DECISION NEUROSCIENCE

This page intentionally left blank

DECISIONNEUROSCIENCE

AN INTEGRATIVE PERSPECTIVE

Edited by

JEAN-CLAUDE DREHER

LEON TREMBLAY

Institute of Cognitive Science (CNRS), Lyon, France

AMSTERDAM • BOSTON • HEIDELBERG • LONDON

NEW YORK • OXFORD • PARIS • SAN DIEGO

SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO

Academic Press is an imprint of Elsevier

Academic Press is an imprint of Elsevier125 London Wall, London EC2Y 5AS, United Kingdom525 B Street, Suite 1800, San Diego, CA 92101-4495, United States50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United StatesThe Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom

Copyright © 2017 Elsevier Inc. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical,including photocopying, recording, or any information storage and retrieval system, without permission in writing fromthe publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and ourarrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can befound at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than asmay be noted herein).

Notices

Knowledge and best practice in this field are constantly changing. As new research and experience broaden ourunderstanding, changes in research methods, professional practices, or medical treatment may become necessary.

Practitioners and researchers may always rely on their own experience and knowledge in evaluating and using anyinformation, methods, compounds, or experiments described herein. In using such information or methods they should bemindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for anyinjury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use oroperation of any methods, products, instructions, or ideas contained in the material herein.

Library of Congress Cataloging-in-Publication DataA catalog record for this book is available from the Library of Congress

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library

ISBN: 978-0-12-805308-9

For information on all Academic Press publicationsvisit our website at https://www.elsevier.com/

Publisher: Mara ConnerAcquisition Editor: April FarrEditorial Project Manager: Timothy BennettProduction Project Manager: Edward TaylorDesigner: Matthew Limbert

Typeset by TNQ Books and Journals

Contents

List of Contributors xiPreface xiii

IANIMAL STUDIES ON REWARDS,PUNISHMENTS, AND DECISION-

MAKING

1. Anatomy and Connectivity of theReward CircuitS.N. HABER

Introduction 3

Prefrontal Cortex 4

The Ventral Striatum 5

Ventral Pallidum 10

The Midbrain Dopamine Neurons 11

Completing the CorticaleBasal Ganglia RewardCircuit 14

References 15

2. Electrophysiological Correlates of RewardProcessing in Dopamine NeuronsW. SCHULTZ

Introduction 21

Basics of Dopamine Reward-Prediction ErrorCoding 21

Subjective Value and Formal Economic UtilityCoding 25

Dopamine and Decision-Making 27

Acknowledgments 29

References 29

3. Appetitive and Aversive Systems in theAmygdalaS. BERNARDI AND D. SALZMAN

Introduction 33

Conclusion 41

References 42

4. Ventral Striatopallidal PathwaysInvolved in Appetitive and AversiveMotivational ProcessesY. SAGA AND L. TREMBLAY

Introduction 47

The Corticobasal Ganglia Functional Circuits: TheRelation Between the Cortex and the BasalGanglia 49

The Direct and Indirect Pathways: From Inhibitionof Competitive Movement to AversiveBehaviors 49

Single-Unit Recording in Awake Animals toInvestigate the Neural Bases 50

Tasks to Investigate Appetitive Approach Behaviorand Positive Motivation 51

Ventral Striatum and Ventral PallidumAre Also Involved in Aversive Behaviors:First Evidence of Local Inhibitory Dysfunction 51

The Ventral Striatum and the Ventral PallidumEncode Aversive Future Events for Preparationof Avoidance and for Controlling Anxiety Level 53

Abnormal Aversive Processing Could BiasDecision-Making Toward PathologicalBehaviors 55

Conclusion 55

Acknowledgments 55

References 55

5. Reward and Decision Encodingin Basal Ganglia: Insights FromOptogenetics and Viral Tracing Studiesin RodentsJ. TIAN, N. UCHIDA AND N. ESHEL

Dopamine 59

Striatum 64

Conclusions 67

References 67

v

6. The Learning and Motivational ProcessesControlling Goal-Directed Action and TheirNeural BasesL.A. BRADFIELD AND B.W. BALLEINE

Learning “What Leads to What”: The Neural Basesof ActioneOutcome Learning 72

How Rewards Are Formed: Neural Bases ofIncentive Learning 74

Deciding What to Do: The Neural Bases ofOutcome Retrieval and Choice 76

How These Circuits Interact: Convergence on theDorsomedial Striatum 77

Conclusion 79

Acknowledgments 79

References 79

7. Impulsivity, Risky Choice, and ImpulseControl Disorders: Animal ModelsT.W. ROBBINS AND J.W. DALLEY

Delayed and Probabilistic Discounting: “ImpulsiveChoice” in Decision-Making 82

Premature Responding in the 5CSRTT 83

Stop-Signal Reaction Time 83

Neural Basis of Impulsivity 84

Neural Substrates of Waiting Impulsivity 84

Neural Substrates of Probability Discounting ofReward: Risky Choice 89

Motor Impulsivity: Stop-Signal Inhibition 90

Conclusion 90

Acknowledgments 91

References 91

8. Prefrontal Cortex in Decision-Making:The PerceptioneAction CycleJ.M. FUSTER

The PerceptioneAction Cycle 96

Inputs to the PerceptioneAction Cycle inDecision-Making 97

Prediction and Preparation Toward Decision 99

Execution of Decision 101

Feedback From Decision: Closure of thePerceptioneAction Cycle 102

References 103

IIHUMAN STUDIES ON MOTIVATION,PERCEPTUAL, AND VALUE-BASED

DECISION-MAKING

9. Reward, Value, and SalienceT. KAHNT AND P.N. TOBLER

Introduction 109

Value 110

Salience 113

Conclusions 118

References 118

10. Computational Principles of ValueCoding in the BrainK. LOUIE AND P.W. GLIMCHER

Introduction 121

Value and Choice Behavior 121

Value Coding in Decision Circuits 123

Context Dependence in Brain and Behavior 125

Neural Computations Underlying ValueRepresentation 127

Temporal Dynamics and CircuitMechanisms 130

Conclusions 133

References 133

11. Spatiotemporal Characteristics andModulators of Perceptual Decision-Makingin the Human BrainM.G. PHILIASTIDES, J.A. DIAZ AND S. GHERMAN

Introduction 137

Factors Affecting Perceptual Decision-Making 139

Conclusion 145

References 145

12. Perceptual Decision-Making: What DoWe Know, and What Do We Not Know?C. SUMMERFIELD AND A. BLANGERO

Introduction 149

What Are Perceptual Decisions? 149

Aims and Scope of This Chapter 150

Decision Optimality 150

Q1: How Is Information Integrated DuringPerceptual Decision-Making? 150

Q2: What Computations Do Cortical NeuronsPerform During Perceptual Decisions? 153

Q3: How Can We Study PerceptualDecision-Making in Humans? 155

Q4: How Do Observers Decide Whento Decide? 157

Q5: How Are Perceptual Decisions Biasedby Prior Beliefs? 158

Conclusions and Future Directions 160

Acknowledgments 160

References 160

13. Neural Circuit Mechanismsof Value-Based Decision-Making andReinforcement LearningA. SOLTANI, W. CHAISANGMONGKON AND X.-J. WANG

Introduction 163

Representations of Reward Value 164

Learning Reward Values 165

CONTENTSvi

Stochastic Dopamine-Dependent Plasticity forLearning Reward Values 167

Foraging With Plastic Synapses 167

Random Choice and Competitive Games 170

Probabilistic Inference With Stochastic Synapses 171

Concluding Remarks 173

Acknowledgments 174

References 174

IIISOCIAL DECISION NEUROSCIENCE

14. Social Decision-Making in NonhumanPrimatesM. JAZAYARI, S. BALLESTA AND J.-R. DUHAMEL

Introduction 179

Behavioral Studies on Social Decision-Making 180

Neuronal Correlates of Decision-Making in aSocial Context 182

Conclusions and Perspectives 185

References 185

15. Organization of the Social Brain inMacaques and HumansM.P. NOONAN, R.B. MARS, F.X. NEUBERT, B. AHMED,

J. SMITH, K. KRUG AND J. SALLET

Introduction 189

Medial Prefrontal Cortex 190

Superior Temporal Sulcus and Temporal ParietalJunction 193

A Social Brain Network 194

Summary and Perspectives 195

References 196

16. The Neural Bases of Social Influence onValuation and BehaviorK. IZUMA

Introduction 199

The Effect of Mere Presence of Others on ProsocialBehavior 199

The Effect of Others’ Opinions on Valuation 203

Concluding Remarks 206

References 206

17. Social Dominance Representations inthe Human BrainR. LIGNEUL AND J.-C. DREHER

Introduction 211

Learning Social Dominance Hierarchies 212

Interindividual Differences and SocialDominance 216

Neurochemical Approaches to SocialDominance and Subordination 218

Conclusion 222

Acknowledgments 222

References 222

18. Reinforcement Learning and StrategicReasoning During Social Decision-MakingH. SEO AND D. LEE

Introduction 225

Model-Free Versus Model-Based ReinforcementLearning 226

Neural Correlates of Model-Free ReinforcementLearning During Social Decision-Making 227

Neural Correlates of Hybrid ReinforcementLearning During Social Decision-Making 227

Arbitration and Switching Between LearningAlgorithms 228

Conclusions 230

References 230

19. Neural Control of Social Decisions:Causal Evidence From Brain StimulationStudiesG. UGAZIO AND C.C. RUFF

Introduction 233

Noninvasive Brain Stimulation Methods Used forStudying Social Decisions 234

Brain Stimulation Studies of Social Emotions 235

Brain Stimulation Studies of Social Cognition 236

Brain Stimulation Studies of Social BehavioralControl 238

Brain Stimulation Evidence for Social-SpecificNeural Activity 240

Conclusions 241

References 242

20. The Neuroscience of Compassion andEmpathy and Their Link to ProsocialMotivation and BehaviorG. CHIERCHIA AND T. SINGER

Introduction 247

The “Toolkit” of Social Cognition 248

The Neural Substrates of Empathy 249

The Psychological and Neural Bases of Compassion 250

Conclusion 255

References 255

CONTENTS vii

IVHUMAN CLINICAL STUDIES

INVOLVING DYSFUNCTIONS OFREWARD AND DECISION-MAKING

PROCESSES

21. Can Models of Reinforcement LearningHelp Us to Understand Symptoms ofSchizophrenia?G.K. MURRAY, C. TUDOR-SFETEA AND P.C. FLETCHER

Introduction 261

Reward Processing in Schizophrenia: A HistoricalPerspective 262

Dopamine, Schizophrenia, and ReinforcementLearning 263

The Possible Importance of Glutamate 263

Studies of Reward Processing/ReinforcementLearning in Psychosis: Behavioral Studies 264

Studies of Reward Processing/ReinforcementLearning in Psychosis: Neuroimaging Studies 265

Can an Understanding of Reward and DopamineHelp Us to Understand Symptoms ofSchizophrenia? 269

Summary 272

Acknowledgments 273

References 273

22. The Neuropsychology of Decision-Making: A View From the Frontal LobesA.R. VAIDYA AND L.K. FELLOWS

Introduction 277

Lesion Evidence in Humans 278

Decision-Making and the Frontal Lobes 279

Component Processes of Decision-Making 279

Summary 286

Acknowledgments 286

References 286

23. Opponent Brain Systems for Reward andPunishment Learning: Causal Evidence FromDrug and Lesion Studies in HumansS. PALMINTERI AND M. PESSIGLIONE

The Neural Candidates for Reward andPunishment Learning Systems 294

Evidence From Drug and Lesion Studies 296

Conclusions, Limitations, and Perspectives 300

References 300

24. Decision-Making and Impulse ControlDisorders in Parkinson’s DiseaseV. VOON

Introduction 305

The Role of Dopaminergic Medications andIndividual Vulnerability in Parkinson’s Disease 305

Reinforcing Effects and Associative Learning 307

Learning From Feedback 308

Risk and Uncertainty 309

Impulsivity 310

Summary 311

References 312

25. The Subthalamic Nucleus in ImpulsivityK. WITT

Introduction 315

Anatomy, Physiology, and Function of CorticobasalGanglia Circuits 315

The Subthalamic Nucleus and Decision-Making:Evidence From Animal Studies 317

The Subthalamic Nucleus and Impulsivity:Evidence From Behavioral Observations 317

The Subthalamic Nucleus and Decision-Making:Evidence From Neuropsychological Studies 319

A Model of the Impact of the Subthalamic Nucleuson Decision-Making 321

References 322

26. Decision-Making in Anxiety and itsDisordersD.W. GRUPE

Introduction 327

Summary and Conclusions 335

References 336

27. Decision-Making in Gambling Disorder:Understanding Behavioral AddictionsL. CLARK

Introduction: Gambling and Disordered Gambling 339

Loss Aversion 340

Probability Weighting 341

Perceptions of Randomness 342

Illusory Control 344

Conclusion 345

Disclosures 345

References 345

CONTENTSviii

VGENETIC AND HORMONAL

INFLUENCES ON MOTIVATION ANDSOCIAL BEHAVIOR

28. Decision-Making in Fish: Genetics andSocial BehaviorR.D. FERNALD

Social System of the African Cichlid Fish,Astatotilapia burtoni 351

Domains of Astatotilapia burtoni Social Decisions 353

Summary 358

References 358

29. Imaging Genetics in Humans: MajorDepressive Disorder and Decision-MakingU. RABL, N. ORTNER AND L. PEZAWAS

Introduction 361

Major Depressive Disorder as a Disorderof Decision-Making 361

Imaging Genetics of Major Depressive Disorder 364

Conclusion 365

References 367

30. Time-Dependent Shifts in NeuralSystems Supporting Decision-Making UnderStressE.J. HERMANS, M.J.A.G. HENCKENS, M. JOELS AND

