Joan

Chiao Northwestern University

 Compassion A feeling that arises in witnessing anothers

suering that motivates a subsequent desire to help. (Goetz, Keltner, Simon-Thomas, 2010)

Empathy [responses] that are more other-focused than

self-focused, including feelings of sympathy, compassion, tenderness and the like (Batson, 1991)

culture cultural psychology

mind social / cognitive / aective neuroscience
brain neurogenetics
genes situation / development / evolution
(Chiao & Ambady, 2007; Chiao et al., 2010; Chiao, 2011)

Integrated Model of Human Behavior

B1, B2

Culture Ecological Pressure

A1, A2

Behavior
E1, E2

Gene
C1, C2

Neuroscience
D1, D2

(Chiao & Immordino-Yang, in press, Perspective in Psychological Science)

Integrated Model of Human Behavior Culture Ecological Pressure Gene Neuroscience

Compassion Empathy Sympathy Concern for others

Prosociality

Amygdala ACC Bilateral AI Somatosensory cortex MPFC TPJ

 How does culture inuence neural processes underlying emotion recognition?

(Chiao, et al, 2008, J. Cog. Neurosci.)

Facial Expressions in Blind and Deaf Children

despite no sensory

input, they can smile, cry and laugh

(Eibl-Eibesfeldt, 1973)

fear

contempt

disgust

surprise

happy

anger
(Ekman & Friesen, 1971)

sadness

 Ingroup advantage Display rules Emotional sensitivity Ideal aect Emotion regulation
(Elfenbein & Ambady, 2000; Mesquita & Leu, 2008; Tsai, 2007; Grossman & Kross, 2010; Butler, Mauss, Gross, 2008)

Ingroup advantage in emotion recognition

People recognize fear facial expressions better when expressed by members of their own culture (Elfenbein & Ambady, 2002). People infer nationality better from emotional expressions relative to neutral expressions (Marsh,
Elfenbein & Ambady, 2003).

Ingroup bias likely akin to phonetic and facial recognition biases observed as a result of critical periods in development (Kuhl et al 1992; Pascalis, De Haan, Nelson, 2002).

Caucasian Number of Publications

non-Caucasian

7
6
5
4
3
2
1
0

8

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

Date of Publication (Phan, et al, 2002, Neuroimage; pubmed search post-2002)

Fear expression
Adaptive social signal Warns others of threat Solicits help from others

Amygdala
Subcortical brain region Evaluation and response to

threat cues Orients attention to eye region of fear expressions

(Adolphs, et al, 2005; Davis & Whalen, 2001; Phelps & LeDoux, 2005)

Method: Cross-cultural neuroimaging

10 Japanese
10 Caucasians

(Chiao, et al, 2008, J. Cog. Neurosci.)

Method: Cross-cultural neuroimaging

(Chiao, et al, 2008, J. Cog. Neurosci.)

(Chiao, et al, 2008, J. Cog. Neurosci.)

Cultural Psychology
Cultural specicity in ingroup biases of emotion

recognition are reected as neural tuning of amygdala response.

Brain Sciences
Cultural specicity in amygdala response to fear faces. Amygdala may be part of larger neural network that

facilitates group selection in empathy and altruistic behavior.

 How does culture inuence empathic neural response to emotional scenes?

(Mathur, Harada, Lipke, Chiao, 2010, Neuroimage)

Aective empathy (ACC/AI)

Aect sharing Empathic resonance

Cognitive empathy (MPFC)

Self evaluation Perspective-taking

Experience Self- and Other-pain

Imagine Self- pain

Imagine Other-pain
Singer, et al. (2004), Jackson et al, (2006)

Method
14 African-Americans 14 Caucasian- Americans

Method
Study Task

Rate empathy for target (1 = not at all; 4 = very much) Post-scan measures
-Altruistic motivation indices (Money & Time) -Interpersonal reactivity index (IRI) -Social dominance orientation (SDO) -Implicit racial bias (IAT)

