolls, E. T. (2004). The functions of the orbitofrontal cortex.
Brain and Cognition, 55, 11–29.
Ruby, P., & Decety, J. (2001). Effect of subjective perspective
taking during simulation of action: A PET investigation of
agency. Nature Neuroscience, 4, 546–550.
Ruby, P., & Decety, J. (2003). What you believe versus what
you think they believe: A neuroimaging study of conceptual
perspective-taking. European Journal of Neuroscience,
17, 2475–2480.
Ruby, P., & Decety, J. (2004). How would you feel versus
how do you think she would feel? A neuroimaging study
of perspective-taking with social emotions. Journal of
Cognitive Neuroscience, 16, 988–999.
Singer, T., Seymour, B., O’Doherty, J., Kaube, H., Dolan,
R. J., & Frith, C. D. (2004). Empathy for pain involves the
affective but not the sensory components of pain. Science,
303, 1157–1161.
Singer, T., Seymour, B., O’Doherty, J. P., Stephan, K. E.,
Dolan, R. J., & Frith, C. D. (2006). Empathic neural responses
are modulated by the perceived fairness of others. Nature,
439, 466–469.
Tice, D. M., Bratslavsky, E., & Baumeister, R. F. (2001).
Emotional distress regulation takes precedence over
impulse control: If you feel bad, do it. Journal of Personality
and Social Psychology, 80, 53–67.
Tzourio-Mazoyer, N., Landeau, B., Papathanassiou, D.,
Crivello, F., Etard, O., Delcroix, N., et al. (2002). Automated
anatomical labelling of activations in SPM using a
macroscopic anatomical parcellation of the MNI MRI
single subject brain. Neuroimage, 15, 273–289.
Underwood, B., & Moore, B. (1982). Perspective-taking and
altruism. Psychological Bulletin, 91, 143–173.
Vogt, B. A. (2005). Pain and emotion interactions in subregions
of the cingulate gyrus. Nature Reviews Neuroscience, 6,
533–544.
Wager, T. D., & Feldman Barrett, L. (2004). From affect to
control: Functional specialization of the insula in motivation
and regulation. Retrieved from www.apa.org/psycextra/ on
24 December 2005.
Zald, D. H. (2003). The human amygdala and the emotional
evaluation of sensory stimuli. Brain Research Reviews, 41,
88–123.
58 Journal of Cognitive Neuroscience Volume 19, Number 1