G. FERNANDEZ

Introduction 371

Large-Scale Neurocognitive Systems and Shiftsin Resource Allocation 372

Salience Network and Acute Stress 373

Executive Control Network and Acute Stress 376

Salience Network and Executive Control Networkand Recovery From Stress 377

Summary and Conclusion 378

References 380

31. Oxytocin’s Influence on SocialDecision-MakingA. LEFEVRE AND A. SIRIGU

Introduction 387

Oxytocin and Perception of Social Stimuli 388

Oxytocin and Social Decisions 389

Oxytocin and Social Reward 390

Oxytocin, Learning, and Memory 392

Perspectives 392

Acknowledgments 393

References 393

32. Appetite as Motivated Choice:Hormonal and Environmental InfluencesA. DAGHER, S. NESELILER AND J.-E. HAN

Introduction 397

Appetitive Brain Systems Promote Food Intake 398

Self-Control and Lateral Prefrontal Cortex: Role inAppetite Regulation and Obesity 401

Interaction Between Energy Balance Signals andDecision-Making 402

Conclusion 404

References 404

33. PerspectivesJ.-C. DREHER, L. TREMBLAY AND W. SCHULTZ

Identifying Fundamental ComputationalPrinciples: Produce Conceptual Foundations forUnderstanding the Biological Basis of MentalProcesses Through Development of NewTheoretical and Data Analysis Tools 411

Understanding the Functional Organization of thePrefrontal Cortex and the Nature of theComputations Performed in Various Subregions:Value-Coding Computations 412

Demonstrating Causality: Linking Brain Activity toBehavior by Developing and Applying PreciseInterventional Tools That Change NeuralCircuit Dynamics 413

Maps at Multiple Scales: Generate CircuitDiagrams That Vary in Resolution FromSynapses to the Whole Brain 414

The Brain in Action: Produce a Dynamic Picture ofthe Functioning Brain by Developing andApplying Improved Methods for Large-ScaleMonitoring of Neural Activity 415

The Analysis of Circuits of Interacting Neurons 415

Develop Innovative Technologies andSimultaneous Measures to Understand How theBrain Makes Decisions 416

Advancing Human Decision Neuroscience:Understanding Neurological/PsychiatricDisorders and Treating Brain Diseases 417

Conclusions 418

Reference 418

Index 419

CONTENTS ix

This page intentionally left blank

List of Contributors

B. Ahmed University of Oxford, Oxford, United Kingdom

B.W. Balleine University of Sydney, Camperdown, NSW,Australia

S. Ballesta Centre National de la Recherche Scientifique,Bron, France; Universite Lyon 1, Villeurbanne, France; TheUniversity of Arizona, Tucson, AZ, United States

S. Bernardi Columbia University, New York, New York,United States

A. Blangero University of Oxford, Oxford, United Kingdom

L.A. Bradfield University of Sydney, Camperdown, NSW,Australia

W. Chaisangmongkon King Mongkut’s University ofTechnology Thonburi, Bangkok, Thailand; New YorkUniversity, New York, NY, United States

G. Chierchia Max Planck Institute for Human Cognitive andBrain Sciences, Leipzig, Germany

L. Clark University of British Columbia, Vancouver, BC,Canada

A. Dagher McGill University, Montreal, QC, Canada

J.W. Dalley University of Cambridge, Cambridge, UnitedKingdom

J.A. Diaz University of Glasgow, Glasgow, United Kingdom

J.-C. Dreher Institute of Cognitive Science (CNRS), Lyon,France

J.-R. Duhamel Centre National de la Recherche Scientifique,Bron, France; Universite Lyon 1, Villeurbanne, France

N. Eshel Harvard University, Cambridge, MA, United States

L.K. Fellows McGill University, Montreal, QC, Canada

R.D. Fernald Stanford University, Stanford, CA, UnitedStates

G. Fernandez Radboud University Medical Centre,Nijmegen, The Netherlands

P.C. Fletcher University of Cambridge, Cambridge, UnitedKingdom

J.M. Fuster University of California Los Angeles, LosAngeles, CA, United States

S. Gherman University of Glasgow, Glasgow, UnitedKingdom

P.W. Glimcher New York University, New York, NY, UnitedStates

D.W. Grupe University of WisconsineMadison, Madison,WI, United States

S.N. Haber University of Rochester School of Medicine,Rochester, NY, United States

J.-E. Han McGill University, Montreal, QC, Canada

M.J.A.G. Henckens Radboud University Medical Centre,Nijmegen, The Netherlands

E.J. Hermans Radboud University Medical Centre,Nijmegen, The Netherlands

K. Izuma University of York, York, United Kingdom

M. Jazayari Centre National de la Recherche Scientifique,Bron, France; Universite Lyon 1, Villeurbanne, France

M. Joels University Medical Center Utrecht, Utrecht,The Netherlands

T. Kahnt Northwestern University Feinberg School ofMedicine, Chicago, IL, United States

K. Krug University of Oxford, Oxford, United Kingdom

D. Lee Yale University, New Haven, CT, United States

A. Lefevre Institut des Sciences Cognitives Marc Jeannerod,UMR 5229, CNRS, Bron, France; Universite Claude BernardLyon 1, Lyon, France

R. Ligneul Institute of Cognitive Science (CNRS), Lyon,France

K. Louie New York University, New York, NY, United States

R.B. Mars University of Oxford, Oxford, United Kingdom

G.K. Murray University of Cambridge, Cambridge, UnitedKingdom

S. Neseliler McGill University, Montreal, QC, Canada

F.X. Neubert University of Oxford, Oxford, United Kingdom

M.P. Noonan University of Oxford, Oxford, UnitedKingdom

N. Ortner Medical University of Vienna, Vienna, Austria

S. Palminteri University College London, London, UnitedKingdom; Ecole Normale Superieure, Paris, France

M. Pessiglione Institut du Cerveau et de la Moelle (ICM),Inserm U1127, Paris, France; Universite Pierre et MarieCurie (UPMC-Paris 6), Paris, France

L. Pezawas Medical University of Vienna, Vienna, Austria

M.G. Philiastides University of Glasgow, Glasgow, UnitedKingdom

U. Rabl Medical University of Vienna, Vienna, Austria

T.W. Robbins University of Cambridge, Cambridge, UnitedKingdom

C.C. Ruff University of Zurich, Zurich, Switzerland

Y. Saga Institute of Cognitive Sciences (CNRS), Lyon, France

J. Sallet University of Oxford, Oxford, United Kingdom

xi

D. Salzman Columbia University, New York, New York,United States; New York State Psychiatric Institute, NewYork, New York, United States

W. Schultz University of Cambridge, Cambridge, UnitedKingdom

H. Seo Yale University, New Haven, CT, United States

T. Singer Max Planck Institute for Human Cognitive andBrain Sciences, Leipzig, Germany

A. Sirigu Institut des Sciences Cognitives Marc Jeannerod,UMR 5229, CNRS, Bron, France; Universite Claude BernardLyon 1, Lyon, France

J. Smith University of Oxford, Oxford, United Kingdom

A. Soltani Dartmouth College, Hanover, NH, United States

C. Summerfield University of Oxford, Oxford, UnitedKingdom

J. Tian Harvard University, Cambridge, MA, United States

P.N. Tobler University of Zurich, Zurich, Switzerland

L. Tremblay Institute of Cognitive Science (CNRS), Lyon,France

C. Tudor-Sfetea University of Cambridge, Cambridge,United Kingdom

N. Uchida Harvard University, Cambridge, MA, UnitedStates

G. Ugazio University of Zurich, Zurich, Switzerland

A.R. Vaidya McGill University, Montreal, QC, Canada

V. Voon University of Cambridge, Cambridge, UnitedKingdom; Cambridgeshire and Peterborough NHSFoundation Trust, Cambridge, United Kingdom

X.-J. Wang New York University, New York, NY, UnitedStates; NYU Shanghai, Shanghai, China

K. Witt Christian Albrecht University, Kiel, Germany

LIST OF CONTRIBUTORSxii

Preface

Decision Neuroscience: an Integrative Perspective ad-dresses fundamental questions about how the brainmakes perceptual, value-based, and more complex de-cisions innonsocial andsocial contexts. This bookpresentsrecent and compelling neuroimaging, electrophysiolog-ical, lesional, and neurocomputational studies, in combi-nationwith hormonal and genetic studies, that have led toa clearer understanding of the neural mechanisms behindhow the brain makes decisions. The neural mechanismsunderlying decision-making processes are of criticalinterest to scientists because of the fundamental rolethat reward plays in a number of cognitive processes(such as motivation, action selection, and learning) andbecause they have theoretical and clinical implications forunderstanding dysfunctions of major neurological andpsychiatric disorders.

The idea for this book grew up from our edition of theHandbook of Reward and Decision Making (Academic Press,2009). We originally thought to revise and reedit thisbook, addressing one fundamental question about thenature of behavior: how does the brain process rewardand makes decisions when facing multiple options?However, given the developments in this active area ofresearch, we decided to feature an entirely different bookwith new contents, covering results on the neural sub-strates of rewards and punishments; perceptual, value-based, and social decision-making; clinical aspects suchas behavioral addictions; and the roles of genes andhormones in these various aspects. For example, anexciting topic from the field of social neuroscience is toknow whether the neural structures engaged withvarious forms of social interactions are cause or conse-quence of these interactions (Fernald, Chapter 28).

A mechanistic understanding of the neural encodingunderlying decision-making processes is of great interestto a broad readership because of their theoretical andclinical implications. Findings in this research field arealso important to basic neuroscientists interested in howthe brain reaches decisions, cognitive psychologistsworking on decision-making, as well as computationalneuroscientists studying probabilistic models of brainfunctions. Decision-making covers a wide range of topicsand levels of analysis, from molecular mechanisms toneural systems dynamics, neurocomputational models,and social system levels. The contributions to this book

are forward-looking assessments of the current andfuture issues faced by researchers. We were fortunate toassemble an outstanding collection of experts whoaddressed various aspects of decision-making processes.The book is divided into five parts that address distinctbut interrelated topics.

STRUCTURE OF THE BOOK

Adecision neuroscience perspective requiresmultiplelevels of analyses spanning neuroimaging, electrophysi-ological, behavioral, and pharmacological techniques, incombination with molecular and genetic tools. Theseapproaches have begun to build a mechanistic under-standing of individual and social decision-making. Thisbook highlights some of these advancements that haveled to the current understanding of the neuronal mech-anisms underlying motivational and decision-makingprocesses.

Part I is devoted to animal studies (anatomical,neurophysiological, pharmacological, and optogenetics)on rewards/punishments and decision-making. In theirnatural environment, animals face a multitude of stimuli,very few of which are likely to be useful as predictors ofreward or punishment. It is thus crucial that the brainlearns topredict rewards,providinga critical evolutionaryadvantage for survival. This first part of the book offers acomprehensive view of the specific contributions ofvarious brain structures as the dopaminergic midbrainneurons, the amygdala, the ventral striatum, and theprefrontal cortex, including the lateral prefrontal cortexand the orbitofrontal cortex, to the component processesunderlying reinforcement-guided decision-making, suchas the representation of instructions, expectations, andoutcomes; the updating of action values; and the evalua-tion process guiding choices between prospective re-wards. Special emphasis is made on the neuroanatomy ofthe reward system and the fundamental roles of dopa-minergic neurons and the basal ganglia in learningstimulusereward associations.

Chapter 1 (Haber SN) describes the anatomy andconnectivity of the reward circuit in nonhuman primates.It describes how corticalebasal ganglia loops are

xiii

topographically organized and the key areas of conver-gence between functional regions.

Chapter 2 describes three novel electrophysiologicalproperties of the classical dopamine reward-predictionerror (RPE) signal (Schultz W). Studies have identifiedthree novel properties of the dopamine RPE signal. Inparticular, concerning its roles in making choices, thedopamine RPE signal may not only reflect subjectivereward value and formal economic utility but could alsofit into formal competitive decision models. The RPEsignal may code the chosen value suitable for updatingor immediately influencing object and action values.Thus, the dopamine utility prediction error signalbridges the gap between animal learning theory andeconomic decision theory.

Chapter 3 focuses on the electrophysiological prop-erties of another important component of the rewardsystem in primates, namely the amygdala (Bernardi Sand Salzman D). The amygdala contains distinct appe-titive and aversive networks of neurons. Processing inthese two amygdalar networks can both regulate and beregulated by diverse cognitive operations.

Chapter 4 extends the concept of appetitive andaversive motivational processes to the striatum (Saga Yand Tremblay L). This chapter describes how the ventralstriatum and the ventral pallidum, two parts of thelimbic circuit in the basal ganglia, are involved not onlyin appetitive rewarding behavior, as classically believed,but also in negative motivational behavior. These resultscan be linked with the control of approach/avoidancebehavior in a normal context and with the expression ofanxiety-related disorders. The disturbance of thispathwaymay induce not only psychiatric symptoms, butalso abnormal value-based decision-making.