(Mathur, et al., 2010, Neuroimage)

(Mathur, et al., 2010, Neuroimage)

(Mathur, et al., 2010, Neuroimage)

Discussion
Empathic neural response for emotional pain of humankind diers from empathic neural response for emotional pain of group members
Aective empathy (ACC/AI) Irrespective of group membership Automatic Cognitive empathy (MPFC) Enhanced for ingroup members Ingroup identication

Racial Identication and the Default Network

Racial Identication and the

[Northoff et al., 2006: ACC: 3, 11, 44; PCC: 3, 59, 31; MPFC: 2, 50, 10]. Parahippocampal gyri ROIs, for which we did not have a priori predictions, were functionally dened as a 4-mm sphere centered on peak voxels identied by our main effect analysis [Fig. 3: L PHG: 33, 47, 3; R PHG (PHG-specic subcluster of large cluster centered on peak voxel 62, 34, 21): 30, 44, 8]. Small volume correction for multiple comparisons was performed using 6 mm spheres centered on the peak voxels of a priori regions of interest Default Network r within the default network, namely MPFC, ACC, and PCC. MNI coordinates were converted

[Northoff et al., 2006: ACC: 3, 11, 44; PCC: 3, 59, 31; MPFC: 2, 50, 10]. Parahippocampal gyri ROIs, for which we did not have a priori predictions, were functionally dened as a 4-mm sphere centered on peak voxels identied by our main effect analysis [Fig. 3: L PHG: 33, 47, 3; R PHG (PHG-specic subcluster of large cluster centered on peak voxel 62, 34, 21): 30, 44, 8]. Small volume correction for multiple comparisons was performed using 6 mm spheres centered on the peak voxels of a priori regions of interest within the default network, namely MPFC, ACC, and PCC. MNI coordinates were converted
Figure 2. African-American participants display increased activity within cortical midline regions of the default network when viewing same-race others. (a) Whole-brain two-sample comparison [AA participants > CA participants], x 0. Red circles highlight independently dened regions submitted to region of interest analyses. (b) Percent signal change extracted from regions of interest [PCC sphere centered on peak voxel: 3, 59, 31; ACC sphere centered on peak voxel: 3, 11, 44; MPFC sphere centered on peak voxel: 2, 50, 10]. Signal change within ROIs extracted from the [Ingroup > Outgroup] contrast image. **P < 0.001; *P < 0.01. contrast [Ingroup (Pain No Pain) > Outgroup (Pain No Pain)] with a threshold of P < 0.005, extant threshold 10 voxels (Tables I and II and Figs. 2 and 3). Whole-brain regression analyses were performed on the contrast images [Ingroup (Pain No Pain) > Outgroup (Pain No Pain)] and [Outgroup (Pain No Pain) > Ingroup (Pain No Pain)] using racial identication (MEIM score) as the covariate of interest. To test hypotheses about Figure covariate effects, the estimates were comregion-specic 2. African-American participants display increased activity within at P < pared using a linear contrast with signicance levels 0.005, the threshold 10 voxels (Tables III and cortical midline regions of extant default network when viewing IV and same-race others. (a) Figs. 4 and 5). two-sample comparison [AA Whole-brain Region of interest (ROI) analyses were performed on participants > CA participants], x 0.using circles highlight inderegions-of-interest Red MarsBar toolbox in SPM2 [Brett

(Mathur, Harada, Chiao, 2011, Human Brain Mapping)

Figure 3. Caucasian American participants display increased activity within bilateral parahippocampal gyrus when viewing same-race others. (a) Whole-brain two-sample comparison [CA participants > AA participants], z 3. Red circles highlight functionally dened regions submitted to region of interest analyses. (b) Percent signal change extracted from regions of interest [L PHG sphere