Chapter 5 (Tian J, Uchida N, and Eshel N) highlightsnew advances in the physiology, function, and circuitmechanism of decision-making, focusing especially onthe involvement of dopamine and striatal neurons. Usingoptogenetics in rodents, molecular techniques, andgenetic techniques, this chapter shows how these toolshave been used to dissect the circuits underlyingdecision-making. It describes exciting new avenues tounderstand a circuit, by recording from neurons withknowledge of their cell type and patterns of connectivity.Furthermore, the ability to manipulate the activity ofspecific neural types provides an important means to testhypotheses of circuit function.

Chapter 6 (Bradfield L and Balleine B) describes theneural bases of the learning and motivational processescontrolling goal-directed action. By definition, the per-formance of such action respects both the current valueof its outcome and the extant contingency between thataction and its outcome. This chapter identifies the neuralcircuits mediating distinct processes, including theacquisition of action-outcome contingencies, the

encoding and retrieval or incentive value, the matchingof that value to specific outcome representations, andfinally the integration of this information for action se-lection. It also shows how each of these individual pro-cesses are integrated within the striatum for successfulgoal-directed action selection.

Chapter 7 (Robbins TW and Dalley JW) describes an-imal models (mostly in rodents) of impulsivity and riskychoices. It reviews the neural and neurochemical basis ofvarious forms of impulsive behavior by distinguishingthree main forms of impulsivity: waiting impulsivity,risky choice impulsivity, and stopping impulsivity. Itshows that dopamine- and serotonin-dependent func-tions of the nucleus accumbens are implicated in waitingimpulsivity and risky choice impulsivity, as well ascortical structures projecting to the nucleus accumbens.For stopping impulsivity, dopamine-dependent functionsof the dorsal striatum are implicated, as well as circuitryincluding the orbitofrontal cortex and dorsal prelimbiccortex. Differences and commonalities between the formsof impulsive responding are highlighted. Importantly,various applications to human neuropsychiatric disor-ders such as drug addiction and attention deficit hyper-activity disorder are also discussed.

Chapter 8 (Fuster JM) proposes that the neuralmechanisms of decision-making are understandableonly in the structural and dynamic context of theperceptioneaction cycle, defined as the biocyberneticprocessing of information that adapts the organism to itsenvironment. It presents a general view of the role of theprefrontal cortex in decision-making, in the generalframework of the perceptioneaction cycle, includingprediction, preparation toward decision, execution, andfeedback from decision.

Part II covers the topic of the neural representation ofmotivation, perceptual decision-making, and value-based decision-making in humans, mostly combiningneurocomputational models and brain imaging studies.

Chapter 9 (Tobler P and Kahnt T) reviews severaldefinitions of value and salience, and describes humanneuroimaging studies that dissociate these variables.Value increases with the magnitude and probability ofreward but decreases with the magnitude and proba-bility of punishment, whereas salience increases with themagnitude and probability of both reward and punish-ment. At the neural level, value signals arise in striatum,orbitofrontal and ventromedial prefrontal cortex, andsuperior parietal areas, whereas magnitude-basedsalience signals arise in the anterior cingulate cortexand the inferior parietal cortex. By contrast, probability-based salience signals have been found in the ventro-medial prefrontal cortex.

Chapter 10 (Louie K and Glimcher PW) reviews anapproach centered on basic computations underlying

PREFACExiv

neural value coding. It proposes that neural informationprocessing in valuation and choice relies on computa-tional principles such as contextual modulation anddivisive normalization. Divisive normalization is anonlinear gain control algorithm widely observed inmultiple sensory modalities and brain regions. Identifi-cation of these computations sheds light on how theunderlying neural circuits are organized, and neuralactivity dynamics provides a link between biologicalmechanism and computations.

Chapter 11 (Philiastides M, Diaz J, and Gherman S)introduces the general principles guiding perceptualdecision-making. Perceptual decisions occur whenperceptual inputs are integrated and converted to form acategorical choice. It reviews the influence of a number offactors that interact and contribute to the decision pro-cess, such as prestimulus state, reward and punishment,speedeaccuracy trade-off, learning and training, confi-dence, and neuromodulation. It shows how these deci-sion modulators can exert their influence at variousstages of processing, in line with predictions derivedfrom sequential-sampling models of decision-making.

Chapter 12 (Summerfield C) reviews the neural andcomputational mechanisms of perceptual decisions. Itaddresses current controversial questions, such as howwe decide when to draw our decisions to a conclusion,and how perceptual decisions are biased by priorinformation.

Chapter 13 (Soltani A, Chaisangmongkon W, andWang XJ) presents possible biophysical and circuitmechanisms of valuation and reward-dependent plas-ticity underlying adaptive choice behavior. It reviewsmathematical models of reward-dependent adaptivechoice behavior, and proposes a biologically plausible,reward-modulated Hebbian synaptic plasticity rule. Itshows that a decision-making neural circuit endowedwith this learning rule is capable of accounting forbehavioral and neurophysiological observations in avariety of decision-making tasks.

Part III of the book focuses on the rapidly developingfield of social neuroscience, integrating neurosciencedata from both nonhuman primates and humans. Pri-mates are fundamentally social animals, and they mayshare common neural mechanisms in diverse forms ofsocial behavior. Examples of such behavior includetracking intentions and beliefs from others, beingobserved by others during prosocial decisions, orlearning the social hierarchy in a group of individuals. Itis also likely that at the macroscopic level, importantdifferences exist concerning social brain structures andconnectivity, and there is a need to directly comparebetween species to answer this fundamental question.Indeed, studies in both humans and monkeys report notonly an increase in gray matter density of specific brain

structures relative to the size of our social network, butalso species differences in prefrontaletemporal brainconnectivity. Furthermore, this part of the book presentsneurocomputational approaches starting to provide amechanistic understanding of social decisions. Forexample, reinforcement learning models and strategicreasoning models can be used when learning social hi-erarchies or during social interactions.

A social neuroscience understanding requires multi-ple approaches, such as electrophysiology and neuro-imaging in both monkeys (Chapters 14, 15, 19) andhumans (Chapters 16, 18, 20), as well as causal (Chapter21), neurocomputational (Chapters 17e19), endocrino-logical, genetics, and clinical approaches (Part V).

Chapter 14 (Duhamel JR and colleagues) presentsmonkey electrophysiological data revealing that theorbitofrontal cortex is tuned to social information. Forexample, in one experiment, macaque monkeys workedto collect rewards for themselves and two monkeypartners. Single neurons encoded the meaning of visualcues that predicted the magnitude of future rewards, themotivational value of rewards obtained in a socialcontext, and the tracking of social preferences and part-ner’s identity and social rank. The orbitofrontal cortexthus contains key neuronal mechanisms for the evalua-tion of social information. Moreover, macaque monkeystake into account the welfare of their peers when makingbehavioral choices bringing about pleasant or unpleasantoutcomes to a monkey partner. Thus, this chapter revealsthat prosocial decision-making is sustained by anintrinsic motivation for social affiliation and controlledthrough positive and negative vicarious reinforcements.

Chapter 15 (Sallet J and colleagues) reviews the sim-ilarities between monkeys and humans in the organiza-tion of the social brain. Using MRI-based connectivitymethods, they compare human and macaque socialareas, such as the organization of the medial prefrontalcortex. They revealed that the connectivity fingerprint ofmacaque area 10 best matched that of the human frontalpole, suggesting that even high-level areas share featuresbetween species. They also showed that animals housedin large social groups had more gray matter volume inbilateral mid-superior temporal sulcus and rostral pre-frontal cortex. Beyond species similarities, there are alsodistinct differences between human and macaqueprefrontaletemporal brain connectivity. For example,functional connections between the temporal cortex andthe lateral prefrontal cortex are stronger in humanscompared to connections with the medial prefrontalcortex in humans, but the opposite pattern is observed inmacaques.

Chapter 16 (Izuma K) focuses on two forms of socialinfluence, the audience effect, which is an increasedprosocial tendency in front of other people, and socialconformity, which consists in adjusting one’s attitude or

PREFACE xv

behavior to those of a group. This chapter discusses fMRIfindings in healthy humans in these two types of socialinfluence and also shows how reputation processing isimpaired in individuals with autism. It also links socialconformity and reward-based learning (reinforcementlearning).

Chapter 17 (Ligneul R and Dreher JC) examines howthe brain learns social dominance hierarchies. Socialdominance refers to relationships wherein the goals of oneindividual prevail over the goals of another individual in asystematic manner. Dominance hierarchies have emergedas a major evolutionary force to drive dyadic asymmetriesin a social group. This chapter proposes that the emer-gence of dominance relationships are learned incremen-tally, by accumulating positive and negative competitivefeedbacks associated with specific individuals and othermembers of the social group. It considers such emergenceof social dominance as a reinforcement learning probleminspired by neurocomputational approaches traditionallyapplied to nonsocial cognition. This chapter also reportshow dominance hierarchies induce changes in specificbrain systems, and it reviews the literature on interindi-vidual differences in the appraisal of social hierarchies, aswell as the underlying modulations of cortisol, testos-terone, and serotonin/dopamine systems, which mediatethese phenomena.

Chapter 18 (Seo H and Lee D) describes reinforcementlearning models and strategic reasoning during socialdecision-making. It shows that dynamic changes inchoices and decision-making strategies can be accountedfor by reinforcement learning in a variety of contexts. Thisframework has also been successfully adopted in a largenumber of neurobiological studies to characterize thefunctions of multiple cortical areas and basal ganglia. Forcomplex decision-making, including social interactions,this chapter shows that multiple learning algorithmsmayoperate in parallel.

Chapter 19 (Ugazio G and Ruff C) reports brainstimulation studies on social decision-making, which testthe causal relationship between neural activity anddifferent types of processes underlying these decisions,including social emotions, social cognition, and socialbehavioral control.

Chapter 20 (Chierchia G and Singer T) shows that twoimportant social emotions, empathy and compassion,engage distinct neurobiological mechanisms, as well asdifferent affective and motivational states. Empathy forpain engages a network including the anterior insula andanterior midcingulate cortex, areas associated withnegative affect; compassionate states engage the medialorbitofrontal cortex and ventral striatum and are associ-atedwith feelings ofwarmth, concern, and positive affect.

Part IV of the book focuses on clinical aspectsinvolving disorders of decision-making and of the

reward system that link together basic research areas,including systems, cognitive, and clinical neuroscience.Dysfunction of the reward system and decision-makingis present in a number of neurological and psychiatricdisorders, such as Parkinson’s disease, schizophrenia,drug addiction, and focal brain lesions. The study ofpathological gambling, for example, and other motivatedstates associated with, and leading to, compulsivebehavior provides an opportunity to learn about thedysfunctions of reward system activity, independent ofdirect pharmacological activation of brain reward cir-cuits. On the other hand, because drugs of abuse directlyactivate brain systems, they provide a unique challengein understanding how pharmacological activation in-fluences reward mechanisms leading to persistentcompulsive behavior.

Chapter 21 (Murray GK, Tudor-Sfetea C, and FletcherPC) shows that principles of reinforcement learning areuseful to understand the neural mechanisms underlyingimpaired learning, reward, and motivational processesin schizophrenia. Two symptoms characteristic of thisdisease is considered in this framework, namely de-lusions and anhedonia.

Chapter 22 (Vaidya AR and Fellows LK) takes aneuropsychological approach to review focal frontal lobedamage effects on value-based decisions. It reveals thenecessary contributions of specific subregions (ventro-medial, lateral, and dorsomedial prefrontal cortex) todecision-making, and provides evidence as to the disso-ciability of component processes. It argues that theventromedial frontal lobe is required for optimal learningfrom reward under dynamic conditions and contributesto specific aspects of value-based decision-making. It alsoshows a necessary contribution of the dorsomedialfrontal lobe in representing action-value expectations.

Chapter 23 (Palminteri S and Pessiglione M) reviewsreinforcement learning models applied to reward andpunishment learning. These studies include fMRI andneural perturbation following drug administration and/or pathological conditions. They propose that distinctbrain systems are engaged, one in reward learning(midbrain dopaminergic nuclei and ventral prefrontos-triatal circuits) and another in punishment learning,revolving around the anterior insula.

Chapter 24 (Voon V) discusses decision-making im-pairments and impulse control disorders in Parkinson’sdisease. The author reports enhancement of the gainassociated with levodopa, reinforcing properties ofdopaminergic medications, and enhancement of delaydiscounting in these patients. Lower striatal dopaminetransporter levels preceding medication exposure, anddecreased midbrain D2 autoreceptor sensitivity, mayunderlie enhanced ventral striatal dopamine release andactivity in response to salient reward cues, anticipatedand unexpected rewards, and gambling tasks.

PREFACExvi

Impairments in decisional impulsivity (delay discount-ing, reflection impulsivity, and risk taking) implicate theventral striatum, orbitofrontal cortex, anterior insula,and dorsal cingulate. These findings provide insight intothe role of dopamine in decision-making processes inaddiction and suggest potential therapeutic targets.

Chapter 25 (Witt K) reports that motor control is theresult of a balance between activation and inhibition ofmovement patterns. It points to a central role of thesubthalamic nucleus within the indirect basal gangliapathway, acting as a brake on the motor system. Thissubthalamic nucleus function occurs when an automaticresponse must be suppressed to have more time tochoose between alternative responses.

Chapter 26 (Grupe DW) discusses value-baseddecision-making as one of a key behavioral symptomspresent in anxiety disorders. This chapter highlights al-terations to specific processes: decision representation,valuation, action selection, outcome evaluation, andlearning. Distinct anxious phenotypes may be charac-terized by differential alterations to these processes andtheir associated neurobiological mechanisms.

Chapter 27 (Clark L) presents a conceptualization ofdisordered gambling as a behavioral addiction driven byan exaggeration of multiple psychological distortionsthat are characteristic of human decision-making, andunderpinned by neural circuitry subserving appetitivebehavior, reinforcement learning, and choice selection.The chapter discusses the neurobiological basis of path-ological gambling behavior in loss aversion, probabilityweighting, perceptions of randomness, and the illusionof control.

Part V focuses on the roles of hormones and genesinvolved in motivation and social decision-making pro-cesses. The combination of molecular genetic, endocri-nology, and neuroimaging has provided a considerableamount of data that help in the understanding of thebiological mechanisms influencing decision processes.These studies have demonstrated that genetic and hor-monal variations have an impact on the physiologicalresponse of the decision-making system. These varia-tions may account for some of the inter- and intra-individual behavioral differences observed in socialcognition.

Chapter 28 (Fernald RD) presents an originalapproach for cognitive neuroscientists by focusing on thedifficult question of how an animal’s behavior orperception of its social and physical surroundings shapesits brain. Using a fish model system that depends oncomplex social interactions, this chapter reports how thesocial context influences the brain and, in turn, alters thebehavior and neural circuitry of animals as they interact.Gathering of social information vicariously producesrapid changes in gene expression in key brain nuclei and

these genomic responses prepare the individual tomodify its behavior to move into a different social niche.Both social success and failure produce changes inneuronal cell size and connectivity in key brain nuclei.This approach bridges the gap between social informa-tion gathering from the environment and the levels ofcellular and molecular responses.

Chapter 29 (Rabl U, Ortner N, and Pezawas L) ex-amines the use of imaging genetics to explore the re-lationships between major depressive disorder anddecision-making.

Chapters 30e32 report neuroendocrinological find-ings in social decision-making, likening variations in thelevels of different types of hormones (cortisol, oxytocin,ghrelin/leptin) to brain systems engaged in social de-cisions and food choices. Chapter 30 (Hermans EJ andcolleagues) integrates knowledge of the effects of stressat the neuroendocrine, cellular, brain systems, andbehavioral levels to quantify how stress-related neuro-modulators trigger time-dependent shifts in the balancebetween two brain systems: a “salience” network, whichsupports rapid but rigid decisions, and an “executivecontrol” network, which supports flexible, elaborate de-cisions. This simple model elucidates paradoxical find-ings reported in human studies on stress and cognition.

Chapter 31 (Lefevre A and Sirigu A) reviews evidencefor a role for oxytocin in individual and social decision-making. It discusses animal and human studies to linkthe behavioral effects of oxytocin to its underlyingneurophysiological mechanisms.

Chapter 32 (Dagher A, Neseliler S, and Han JE) ex-amines the neurobehavioral factors that determine foodchoices and food intake. It reviews findings on the in-teractions between brain systems that mediate feedingbehavior and the gut and adipose peptides that signalthe current state of energy balance.

Chapter 33 (Dreher, Tremblay, and Schultz) concludesthis decision neuroscience book by integrating perspec-tives from all contributors.

We anticipate that while some readers may read thevolume from the first to the last chapter, other readersmay read only one or more chapters at a time, and notnecessarily in the order presented in the book. This iswhy we encouraged an organization of this volumewhereby each chapter can stand alone, while makingreferences to others andminimizing redundancies acrossthe volume. Given the consistent acceleration of ad-vances in the various approaches described in this bookon decision neuroscience, you are about to be dazzled bya first look at the new stages of an exciting era in brainresearch. Enjoy!

Jean-Claude DreherLeon Tremblay

PREFACE xvii

This page intentionally left blank

C H A P T E R

17

Social Dominance Representations inthe Human BrainR. Ligneul, J.-C. Dreher

Institute of Cognitive Science (CNRS), Lyon, France

AbstractSocial dominance refers to relationships wherein the goalsof one individual prevail over the goals of another indi-vidual in a systematic manner. Dominance hierarchies haveemerged as a major evolutionary force to drive dyadicasymmetries in a social group. Understanding how thebrain detects, represents, implements, and monitors socialdominance hierarchies constitutes a fundamental topic forsocial neuroscience as well as a major challenge for thefuture of clinical psychiatry. In this chapter, we argue thatthe emergence of dominance relationships is learnedincrementally, by accumulating positive and negativecompetitive feedback associated with specific individualsand other members of the social group. We consider suchemergence of social dominance as a reinforcement learningproblem inspired by neurocomputational approachestraditionally applied to nonsocial cognition. We also reporthow dominance hierarchies induce changes in specificbrain systems, and we review the literature on interindi-vidual differences in the appraisal of social hierarchies, aswell as the underlying modulations of the cortisol, testos-terone, and serotonin/dopamine systems that mediatethese phenomena.

INTRODUCTION

Social Hierarchies in Health and Well-Being

Social dominance refers to situations in which an “in-dividual or a group controls or dictates others’ behavior,primarily in competitive situations” [1,2]. Social domi-nance hierarchies influence access to resources and mat-ing partners and therefore constitute a potent biologicalforce binding together social behavior, well-being, andevolutionary success. The concept of social dominanceis most often applied to “learned relationships,” shapedby a history of social victories and defeats within dyads

of individuals [3]. Together with other forms of power,social dominance asymmetries constitute a pivotalconcept for understanding social organizations and pre-dicting individual behaviors.

Many animal studies indicate that iterated social de-feats can trigger maladaptive social avoidance, behav-ioral inhibition, elevated glucocorticoid levels, andhigher vulnerability to addiction, anxiety, or depression[4e7]. Epidemiological approaches in humans have sub-sequently confirmed that suffering from a chronicallylow socioeconomic status or enduring transient status-lowering threats facilitates both somatic and psychiatricdisorders [8,9]. Unfortunately, it has long been difficultto disentangle the specific contributions of stress, socio-economic status, and social dominance on the humanbrain. In particular, very little is known about the cere-bral mechanisms governing the progressive establish-ment of social dominance hierarchies and associatedneurobehavioral changes through real-life interactions(for a review of such mechanisms in nonhuman pri-mates and of genetic mechanisms in zebra finches, seeChapters 15 and 28).

In this chapter, we will first consider the learning ofsocial dominance as an incremental process, allowingus to develop a neurocomputational approach to akey decision problem (i.e., to initiate or not a compet-itive interaction). Second, we will review importantinterindividual differences that naturally derive andinfluence social dominance hierarchies. Third, wewill highlight the tight relationship of social domi-nance with stress and neuroplasticity [6] by reportingits effects on the hypothalamicepituitaryeadrenal(HPA)/hypothalamicepituitaryegonadal (HPG)axes as well as on the serotonin and dopaminesystems [10].

211Decision Neuroscience

Copyright © 2017 Elsevier Inc. All rights reserved.http://dx.doi.org/10.1016/B978-0-12-805308-9.00017-8

Social Hierarchies as a Major EvolutionaryPressure and Pivotal Feature of Societies

In folk psychology, dominance is often considered afundamental motive of social organisms. However,while many primatologists agree that social rankscorrelate positively with offspring production inmany primate and nonprimate species, effect sizesare usually small and many counterexamples exist,indicating that subordinate individuals often achievedecent reproductive success [11,12]. Moreover, domi-nance hierarchies spontaneously reemerge, even ifonly subordinate or only dominant individuals areput together to form a new social group at each gener-ation [13,14], implying that the social environmentdynamically tunes individuals’ brains to promote thebest behavioral strategy given the social context,through synaptic and epigenetic plasticity mecha-nisms. Dominance and subordination can thus be betterdescribed as life-history strategies [15,16], because bothconstitute adaptations to the social environment andboth can increase evolutionary fitness.

The importance of social dominance for domain-general cognition is primarily rooted in the so called “so-cial brain hypothesis,” which stipulates that the need tooptimize behavior within complex social environmentslargely constrained the evolution of the primate brain.This theory was first outlined in the pioneering studyof Alison Jolly in lemurs [17], and it was popularizedby Humphrey [18], Byrne and Whiten [19], and Dunbar[20], who provided correlational evidence for the coevo-lution of social complexity and various markers of braindevelopment. In evolved animals, other group membersconstitute stochastically behaving entities guided byhidden mental states and obeying complex sets of rules.Consequently, predicting their actions and improvingour transactions with them requires very elaboratedcomputations which have long been overlooked incognitive neurosciences. A similar argument alsoapplies to simpler behaviors: the neural mechanismthat enables human subjects to avoid selecting a bluesquare associated with electric shock delivery in thelab might have partly been sculpted, throughout evolu-tion, to enable efficient avoidance of aggressive domi-nant individuals within one’s social group. Thepossibility that “social life provided the evolutionarycontext of primate intelligence” [17] is thus key to fore-seeing the importance of social dominance for domain-general, high level human cognition.

Previously, a large neural circuit activated when peo-ple make decisions in social settings has been identifiedusing functional magnetic resonance imaging (fMRI) inhumans. Key components include the orbitofrontal cor-tex, the ventromedial prefrontal cortex (VMPFC), thedorsomedial and dorsolateral prefrontal cortex(DLPFC), parts of the superior temporal sulcus (STS)

including a region near the temporoparietal junction(TPJ), and the anterior cingulate gyrus. In sum, there isnow extensive evidence that social decision-making re-lies on many “nonsocial” subcortical and brain stem cir-cuits. Social decision-making also overtakes thecanonical cortical network of social cognition encom-passing the medial prefrontal cortex (MPFC), the STS,and the TPJ (Fig. 17.1A).

Since 2005, a number of fMRI studies have investi-gated the perception of social ranks based on noncom-petitive cues, such as wealth [23,24], postures [25],uniforms [23], facial traits [26], and celebrity, height,or intelligence [27,28]. Completing the pioneeringwork of Zink et al. [29], these studies have also demon-strated the engagement of a large brain networkinvolved in social hierarchy processing, including theamygdala, hippocampus, striatum, ventrolateral pre-frontal cortex (VLPFC), rostromedial prefrontal cortex(RMPFC), inferior parietal lobule (IPL), and the fusi-form gyrus (Fig. 17.1BeC). Although these perceptualprocesses linked with social dominance raise manyimportant questions (for a review see [30]), we willfocus in this chapter on the neurocomputational pro-cesses that underlie the learning of social dominance sta-tuses (SDSs).

LEARNING SOCIAL DOMINANCEHIERARCHIES

In what follows, we will propose that one’s own domi-nance status is learned incrementally by accumulatingthe numerous competitive feedbacks (victories and de-feats) obtained against other group members. Put sim-ply, individuals who experience on average negativefeedback following competitive encounters will developan adaptive subordinate profile, which may take themaway from social conflicts by promoting submission,and vice versa for dominant individuals. Importantly,the same principle can also be applied to the rapid up-date of others’ SDS during competitive interactions.These assumptions allowed us to develop a neurocom-putational approach to characterize the emergence of so-cial dominance relationships through time and providecomputational neuroscientists with an adequate frame-work to test quantitative and mechanistic hypothesesabout this process [21,31].

Reinforcement Learning Approaches toSocial Cognition

Mathematical models are increasingly used in socialneuroscience as they can probe, simultaneously, severalcognitive processes which would otherwise not be sepa-rable [21,31] (see also Chapters 18 and 19). Three main

17. SOCIAL DOMINANCE REPRESENTATIONS IN THE HUMAN BRAIN212

III. SOCIAL DECISION NEUROSCIENCE

subsystems have been unraveled by this approach: theVMPFC, together with the ventral striatum, may beresponsible for observational learning and reward pre-diction error signaling [32,33]. Second, the anteriorcingulate cortex, the DLPFC, and the IPL may computeprediction errors elicited when others’ behaviors deviatefrom our predictions about them. Third, the RMPFC andTPJ/posterior STS may compute the updating of one’sown and others’ mental states, [33,34] as well as the de-gree of interpersonal influence during social interactions,that is, the degree to which one’s own actions and utili-ties are determined by others’ skills and strategies [35].