When MEIM scores were entered as a covariate in this model, these effects were attenuated. The target race by target pain interaction was no longer signicant F(1,17) 1.15, P 0.30. Controlling for ethnic identication appeared to control for the effect of target race on MPFC response as well, F(1,17) 3.56, P 0.08. ACC. Within the ACC ROI, only the main effects of target race [F(1,18) 9.81, P 0.006] and target pain [F(1,18) 16.39, P 0.001] were signicant. Paired sample t-tests revealed that, on average, participants showed a signicantly greater signal change within the ACC ROI in response to AA (M 0.55, SE 0.07), relative to CA (M Figure 5. Racial identication negatively predicts activity within medial temporal regions of the default network viewing same-race others. (a) Whole-brain regression analysis of group contrast image [Outgroup (Same race Pain Same race No Pain) > r Racial Identication and the Default Network r Ingroup (Same race Pain Same race No Pain)] with individual differences in racial identication as the covariate. (b, c) Inderesponse to AA targets in pain, relative to AA targets in pendent regression analyses [(b) L PHG regression centered no pain [t(19) 1.60, P > 0.05]; and greater MPFC activity around peak voxel: 33, 47, 3; (c) R PHG regression cenin response to CA targets in no pain, relative to CA targets tered around peak voxel: 30, 44, 8]. Signal change within in pain [t(19) 0.69, P > 0.05]. There was also a signi- ROIs extracted from the [Ingroup > Outgroup] contrast image.
cant main effect of target race, F(1,18) 24.92, P < 0.001. Paired sample t-tests revealed that, on average, participants showed a signicantly greater signal change within the MPFC ROI in response to CA (M 0.45, SE 0.12), relative to AA (M 0.09, SE 0.11) scenes (Pain No Pain), t(19) 5.01, P < 0.001. No other comparisons were signicant (all Ps > 0.05). When MEIM scores were entered as a covariate in this model, these effects were attenuated. The target race by target pain interaction was no longer signicant F(1,17) 1.15, P 0.30. Controlling for ethnic identication appeared to control for the effect of target race on MPFC response as well, F(1,17) 3.56, P 0.08.

0.35, SE 0.06) scenes (Pain No Pain) [t(19) 3.18, P 0.005]; and in response to Pain (M 0.68, SE 0.10), relative to No Pain (M 0.22, SE 0.06) [t(19) 4.12, P 0.001]. No other comparisons were signicant (all Ps > 0.05). Controlling for ethnic identication suppressed these effects [Target Race: F(1,17) 0.46, P 0.51; Target Pain: F(1,17) 0.05, P 0.38].

PCC. Within the PCC ROI, only the main effect of target race was signicant, F(1,18) 29.68, P < 0.001. A paired sample t-test revealed that participants showed a signiACC. Within the ACC ROI, only the main effects of target cantly greater signal change within the PCC ROI in Figure 5. race [F(1,18) 9.81, Figure 4. and target pain [F(1,18) P 0.006] response to AA (M 0.37, SE 0.17), relative to CA (M Racial 16.39, P 0.001] were signicant. Paired sample t-tests Racial identication negatively predicts activity within medial identication positively predicts activity within cortical temporal 0.16) of the default No Pain) [t(19) same-race revealed of the default network when showed a signi- 0.08, SE regions scenes (Pain network viewing 5.37, P < midline regions that, on average, participantsviewing same-race others. (a) other comparisons analysis of group (all Ps cantly greater signal change within the ACC ROI in 0.001]. No Whole-brain regressionwere signicant contrast > others. (a) Whole-brain regression analysis of group contrast (Mathur, et al., 2011, Human Brain Mapping) 0.05). image response to AA (M Pain Same racerelative to CAOut- image [Outgroup (Same race Pain Same race No Pain) > [Ingroup (Same race 0.55, SE 0.07), No Pain) > (M Controlling for Pain Same race No attenuated the effect Ingroup (Same race ethnic identication Pain)] with individual group (Other race Pain Other race No Pain)] with individual of target race, though it remained statistically signicant differences in racial identication as the covariate. (b, c) Indedifferences in racial identication as the covariate, x 0. Red F(1,17) regression analyses [(b) L PHG regression centered pendent 5.33, P 0.03.