Within iterated competitive games, rmPFC activitywas often shown to encode second-order variables(i.e., variables inferred from other learned variables)[31,34,36]. For example, in the inspector game or in the

matching pennies task respectively used in Hamptonet al. [34] and Seo et al. [36], a player able to take into ac-count the influence of his or her past choices over theevolution of another’s strategy will be able to increasehis or her chances of winning to the detriment of theother, because this second-order information enablesthe player to better predict the other’s choice (seeChapter 18 from Lee et al.). Given that the ability to con-trol others’ behaviors and outcomes is at the core of thesocial dominance concept, it would be tempting to char-acterize such player as “mentally dominating” his or heropponent. Interestingly, if the “subordinate” opponentstarts to play on a purely random basis, the opportunityfor interpersonal control disappears, possibly contrib-uting to the frequency of inconsistent or irrational socialbehaviors in humans. Moreover, the consequences of

FIGURE 17.1 (A) The functional neuroanatomy of social behavior. Primary colors denote brain regions activated by reward and valuation,frequently identified in studies of social interaction within the frame of reference of the subject’s own actions: anterior cingulate cortex sulcus(ACCs), ventromedial prefrontal cortex (VMPFC), amygdala, and ventral striatum (VStr). Pastels denote brain regions activated by considering theintentions of another individual: anterior cingulate cortex gyrus (ACCg), dorsomedial prefrontal cortex (DMPFC), temporoparietal junction (TPJ ),and superior temporal sulcus (STS). (From Behrens TEJ, Hunt LT, Rushworth MFS. The computation of social behavior. Science 2009;324:1160e64, with

permission.) (B) Cerebral substrates of social comparison processes. Comparative judgments about the height or the intelligence of others activatespecifically the anterior prefrontal cortex, the amygdala, and the TPJ. (From Lindner M, Hundhammer T, Ciaramidaro A, Linden DEJ, Mussweiler T. Theneural substrates of person comparisonean fMRI study. Neuroimage 2008;40:963e71, with permission.) (APCC, anterior paracingulate cortex; prMFC,posterior medial prefrontal cortex; arMFC, anterior portion of the rostral medial frontal cortex; OFC, orbitofrontal cortex; oMFC, orbital medialprefrontal cortex) (C) Statistical maps of the brain regions more engaged in the comparison “superior player > inferior player.” Compared toinferior individuals, the perception of superior individuals elicited stronger activations inmany brain regions including the dorsolateral prefrontalcortex, the medial prefrontal cortex, the striatum, and the occipitoparietal cortices. (From Zink CF et al. Know your place: neural processing of social

hierarchy in humans. Neuron 2008;58:273e83, with permission.)

III. SOCIAL DECISION NEUROSCIENCE

LEARNING SOCIAL DOMINANCE HIERARCHIES 213

FIGURE 17.2 The emergence of social dominance through reinforcement learning. (A) Typical trial and experiment time courses for the

fMRI experiment. During 15 trials of a “miniblock,” subjects played against (or with) the same player in the competitive (or control) situation. Thecompetitive task required subjects to evaluate a series of stationary arrows, indicating which direction the majority of these arrows pointed (left orright). The task was performed against one of three virtual opponents implicitly associated with three frequencies of winning and losing. Tosucceed in the competition, subjects were instructed to answer accurately and faster than their opponent. (B) Typical trial and experiment time

courses of the brain stimulation experiment. Subjects performed a similar perceptual task but now opponents were marked by visual symbolsand artificial names rather than a face photograph. Subjects could now choose which opponent to defy among two alternatives (three opponentsper block), in two types of trials designed to distinguish dominance-based (spontaneous) and reward-based (control) choices. In half of thesubjects, the rostromedial prefrontal cortex (RMPFC) was monitored with the excitatory anodal electrode of the transcranial direct currentstimulation (tDCS) apparatus (magenta; the reference electrode on the vertex is in blue). (C) Striatal encoding of competitive defeats.Competition-specific outcome signals revealed by the interaction competitive victory and control failure> competitive defeat and control success

III. SOCIAL DECISION NEUROSCIENCE

17. SOCIAL DOMINANCE REPRESENTATIONS IN THE HUMAN BRAIN214

this interaction between social dominance and mentali-zation of interpersonal influence within strategic gameshas also been explored based on evolutionary reinforce-ment learning models [37], which showed that mentaliz-ing abilities may indeed promote social status when therate of cooperation among group members is low.

Finally, neuroimaging studies indicate that theRMPFC (as well as the TPJ and the medial STS) is gener-ally more engaged when subjects are invited to competeagainst other humans compared to computers[29,38,39,40e42], suggesting that the RMPFC might beinvolved in representing the mental states of othersand/or the dominance relationship emerging betweentwo individuals. Competitive interactions also tend tolead to higher activity of the RMPFC compared to coop-erative interactions, which indicates a possible competi-tion specificity of the cognitive processes implementedin this structure [40,43].

The Emergence of Dominance andSubordination by Reinforcement Learning

According to the pioneering theory of Bernstein [3],dominance relationships are learned progressively, byintegrating the positive and negative competitive feed-back associated with specific individuals and othermembers of the social group taken as a whole. Acompetitive outcome thus provides information aboutothers’ behaviors (which underlies the representationof objective social hierarchies) and information aboutoneself (which may lead to adjustments in subjectivesocial status and related variables such as self-esteem).In other words, the former information may fostertarget- or dyad-specific dominance behaviors while thelatter may foster target-independent behavioral profilesof dominance and subordination with individuals.Although both can be described by a similar reinforce-ment learning scheme, the time constants (and learningrate) associated with those two processes might bedifferent, as self-representations intuitively appear lessvolatile than representations about others, in mostpeople.

By learning social dominance representations, indi-viduals can anticipate their probability of winningversus losing and therefore decide whether they shouldcarry on fighting or disengage from a confrontation inorder to limit the physical and social costs associatedwith a defeat. Outside of agonistic interactions, moni-toring such probability can also prevent the escalationof social conflicts for which the risks of losing outweighthe expected benefits of winning. Interestingly, in thecostebenefit trade-off that underlies the decision tocompete against another conspecific, the cost is typicallyassociated with a property of the opponent (i.e., his orher strength or skill, which translates into a given prob-ability of losing and a given effort to be exerted whentrying to win), whereas the benefit usually refers to theexternal resource at stake (a notable exception being so-cial competitive play, in which no such external incen-tive is present). The anticipated energy costs associatedwith competitive interactions and external resourcesmotivating the social conflict are thus pivotal to makingoptimal decisions. Yet we will here restrict the problemto winning-losing probability estimation, which alreadyprovide strong empirical evidence to the aforemen-tioned reinforcement-learning model.

In a recent study, we have induced an implicit domi-nance hierarchy in men through a competitive gameinvolving three opponents of different strengths, com-plemented by a noncompetitive control condition(Fig. 17.2A). Using fMRI, we first observed specific re-sponses to social defeats in the ventral striatum andother subcortical regions, which were correlated withtrait inhibition across subjects (Fig. 17.2C; see also“Interindividual Differences Resulting From Social Sta-tus and Personality”). Second, and more importantly,model-based analyses highlighted the functions of therostromedial cortices in tracking the dominance statusof opponents (i.e., anticipated winning-losing probabili-ties). More specifically, the RMPFC encoded opponent-specific prediction errors and appeared to monitor theprobability of winning against each player in a dynamicfashion, throughout the competitive task (Fig. 17.2D).

These findings were obtained by applying a classicalRescorlaeWagner rule [Eq. (17.1); Fig. 17.2E] tracking

were observed in the bilateral ventral striatum. The amplitude of defeat-related deactivations was correlated with the behavioral inhibitionpersonality trait across subjects. (D) Encoding of competitive prediction errors (cPE) in the RMPFC. Analyses showed that the activity changesobserved in the RMPFC encoded a signed competitive prediction error, which did not reflect winning or losing per se, the identity of the opponent,or interactions of these two factors. ncPE, noncompetitive prediction error; SDS, social dominance status (Id, opponent type). (E)Overview of the

computational model. Our reinforcement learning algorithm assumed that decisions are taken probabilistically (softmax policy) according to thevalue of each available opponent. Once the competition occurred, the value of the selected opponent was updated for the next trial proportional tothe prediction error elicited by the outcome [i.e., (Re SDS) with victory R¼ 1 and defeat R¼ 0] multiplied by the learning rate a. (F) Effects ofRMPFC tDCS on the parameters governing social dominance learning. Whereas average learning rates related to defeats and victories werebalanced in the sham group, stimulating the RMPFC using anodal tDCS induced a significant imbalance in the learning rates, with more weightplaced on victories and less weight place on defeats. From Ligneul R, Obeso I, Ruff CC, Dreher JC. Dynamical representation of dominance relationshipsin the human medial prefrontal cortex. Current Biology, in press.

=

III. SOCIAL DECISION NEUROSCIENCE

LEARNING SOCIAL DOMINANCE HIERARCHIES 215

the probability of winning in such agonistic interactions,by simply initializing Pwin (or SDS) at 0.5 and moni-toring the outcome R (0 for defeat, 1 for victory);

SDS tþ 1ð Þ ¼ SDS tð Þ þ a � R� SDS tð Þð Þ (17.1)

The prediction error term ReSDS(t) multiplied by thelearning rate a allows updating the anticipated chancesof winning (or SDS) in the next encounter at t þ 1. Alsocalled momentary social dominance status, or SDS, this“probability estimate could then be used to decidewhether one should defy another conspecific, accordingto a decisional policy such as the probabilistic softmaxrule. In real-life competitive settings, when the chancesof winning are deemed too low to initiate or continue thefight, the decision-maker may start to submit, hencemeeting the criterion of a dominance relationship [3,44].Moreover, generalizing or averaging of target-specificSDSs over all members of the group would naturallyunderlie the emergence of chronically dominant orsubordinate personality profiles.

Next, using an adapted version of our competitiveperceptual decision-making task (Fig. 17.2B), wedemonstrated that transcranial direct current stimula-tion (tDCS) applied over the RMPFC exerted a causal in-fluence over social dominance learning, reflected inhigher learning rates associated with victories and lowerlearning rates associated with defeats (Fig. 17.2F). Thisresult paralleled a study in mice in which viral injec-tions, producing an increase or decrease in overallMPFC activity, led to increases or decreases in the domi-nance ranks of animals [45].

In the past, it has been difficult to ascertain whetherneural network covarying with learning of social domi-nance was a cause or a consequence of the emergence ofsocial dominance in humans although original experi-ments in animals suggested a profound impact of domi-nance on brains and bodies. For example, it was shownthat selective lesions of the amygdala or the administra-tion of a serotonergic antidepressant can induce changesin the behavioral expression of dominance in monkeys[46,47] and human patients with lesions of the MPFCare also impaired in their ability to make social domi-nance attributions based on the narrative descriptionof diverse social interaction [48]. Beyond those classicalfindings, brain stimulation techniques such as tDCSshall thus constitute an important tool for the study ofcausality in fine-grained social dominance behaviorsamong healthy human subjects.

INTERINDIVIDUAL DIFFERENCES ANDSOCIAL DOMINANCE

Cognitive neuroscientists tend to rely on the conve-nient assumption that human brains are largely similar

across genders, ethnicities, sexes, ages, and socialgroups. This assumption facilitates the generalizationof observations typically made in a few dozens of stu-dents to the whole human population or, at least, tothe whole student population. Impeding predictive po-wer and reproducibility of empirical findings, thisassumption is unfortunately often invalid. Therefore,the systematic study of interindividual differences hasbegun, so that many neuroscientists now routinelyreport differences in personality traits, gray matter vol-umes, functional activities, or connectivity estimates tobetter explain the behavioral variability of their subjects[49]. In this perspective, social dominance holds a strongpotential to explain variability observed within socialgroups, which would otherwise be envisioned as homo-geneous (such as psychology students participating insocial learning experiments). Indeed, the study of socialdominance promises more than simply accounting forinterindividual differences in behavior and physiology:it may also offer mechanistic explanations for the emer-gence of neural, behavioral, cognitive, and social vari-ability. For example, the existence of clear-cutdominance hierarchies in pure strains of rodents (i.e.,genetically identical) stresses that the behavioral andphysiological features of dominant and subordinate an-imals derive largely from experience and adaptation.Moreover, although social dominance is a universalprinciple structuring social groups, it is also highlydependent on the culture, the gender, and the personal-ity of the participants.

Intercultural Differences in the Appraisal ofSocial Dominance

Human cultures vastly differ in how they value per-sonality traits related to social dominance. For example,the construction of self in collectivistic east Asiancultures tends to be more interdependent upon othergroup members than in individualistic societies inwhich the construction of self appears more as a questfor independence and autonomy with respect to othergroup members [50]. As Sedikides et al. [51] wrote, “inindividualistic cultures, the relevant dimension [ofself-construal] is agency, defined as a concern withpersonal effectiveness and social dominance. In collec-tivistic cultures, however, the relevant dimension iscommunion, defined as a concern with personal integra-tion and social connection.” Acknowledging culturaldifferences in the promotion of person-centric versusnormative-contextual models of self-construal is thuscrucial to avoid a partial, Western-centric conception ofbehaviors related to social dominance. For example, aninfluential study demonstrated that European Americanchildren are more motivated to solve anagrams whenthey could choose the category of problem to be solved

III. SOCIAL DECISION NEUROSCIENCE

17. SOCIAL DOMINANCE REPRESENTATIONS IN THE HUMAN BRAIN216

compared to when their mother or the experimenterchose it for them, whereas Asian American children dis-played the opposite pattern [52]. This finding indicatesthat, early in development, the motivational attitudes to-ward dominant others (here, the experimenter or theparent) vary greatly across cultures. It is also consistentwith the fact that high external locus of control (i.e., thefeeling that one’s own life is controlled by others) ismuch more strongly correlated with trait anxiety andnegative emotions in individualistic compared to collec-tivistic societies [53].

In the only available neuroimaging study (as of thiswriting) probing intercultural differences in the appraisal

of social dominance (Fig. 17.3A), Freeman and coauthorsdemonstrated that the neural correlates of social domi-nance expressed by body postures were reversed in theventral striatum and the MPFC when comparing Amer-ican (more activity in response to dominant postures)and Japanese subjects (more activity to subordinate pos-tures). Interestingly, these responses to dominant andsubordinate postures were also correlated with thebehavioral tendency of the subjects: a stronger responseto dominant postures predicted (self-reported) dominantbehaviors typical of Western cultures, whereas a strongerresponse to subordinate cues predicted the opposite be-haviors. Although we firmly believe that most of those

(B)

(C)

(A)90

80

70

60

50

40

30

20

10

0

1,5

1,0

0,5

0,75

0,50

0,25

0,00 -0,5

0,0

10 20 30 40SDO

SUP

Sig

nal c

hang

e (%

)

Con

trast

est

imat

es S

UP

>IN

F

MID INF 50 60 70

5-HTTLPR (% S allele)

BrazilRussia

UKAustraliaUSA

Hungary

S. AfricaSweden

Denmark

EstoniaAustria

SpainArgentina India Japan

IsraelPoland

FinlandGermanyFrance

Italy

SingaporeTaiwan

Mexico

PR ChinaKorea

Turkey

indi

vidu

alis

m–c

olle

ctiv

ism

0 20 40 60 80 100

L caudate

Japanese American Japanese American

mPFC

Bet

a va

lues

Bet

a va

lues

0.5

-0.5

0

0.5

-0.5

0

FIGURE 17.3 Brain-related interindividual differences in the appraisal of social hierarchies. (A) Whole-brain analysis testing a status

display � culture interaction effect. In Americans, the medial prefrontal cortex (mPFC) and the caudate nucleus exhibited reliably stronger bloodoxygen level-dependent (BOLD) responses to dominant stimuli relative to subordinate stimuli; in the Japanese, these same regions exhibited theopposite pattern, showing reliably stronger BOLD responses to subordinate stimuli relative to dominant stimuli. (From Freeman JB, Rule NO, Adams

RB, Ambady N. Culture shapes a mesolimbic response to signals of dominance and subordination that associates with behavior. Neuroimage 2009;47:353e59,

with permission.) (B) Correlation analysis between Hofstede’s individualismecollectivism index and frequency of S allele carriers of 5-

HTTLPR across 29 nations. Collectivistic nations showed higher prevalence of S allele carriers. (From Chiao JY, Blizinsky KD. Culture-gene

coevolution of individualismecollectivism and the serotonin transporter gene. Proc Biol Sci 2010;277:529e37, with permission.) (5-HTTLPR, serotonin-transporter-linked polymorphic region). (C) The relationship between political orientation and the neural sensitivity to competitive ranks. Inone of our experiments, the right anterior dorsolateral prefrontal cortex encoded social rank as induced by a prior competitive task against threeopponents. In addition, the sensitivity of this brain region to social rank was strongly correlated with the social dominance orientation acrosssubjects, thereby indicating that subjects more prone to legitimizing and reinforcing social inequalities are also more sensitive to competitivehierarchies [67]. INF, inferior; LPP, late positive potential; MID, middle; SDO, social dominance orientation; SUP, superior; tDCS, transcranialdirect current stimulation.

III. SOCIAL DECISION NEUROSCIENCE

INTERINDIVIDUAL DIFFERENCES AND SOCIAL DOMINANCE 217

inter-individual differences derive from the reinforce-ment-learning process described above, a cross-culturalgenetic analysis revealed that endogenous serotonin re-uptake capacity covaried with the individualisticecollectivistic opponency and the steepness of social hier-archies from one country to another (Fig. 17.3B) ([54]; seealso “Dopamine, Serotonin, and Social Hierarchies in Ro-dents and Nonhuman Primates”). Located in the pro-moter region of the serotonin transporter, thispolymorphism is also predictive of stress-coping strate-gies and resilience [55]. Cross-cultural differences in theperception of social dominance may thus be partly basedon polymorphisms of genes engaged in serotoninergic(but also dopaminergic) transmission. Yet, as we willsee, these polymorphisms may actually act by alteringthe learning process which is a the core of social domi-nance relationships.

Interindividual Differences Resulting FromSocial Status and Personality

By definition, the existence of a social dominance hi-erarchy means that different members of a given groupexperience different social environments and attributedifferent motivational values to specific social behav-iors. Social hierarchies define the type of social dilemmamost often faced by individuals and the range of optionsavailable to solve them. Thus, it is reasonable to expectthat different social ranks would turn into different pat-terns of brain activity anddpossiblyddifferent brainanatomies.

A paper by Noonan, Sallet, Mars, and coauthorsstudied 36 macaques living in groups of two to sevenmembers in which social dominance hierarchies couldbe reliably assessed [56] (see also Chapter 15). Their an-alyses showed that gray matter volumes markedly andreproducibly differed in several brain regions as a func-tion of social rank. Higher ranked animals had moregray matter in the hippocampus, the amygdala, andthe serotonergic brainstem, as well as in the medial tem-poral sulcus and the rostral prefrontal cortex. Becausethe last two were also correlated with the size of thegroup in which those animals were housed, the authorssuggested that they might be “linked to the social cogni-tive processes that are taxed by life in more complex so-cial networks and that must also be used if an animal isto achieve a high social status.” In addition, lowerranked animals had more gray matter in three regionsof the basal ganglia involved in habit learning and aver-sive processing: the dorsal striatum, the caudate nu-cleus, and the posterior putamen. As of this writing,the exact mechanisms that drive these correlations areunclear, but the ability of endogenous and drug-induced serotonin release to alter gray matter volumesand/or to stimulate neurogenesis in several brain re-gions offers promising perspectives [57,58]. In addition,

the pervasive effects of chronic stress on brain circuitryshould be taken into account, as it is well known thatsubordinate individuals tend to be more stressed thandominant individuals and more prone to developstereotypical behaviors, especially in captivity settings(see “Stress Asymmetries Paralleling Social HierarchyRank Have Adverse Consequences on Adrenocortical,Reproductive, and Neural Systems”).

In humans, the existence of neuroanatomical corre-lates of social ranks per se remains underexplored.Some developmental neuroimaging studies indicatethat, even after correction for several confounding fac-tors, higher parental socioeconomic status still pre-dicts higher prefrontal cortical thickness in children[59] and increased gray matter volumes in several brainregions, including prefrontal but also occipital, parietal,and limbic areas [60]. Behavioral sensitivity to socialdominance expressed by facial traits or induced bycompetitive games may also be predicted by neuroana-tomical variations in the insula and other regions [61,62].

In addition to the (sustained) neuroanatomical signa-ture of social hierarchies, humans also process socialevents differently, depending on their own social stand-ing. For example, Ly and coauthors demonstrated thatlow-status subjects had stronger striatal responseswhen presented with low-status faces, whereas theopposite was true for high-status subjects [63]. In oneof our experiments [64], we found that the sensitivityof the ventral striatum to social defeats in a competitiveperceptual decision-making game was correlated withthe behavioral inhibition personality trait [65], oftenlinked with social subordination and anxiety [66].Because more inhibited individuals had more salient de-activations in response to defeats in this structure, onecould infer that the repeated experience of social defeatsnot only lowers social status and social dominance, butalso heightens the overall sensitivity of the motivationalsystem to threats and negative events (Fig. 17.2C). More-over, in another experiment [67], we observed thatthe sensitivity of the right anterior prefrontal cortex tosocial rank of neutral faces was strongly correlatedwith the Social Dominance Orientation questionnaire(Fig. 17.3E), which reflects the degree to which one envi-sions social hierarchy and economic inequalities as legit-imate and necessary phenomena [68]. Deciphering theneurocognitive mechanisms involved in the appraisalof social hierarchy may thus help us understand real-world political divides [69].

NEUROCHEMICAL APPROACHES TOSOCIAL DOMINANCE AND

SUBORDINATION

The neurochemical processes involved in the emer-gence, maintenance, and consequences of social

III. SOCIAL DECISION NEUROSCIENCE

17. SOCIAL DOMINANCE REPRESENTATIONS IN THE HUMAN BRAIN218

dominance hierarchies by reinforcement-learning arekey to translating fundamental social neurosciencesinto new therapeutic options and to improve our un-derstanding of psychosocial disorders. Those disor-ders are largely mediated by pharmacologicalmodulation of plastic, stress-sensitive systems suchas hormonal and monoaminergic signaling. All“culture-like” features of animal societies can influ-ence the dynamical form taken by a social hierarchyand the manifold individual profiles composing it.Nonetheless, the hormonal and neural systems thatunderlie the variable expressions of dominance acrossspecies, groups, and individuals have been highlyconserved through evolution. Consequently, the acuteand long-term consequences of social defeats are nowwidely studied because of their ability to triggerrobust anxiety- and depression-like symptoms in ani-mals, thereby providing a useful translational modelof affective disorders [70].

Stress Asymmetries Paralleling Social HierarchyRank Have Adverse Consequences onAdrenocortical, Reproductive, andNeural Systems

In the attempt to explain dominance hierarchies in ro-dents, nonhuman primates, and humans, cortisol andtestosterone have long played the first roles. Relativelyeasy to quantify, they reproducibly covary with socialrank across species and experimental conditions. Theyare often jointly investigated because they interact atthe physiological and behavioral levels. Cortisol andtestosterone are the end products of two hormonalaxes reciprocally inhibiting each other: the HPA axisand the HPG axis, respectively [71,72] (Fig. 17.4A).Moreover, exposure to stressors activates a chain ofendocrine reactions, including secretion of glucocorti-coids by the adrenal glands, which reallocate energy re-sources necessary to adapt rapidly to the stressor. In thelong term, high levels of glucocorticoids can howeverdisrupt an essential negative feedback loop, hence lead-ing to immune function suppression as well as impair-ments in hippocampal and prefrontal functioning[73,74]. On the other hand, testosterone largely contrib-utes to muscle mass, male secondary sexual characteris-tics, and reactive aggressive behaviors, which arerelevant to predict the onset and the outcome of agonistinteractions [6,72].

Modern ethology has reported a great variability be-tween and even within species, regarding whetherhigh- or low-ranking animals are the ones who are themost stressed in a dominance hierarchy. Many factorsinfluence such rank-associated stress in stratifiedmammalian societies. Such factors include, but are not

limited to, species-level variations in style of breedingsystem (cooperative/competitive), social and matingsystems, housing, despotic versus egalitarian hierarchystyle [75], and hierarchy stability within species [76]. Indespotic hierarchies, resource access is skewed mark-edly and dominant positions are attained throughaggression and intimidation, whereas in “egalitarian”hierarchies resource distribution is more equal anddominance is attained with the support of subordinateindividuals. A general concept to help resolve these dif-ferences in the relationship between rank and stressacross species is that it is the rank that experiences themost physical and psychological stressors that tends todisplay the most severe stress-related response.

In primates, glucocorticoid levels are often higher insubordinate males whenever a dominance hierarchy isstabilized and testosterone levels are generally indepen-dent of social rank [76,77]. However, higher-rankingmales tend to experience higher testosterone and gluco-corticoid (stress hormone) levels than lower-rankingmales whenever their dominance rank is threatened(i.e., in a period of social instability) [76,77]. Togetherwith the impact of living conditions (i.e., captivity, semi-captivity, free ranging, access to resources, size of thegroups, etc.), this phenomenon probably explains thevariability in empirical findings between and withinspecies regarding whether high- or low-ranking animalsendure more stress in a dominance hierarchy [6]. Ulti-mately, the reason why a psychosocial stressor is experi-enced as such by a given individual may depend on theamount of control exerted over its termination and thepredictability of its occurence.

In humans, absolute dominance ranks have littlemeaning because of the multidimensional nature of so-cial success in our species. It has been found that low so-cioeconomic status (SES) is reliably associated with adisruption of endogenous circadian fluctuations incortisol levels, suggesting that cortisol might be linkedwith social hierarchy in our species as well [78,79]. Theinfluence of parental SES and parent education in thisphenomenon [80] may suggest the existence of a trans-generational epigenetic mechanism as observed instress-related disorders [81], as the relationship holds af-ter controlling for many confounding factors includingthe offspring’s actual SES. This finding is also consistentwith the well-established observation that uncontrolla-ble psychosocial stressors involving a real or possible so-cial subordination component invariably induce stressand cortisol release in humans [82,83].

The exact cognitive role of testosterone in humans isstill debated. Although early studies proposed thattestosterone plays a role in reactive aggression ratherthan aggression per se, studies proposed that it has animportant function to establish social status in bothmen and women [71,84]. Yet, testosterone does not

III. SOCIAL DECISION NEUROSCIENCE

NEUROCHEMICAL APPROACHES TO SOCIAL DOMINANCE AND SUBORDINATION 219

correlate linearly with socioeconomic status in humans,for two reasons. First, the aggressive tendencies of high-testosterone individuals are generally counterselected inmany social organizations. Second, the net behavioral ef-fects of testosterone on dominance-related behaviorsdepend upon cortisol levels: indeed, when restingcortisol levels are high, the positive association seen be-tween dominance behaviors and testosterone is lost oreven reversed [10,84]. A plausible role for testosteronewould thus be to regulate the salience of and the

reactivity to social threats as a function of dominanceranks [85], whereas glucocorticoids would modulatethe ability to shift flexibly between a “salience” network,which supports rapid but rigid decisions, and an “exec-utive control” network, which supports flexible, elabo-rate social decisions (see also Chapter 30). Accordingto this hypothesis, one may expect that the high anddisruptive cortisol levels observed in chronic stressdiminish the social and biological value oftestosterone-mediated dominance behaviors because of

FIGURE 17.4 Hormonal and neuromodulatory bases of social dominance. (A) Complexities of the cortisoletestosterone relationship

involved in the maintenance of social dominance. The hypothalamicepituitaryeadrenal (HPA) and hypothalamicepituitaryegonadal (HPG)axes are represented with the brain structures involved, hormonal cascades, and functional interrelations. (From Terburg D et al. The testosterone-

cortisol ratio: a hormonal marker for proneness to social aggression. Int J L Psychiatry 2009;32:216e23, with permission.) (B) Role of dopamine in the

emergence of social dominance and facilitation of cocaine addiction in subordinate individuals. [18F]FCP binding potential increases indominant monkeys (left). (Right) Mean intake of cocaine per session for dominant (white symbols) and subordinate (black symbols) monkeys, as afunction of the cocaine concentration in the self-administered solution. (FromMorgan D et al. Social dominance in monkeys: dopamine D2 receptors and

cocaine self-administration. Nat Neurosci 2002:169e74.) (FCP, fluoroclebopride) (C) Involvement of serotonin neurons in the behavioral conse-

quences of social defeats. Optogenetic targeting showed that the serotonin neurons of susceptible (SUS) mice showing anxiety-like symptomsfollowing social defeats were more inhibited than control (CTRL) or resilient (RES) mice showing no such symptoms. 5-HT, serotonin; vmPFC,ventromedial prefrontal cortex. (dm/vmDR, dorsomedial/ventromedial dorsal raphe; IPSC, inhibitory post synaptic potential) (From Challis C, et al.