 How does culture inuence empathic neural response to emotional scenes?

( Chiao, Harada, Mathur, Lipke, 2009, NYAS; Cheon et al, 2011, Neuroimage)

Cultural variation in hierarchy preference

Culture vary in preference for hierarchy Hierarchical relationships may maintain social harmony in some cultures

Method: Cross-cultural neuroimaging

13 Caucasians
14 Koreans

Cultural variation in hierarchy preference predicts empathic neural response

( Chiao, Mathur, Harada, Lipke, 2009, NYAS)

fMRI study design

( Cheon et al, 2011, Neuroimage)

Cultural moderators of empathy

Social Dominance
Social Dominance Orientation Scale (SDO; Pratto et al, 1994) Empathic Concern (Davis, 1983)

Dispositional Empathy

Individualism-Collectivism Self-Construal Scale (SCS; Singelis et al, 1995) Ethnic Identity Display Rules

Multigroup Ethnic Identity Measure (MEIM) Display Rule Inventory (DRAI; Matsumoto et al, 2008) Feeling Thermometer towards Koreans and Caucasian-Americans
( Cheon et al, 2011, Neuroimage)

Explicit Attitudes

Cultural moderators of empathy

Social Dominance
Social Dominance Orientation Scale (SDO; Pratto et al, 1994) Empathic Concern (Davis, 1983)

Dispositional Empathy

Individualism-Collectivism Self-Construal Scale (SCS; Singelis et al, 1995) Ethnic Identity Display Rules

Multigroup Ethnic Identity Measure (MEIM) Display Rule Inventory (DRAI; Matsumoto et al, 2008) Feeling Thermometer towards Koreans and Caucasian-Americans
( Cheon et al, 2011, Neuroimage)

Explicit Attitudes

( Cheon et al, 2011, Neuroimage)

Social dominance orientation predicts ingroup empathy bias

( Cheon et al, 2011, Neuroimage)

Cultural variation in neural basis of intergroup empathy

( Cheon et al, 2011, Neuroimage)

Cultural variation in neural basis of intergroup empathy

( Cheon et al, 2011, Neuroimage)

Discussion
Social dominance orientation is the only predictor of ingroup empathy bias due to neural activity in L TPJ. Cultural variation in social dominance predicts ingroup empathy bias due to cultural variation in L TPJ.
Greater response in L TPJ among Koreans for ingroup empathy bias may represent greater conceptual processing of ingroup members pain (i.e. theory of mind)

Hierarchy-based cultures: Conceptual processing Egalitarian cultures: Simulation processing

(Hein & Singer, 2008, Current Opinion in Neurobiology)

Integrated Model of Human Behavior

SDO, Ethnic ID

Culture Ecological Pressure Gene Neuroscience

Compassion Empathy Sympathy Concern for others

Behavior
Prosociality

Amygdala ACC Bilateral AI Somatosensory cortex MPFC, TPJ

Population variation in mental health provides a natural window into questions that can be addressed cultural neuroscience. Two approaches to closing the disparity gap:
Basic mechanisms Access to treatment

Group selection theory of intergroup conict Understanding how and why groups dier in how they think about themselves and others is key to conict resolution

Mission
To promote the sophistication of theoretical and

methodological approaches in cultural neuroscience

Membership (over 50 scientists from 5 continents)

Anthropologists, psychologists, neuroscientists,

geneticists, epidemiologists, public health experts

Funded by NIH

Tetsuya Iidaka Ahmad Hariri Tokiko Harada Hide Komeda Norihiro Sadato Nalini Ambady

Dong-mi Im Ji-Sook Kim Jason Scimeca Todd Parrish Hyun Wook Park Moshe Bar

Heather Gordon Junpei Nogawa Elissa Amino Mary Helen Immordino-Yang