Raphe GABAergic neurons mediate the acquisition of avoidance after social defeat. J Neurosci 2013;33:13978e88. 13988a, with permission.)

III. SOCIAL DECISION NEUROSCIENCE

17. SOCIAL DOMINANCE REPRESENTATIONS IN THE HUMAN BRAIN220

the loss in behavioral flexibility. Moreover, transientfluctuations in cortisol levels seem causally involved inthe adaptive memorization of social dominance relation-ships induced by competitive encounters [86], henceimplying that disruption of cortisol signaling leads toimprecise representations of social dominancerelationships.

While the HPG and HPA axes certainly play animportant role in the implementation of proximal domi-nance behaviors, such as the arbitration of the “fleeefightethink” dilemma elicited by any social conflict,their functional physiology seems incompatible withthe implementation of higher-order cognitive processesmodulated by learned social hierarchies. Thephenomena reported above are more likely to be medi-ated by central dopaminergic and serotonergic systems.Indeed, their modes of release in the forebrain enablealso a refined coding of social information. To date,only a limited literature has investigated their roles inthe social hierarchies of humans and nonhumansbecause of the methodological constraints associatedwith the measurement of central neuromodulators.

Dopamine, Serotonin, and Social Hierarchies inRodents and Nonhuman Primates

In rodents, the emergence of an avoidant, subordinatebehavior following social defeat is causally mediated byplasticity in the ventral tegmental area (VTA; containingdopamine neurons) occurring during and after acompetitive interaction with negative outcome. Moreprecisely, the sensitization of dopamine (DA) neuronsoccurs only in “susceptible” mice, which display a sub-ordinate behavioral pattern following social defeats [87],and transient light stimulation of the VTA 1 day after so-cial competition can reinstate avoidance and anhedoniasymptoms induced by social defeats in most mice(Chaudhury et al. [88]). Suppression of dopaminergicfiring may thus doubly contribute to the emergence ofsubmissive behavior by inhibiting reward-related pro-cesses and by exacerbating the avoidance of subsequentsocial contacts with others, especially when they aredominant.

An outstanding question is whether postsynapticsites to dopaminergic neurons are modulated by DAduring social interaction. Does DA encode the pres-ence/absence of dominant individuals? Does it encodesocial prediction errors (see “Social Hierarchies as aMajor Evolutionary Pressure and Pivotal Feature ofSocieties”) as in nonsocial settings? In dominant rats,microdialysis experiments showed that the imminenceof a social conflict induces a strong release of DA inthe nucleus accumbens [89,90]. In monkeys, no studyhas investigated dopaminergic firing per se as of this

writing, but the perception of dominant individualswas shown to interfere with a reward valuation processtypically controlled by DA neurons [91,92], and onestudy showed that neurons in the ventral striatumdintensely innervated by DA neuronsdmay encode theexperience of social subordination and dominance dur-ing social conflicts over reward [93]. These findings arehighly relevant for clinicians, because the emergence ofdominance hierarchies within groups of monkeysinduced reversible changes in D2/D3 receptor availabil-ity, which mediates detrimental behavioral changes insubordinate individuals, including enhanced suscepti-bility to cocaine addiction (Fig. 17.4B) [94,95]. Interest-ingly, a later study extended this finding to humans,using a subjective social status questionnaire [96].Beyond the evidence they provide for an involvementof DA in learning social dominance relationships, thesefluctuations of D2/D3 receptors may also explain whyrecreational dopaminergic drugs are able to artificiallyupregulate self-esteem, self-confidence, and socialdominance in the short term; why they result in degen-erate social behavior patterns when used for a longerterm; and why the experience of juvenile social stress,dominance motives, and low SES predispose to psychos-timulant usage [5,97].

Compared to DA, the exact roles played by serotoninin reinforcement learning are less clear and constitute anactive area of research. Nonetheless, this neurotrans-mitter is undoubtedly involved in the adaptive regula-tion of many aspects of social and nonsocial behaviors[98]. An influential theory has long maintained that se-rotonin (5-HT) might implement the coupling betweenthe anticipation of aversive events and behavioral inhi-bition [99,100], which strongly resonates with situa-tions of social subordination in which one has toinhibit the decision to compete for resources in frontof threatening and powerful conspecifics. An outsidertheorydwhich recently gained strong empirical sup-port from electrophysiological recordings and optoge-netic manipulation of 5-HT neurons[101,102]dproposed that serotonin firing might insteadpromote patience and cognitive control within bothappetitive and aversive contexts [103]. Interesting,this second theory also resonates with dominance rela-tionships, as dominant individuals tend to be impul-sive decision-makers, whereas subordinates typicallyhave to “wait their turn,” because of the core impor-tance of pecking orders in any dominance hierarchy(for both nutritional and social resources).

To date, the most striking demonstration of the causalrole played by serotonin in the establishment of social hi-erarchies comes perhaps from the study of Raleigh andcollaborators [46]. In this series of experiments performedin 12 groups of three vervet monkeys, the authorsshowed that the “enhancement of serotonin signaling”

III. SOCIAL DECISION NEUROSCIENCE

NEUROCHEMICAL APPROACHES TO SOCIAL DOMINANCE AND SUBORDINATION 221

by a selective reuptake inhibitor and the suppression ofserotonin signaling by a nonselective serotonin antago-nist could induce dominance or subordination, respec-tively, in treated monkeys. More recently, it wasconfirmed that social defeats trigger sensitization ofGABAergic neurons in the main serotonergic dorsalraphe nucleus (DRN) irrigating the forebrain [104]. Mir-roring the aforementioned effect in the VTA [87,88], thissensitization phenomenon was visible only in the “sus-ceptible” mice, which displayed an avoidant,subordinate-like behavioral pattern following socialdefeats (Fig. 17.4C). In humans, evidence supporting arole of 5-HT in social dominance is still sparse, but itwas shown that enhancing 5-HT level through antide-pressant medications or tryptophan supplementation(i.e., the precursor of 5-HT biosynthesis) might increasethe frequency of dominance-related behaviors ineveryday life [105,106]. Finally, in line with the definitionof social dominance as an asymmetry of control over so-cial stressors and social rewards, strong empirical evi-dence has emphasized that the reciprocal connectionsbetween the serotonergic DRN and the MPFC are crucialto adapt behavior in front of controllable stressors[87,107].

The joint involvement of DA and 5-HT in social domi-nance thus suggests (1) that the learning and decision-making processes controlled by these neuromodulatorsare central to the emergence of social hierarchies in mam-mals, (2) that social dominance might affect domain-general learning and decision-making through its influ-enceon thoseneuromodulatory systems, and (3) that theseneuromodulatory systems might have been sculptedthroughout evolution to facilitate high flexibility in socialbehaviors, as required in species forming a dominance hi-erarchy. However, more research is needed to elucidatetheir exact computational roles, because no study hasinvestigated how DA and 5-HT neurons react to conspe-cifics of different social ranks nor how they would imple-ment the decision to compete or not against others.

CONCLUSION

The study of social dominance promises much morethan simply accounting for interindividual differencesin behavior and neurophysiology. Indeed, social domi-nance processes may offer mechanistic explanations forthe emergence of such differences. Animal research in-dicates that social dominance affects serotonergic anddopaminergic neuromodulatory pathways responsiblefor behavioral and neural plasticity. It also affects theanatomical and functional properties of several brainstructures traditionally linked with social and nonso-cial perception, learning, and decision-making. It isthe very nature of dominance hierarchies to shape

social behaviors and to promote the coexistence ofvarious profiles within a single social group. In thisperspective, the development of refined computationalmodels and social learning tasks probing social domi-nance in humans (coupled with neuroimaging) mayhelp us understand and treat specific psychosocial dis-orders that seem particularly prevalent in humans rela-tive to other apes, such as pathological aggression,social anxiety, schizophrenia, psychopathy, and someforms of depression.

Acknowledgments

This workwas performedwithin the framework of the LABEXCORTEX(ANR-11-LABX-0042) of Universite de Lyon,within the program “Inves-tissements d’Avenir” (ANR-11-IDEX-0007) operated by the French Na-tional Research Agency (ANR). JCD was also supported by GrantANR-14-CE13-006 and by the European Institute for Advanced StudyFellowship Programme at the Hanse-Wissenschaftskolleg.

References

[1] Wilson R, Keil F. The MIT Encyclopedia of the cognitive sciences.Comput Linguist, vol. 26. Cambridge (Massachusetts): MITPress; 1999.

[2] Ligneul R. Bases cerebrales des processus de competition et dehierarchisation sociales. 2014 [Doctoral dissertation], http://www.theses.fr/s101849.

[3] Bernstein IS. Dominance: the baby and the bathwater. BehavBrain Sci 1981;4:419e57.

[4] Price J. The dominance hierarchy and the evolution of mentalillness. The Lancet 1967;290:243e6.

[5] Sandi C, Haller J. Stress and the social brain: behavioural effectsand neurobiological mechanisms. Nat Rev Neurosci 2015;16:290e304.

[6] Sapolsky RM. The influence of social hierarchy on primatehealth. Science 2005;308:648e52.

[7] Sloman L, Gilbert P. Subordination and defeat: an evolutionaryapproach to mood disorders and their therapy. Lawrence Erl-baum Associates; 2000.

[8] Singh-Manoux A, Marmot MG, Adler N. Does subjective socialstatus predict health and change in health status better thanobjective status? Psychosom Med 2005;67:855e61.

[9] Marmot M, Wilkinson R. Social Determinants of health. OxfordUniversity Press; 2005. At: http://books.google.com/books?id¼AmwiS8HZeRIC&pgis¼1.

[10] Mehta PH, Josephs RA. Testosterone and cortisol jointly regulatedominance: evidence for a dual-hormone hypothesis. HormBehav 2010;58:898e906.

[11] Cowlishaw G, Dunbar RIM, Street G. Dominance rank and mat-ing success in male primates. Anim Behav 1991:1045e56. http://dx.doi.org/10.1016/S0003-3472(05)80642-6.

[12] Ellis L. Dominance and reproductive success among nonhumananimals: a cross-species comparison. Ethol Sociobiol 1995;16:257e333.

[13] McGuire MT, Brammer GL, Raleigh MJ. Resting cortisol levelsand the emergence of dominant status among male vervetmonkeys. Horm Behav 1986;20:106e17.

[14] van der Kooij MA, Sandi C, Kooij M, Van Der A, Sandi C. The ge-netics of social hierarchies. Curr Opin Behav Sci 2015;2:52e7.

[15] McNamara JM, Houston AI. State-dependent life histories. Na-ture 1996;380:215e21.

III. SOCIAL DECISION NEUROSCIENCE

17. SOCIAL DOMINANCE REPRESENTATIONS IN THE HUMAN BRAIN222

[16] Wolf M, van Doorn GS, Leimar O,Weissing FJ. Life-history trade-offs favour the evolution of animal personalities. Nature 2007;447:581e4.

[17] Jolly A. Lemur social behavior and primate intelligence. Science1966;153:501e6.

[18] Humphrey NK. The social function of intellect. 1976. p. 303e17.http://dx.doi.org/10.2307/375925.

[19] Whiten A, Byrne R. Machiavellian intelligence hypothesis.Machiavellian Intell 1988. At: http://www.rm-f.net/wpennywis/MITECS/Articles/whiten.html.

[20] Dunbar RIM. Neocortex size as a constraint on group size inprimates. J Hum Evol 1992;22:469e93.

[21] Behrens TEJ, Hunt LT, Rushworth MFS. The computation of so-cial behavior. Science 2009;324:1160e4.

[22] Frith CD. The social brain?. 2007. http://dx.doi.org/10.1098/rstb.2006.2003.

[23] Chiao JY, et al. Neural representations of social status hierarchyin human inferior parietal cortex. Neuropsychologia 2009;47:354e63.

[24] Kumaran D, Melo HL, Duzel E. The emergence and representa-tion of knowledge about social and nonsocial hierarchies.Neuron 2012;76:653e66.

[25] Marsh AA, Blair KS, Jones MM, Soliman N, Blair RJR. Domi-nance and submission: the ventrolateral prefrontal cortex and re-sponses to status cues. J Cogn Neurosci 2009;21:713e24.

[26] Todorov A, Engell AD. The role of the amygdala in implicit eval-uation of emotionally neutral faces. Soc Cogn Affect Neurosci2008;3:303e12.

[27] Lindner M, Hundhammer T, Ciaramidaro A, Linden DEJ,Mussweiler T. The neural substrates of person comparisoneanfMRI study. Neuroimage 2008;40:963e71.

[28] Farrow TFD, et al. Higher or lower? The functional anatomy ofperceived allocentric social hierarchies. Neuroimage 2011;57:1552e60.

[29] Zink CF, et al. Know your place: neural processing of social hier-archy in humans. Neuron 2008;58:273e83.

[30] Watanabe N, Yamamoto M. Neural mechanisms of socialdominance. Front Neurosci 2015;9.

[31] Dunne S, Doherty JPO. Insights from the application of compu-tational neuroimaging to social neuroscience. Curr Opin Neuro-biol 2013;23:1e6.

[32] Burke CJ, Tobler PN, Baddeley M, Schultz W. Neural mecha-nisms of observational learning. Proc Natl Acad Sci USA 2010;107:14431e6.

[33] Behrens TEJ, Hunt LT, Woolrich MW, Rushworth MFS. Associa-tive learning of social value. Nature 2008;456:245e9.

[34] Hampton AN, Bossaerts P, O’Doherty JP, Doherty JPO. Neuralcorrelates of mentalizing-related computations during strategicinteractions in humans. Proc Natl Acad Sci USA 2008;105:6741e6.

[35] Renes RA, van Haren NEM, Aarts H, Vink M. An exploratoryfMRI study into inferences of self-agency. Soc Cogn Affect Neu-rosci 2014. Nsu106.

[36] Seo H, Cai X, Donahue CH, Lee D. Neural correlates of strategicreasoning during competitive games. Science 2014;364:340e5.

[37] Devaine M, Hollard G, Daunizeau J. Theory of mind: did evolu-tion fool us? PLoS One 2014;9.

[38] Fareri DS, Delgado MR. Differential reward responses duringcompetition against in- and out-of-network others. Soc CognAffect Neurosci 2013;9:412e20.

[39] Katsyri J, Hari R, Ravaja N, Nummenmaa L. The opponent mat-ters: elevated fMRI reward responses to winning against a hu-man versus a computer opponent during interactive videogame playing. Cereb Cortex 2012;23:2829e39.

[40] Le Bouc R, Pessiglione M. Imaging social motivation: distinctbrain mechanisms drive effort production during collaborationversus competition. J Neurosci 2013;33:15894e902.

[41] Polosan M, et al. An fMRI study of the social competition inhealthy subjects. Brain Cogn 2011;77:401e11.

[42] Chaminade T, et al. How do we think machines think? An fMRIstudy of alleged competition with an artificial intelligence. FrontHum Neurosci 2012;6:103.

[43] Decety J, Jackson PL, Sommerville JA, Chaminade T,Meltzoff AN. The neural bases of cooperation and competition:an fMRI investigation. Neuroimage 2004;23:744e51.

[44] Rowell TE. The concept of social dominance. Behav Biol 1974;11:131e54.

[45] Wang F, et al. Bidirectional control of social hierarchy bysynaptic efficacy in medial prefrontal cortex. Science 2011;334:693e7.

[46] Raleigh MJ, McGuire MT, Brammer GL, Pollack DB, Yuwiler A.Serotonergic mechanisms promote dominance acquisition inadult male vervet monkeys. Brain Res 1991;559:181e90.

[47] Bauman MD, Toscano JE, Mason WA, Lavenex P, Amaral DG.The expression of social dominance following neonatal lesionsof the amygdala or hippocampus in rhesus monkeys (Macaca

mulatta). Behav Neurosci 2006;120:749e60.[48] Karafin MS, Tranel D, Adolphs R. Dominance attributions

following damage to the ventromedial prefrontal cortex.J Cogn Neurosci 2004;16:1796e804.

[49] Kanai R, Rees G. The structural basis of inter-individual differ-ences in human behaviour and cognition. Nat Rev Neurosci2011;12:231e42.

[50] Cross SE, Madson L. Models of the self: self-construals andgender. Psychol Bull 1997;122:5e37.

[51] Sedikides C, Gaertner L, Toguchi Y. Pancultural self-enhancement. J Pers Soc Psychol 2003;84:60e79.

[52] Iyengar SS, LepperMR. Rethinking the value of choice: a culturalperspective on intrinsic motivation. J Pers Soc Psychol 1999;76:349e66.

[53] Cheng C, Cheung S-F, Chio JH, Chan MS. Cultural meaning ofperceived control: a meta-analysis of locus of control and psy-chological symptoms across 18 cultural regions. Psychol Bull2013;139.

[54] Chiao JY, Blizinsky KD. Culture-gene coevolution ofindividualism-collectivism and the serotonin transporter gene.Proc Biol Sci 2010;277:529e37.

[55] Homberg JR, Lesch K-PP. Looking on the bright side of serotonintransporter gene variation. Biol Psychiatry 2011;69:513e9.

[56] Noonan MP, et al. A neural circuit covarying with social hierar-chy in macaques. PLoS Biol 2014;12:e1001940.

[57] Frodl T, et al. Reduced gray matter brain volumes are associatedwith variants of the serotonin transporter gene in majordepression. Mol Psychiatry 2008;13:1093e101.

[58] Mahar I, Bambico FR, Mechawar N, Nobrega JN. Stress, seroto-nin, and hippocampal neurogenesis in relation to depressionand antidepressant effects. Neurosci Biobehav Rev 2014;38:173e92.

[59] Lawson GM, Duda JT, Avants BB, Wu J, Farah MJ. Associationsbetween children’s socioeconomic status and prefrontal corticalthickness. Dev Sci 2013;16:641e52.

[60] Yeo RA, et al. The impact of parent socio-economic status on ex-ecutive functioning and cortical morphology in individuals withschizophrenia and healthy controls. Psychol Med 2014;44:1257e65.

[61] Getov S, Kanai R, Bahrami B, Rees G. Human brain structure pre-dicts individual differences in preconscious evaluation of facialdominance and trustworthiness. Soc Cogn Affect Neurosci2014. http://dx.doi.org/10.1093/scan/nsu103.

[62] Santamaria-Garcia H, Burgaleta M, Sebastian-Galles N. Neuro-anatomical markers of social hierarchy recognition in humans:a combined ERP/MRI study. J Neurosci 2015;35:10843e50.

III. SOCIAL DECISION NEUROSCIENCE

REFERENCES 223

[63] Ly M, Haynes MR, Barter JW, Weinberger DR, Zink CF. Subjec-tive socioeconomic status predicts human ventral striatalresponses to social status information. Curr Biol 2011;21:794e7.

[64] Ligneul R, Obeso I, Ruff CC, Dreher JC. Dynamical representa-tion of dominance relationships in the human medial prefrontalcortex. Current Biology, in press

[65] Carver CS, White TL. Behavioral inhibition, behavioral activa-tion, and affective responses to impending reward and punish-ment: the BIS/BAS Scales. J Pers Soc Psychol 1994;67:319e33.

[66] Keltner D, Gruenfeld DH, Anderson C. Power, approach, andinhibition. Psychol Rev 2003;110:265e84.

[67] Ligneul R, Girard R, Dreher J-C. Social brains and divides: theinterplay between political orientation and the neural sensitivityto social ranks. 2016 [Submitted for publication].

[68] Pratto F, Sidanius J, Stallworth LM, Malle BF. Social dominanceorientation: a personality variable predicting social and politicalattitudes. J Pers Soc Psychol 1994;67:741e63.

[69] Stanton SJ, Beehner JC, Saini EK, Kuhn CM, Labar KS. Domi-nance, politics, and physiology: voters’ testosterone changes onthe night of the 2008 United States presidential election. PLoSOne 2009;4:e7543.

[70] Chaouloff F. Social stress models in depression research: what dothey tell us? Cell Tissue Res 2013. http://dx.doi.org/10.1007/s00441-013-1606-x.

[71] Dreher J-C. Neuroimaging evidences of gonadal steroid hor-mone influences on reward processing and social decision-making in humans. Brain Mapp 2015;3. Elsevier Inc.

[72] Terburg D, et al. The testosterone-cortisol ratio: a hormonalmarker for proneness to social aggression. Int J Law Psychiatry2009;32:216e23.

[73] Sapolsky RMM, Krey LC, McEwen BS. Glucocorticoid-sensitivehippocampal neurons are involved in terminating the adrenocor-tical stress response. Proc Natl Acad Sci USA 1984;81:6174e7.

[74] Arnsten AFT. Stress signalling pathways that impair prefrontalcortex structure and function. Nat Rev Neurosci 2009;10:410e22.

[75] Thierry B, Singh M, Kaumanns W. Macaque societies: a modelfor the study of social organization. Cambridge UniversityPress,; 2004. 41.

[76] Gesquiere LR, et al. Life at the top: rank and stress in wild malebaboons. Science 2011;333:357e60.

[77] Sapolsky RM, Ray JC. Styles of dominance and their endocrinecorrelates among wild olive baboons (Papio anubis). Am J Prima-tol 1989;18:1e13.

[78] Cohen S, et al. Socioeconomic status is associated with stresshormones. Psychosom Med 2006;68:414e20.

[79] Agbedia OO, et al. Blunted diurnal decline of cortisol amongolder adults with low socioeconomic status. Ann NY Acad Sci2011;1231:56e64.

[80] Desantis AS, Kuzawa CW, Adam EK. Developmental origins offlatter cortisol rhythms: socioeconomic status and adult cortisolactivity. Am J Hum Biol 2015;27:458e67.

[81] Yehuda R, et al. Parental posttraumatic stress disorder as avulnerability factor for low cortisol trait in offspring of holocaustsurvivors. Arch Gen Psychiatry 2007;64:1040e8.

[82] Dickerson SS, Kemeny ME. Acute stressors and cortisol re-sponses: a theoretical integration and synthesis of laboratoryresearch. Psychol Bull 2004;130:355e91.

[83] Wirth MM, Welsh KM, Schultheiss OC. Salivary cortisol changesin humans after winning or losing a dominance contest dependon implicit power motivation. Horm Behav 2006;49:346e52.

[84] Montoya ER, Terburg D, Bos PA, van Honk J. Testosterone,cortisol, and serotonin as key regulators of social aggression:a review and theoretical perspective. Motiv Emot 2012;36:65e73.

[85] Hermans EJ, Ramsey NF, van Honk J. Exogenous testosteroneenhances responsiveness to social threat in the neural circuitryof social aggression in humans. Biol Psychiatry 2008;63:263e70.

[86] Cordero MI, Sandi C, Cordero I, Sandi C. Stress amplifies mem-ory for social hierarchy. Front Neurosci 2007;1:175e84.

[87] Cao JL, et al. Mesolimbic dopamine neurons in the brain rewardcircuit mediate susceptibility to social defeat and antidepressantaction. Journal Neurosci 2010;30(49):16453e8.

[88] Chaudhury D, et al. Rapid regulation of depression-related be-haviours by control of midbrain dopamine neurons. Nature2013;493:532e6.

[89] Tidey JW, Miczek KA. Social defeat stress selectively alters mes-ocorticolimbic dopamine release: an in vivo microdialysis study.Brain Res 1996;721:140e9.

[90] Ferrari PF, et al. Accumbal dopamine and serotonin in anticipa-tion of the next aggressive episode in rats. Eur J Neurosci 2003;17:371e8.

[91] Deaner RO, et al. Monkeys pay per view: adaptive valuation ofsocial images by rhesus macaques. Curr Biol 2005;15:543e8.

[92] Klein JT, Platt ML. Social information signaling by neurons inprimate striatum. Curr Biol 2013;23:691e6.

[93] Santos GS, Nagasaka Y, Fujii N, Nakahara H. Encoding of socialstate information by neuronal activities in the macaque caudatenucleus. Soc Neurosci 2012;7:42e58.

[94] Morgan D, et al. Social dominance in monkeys: dopamine D2 re-ceptors and cocaine self-administration. Nat Neurosci 2002:169e74. http://dx.doi.org/10.1038/nn798.

[95] Nader MA, et al. Social dominance in female monkeys: dopa-mine receptor function and cocaine reinforcement. Biol Psychia-try 2012;72:414e21.

[96] Martinez D, et al. Dopamine type 2/3 receptor availability in thestriatum and social status in human volunteers. Biol Psychiatry2010;67:275e8.

[97] Sinha R. Chronic stress, drug use, and vulnerability to addiction.Ann NYAcad Sci 2008;1141:105e30.

[98] Kiser D, Steemers B, Branchi I, Homberg JR. The reciprocal inter-action between serotonin and social behaviour. Neurosci Bio-behav Rev 2012;36(2):786e98.

[99] Dayan P, Huys QJM. Serotonin in affective control. Annu RevNeurosci 2009;32:95e126.

[100] Dayan P, Huys Q. Serotonin’s many meanings. 2015. p. 3e5.http://dx.doi.org/10.7554/eLife.06346.

[101] Miyazaki KW, et al. Optogenetic activation of dorsal raphe sero-tonin neurons enhances patience for future rewards. Curr Biol2014;24:2033e40.

[102] Miyazaki KW, Miyazaki K, Doya K. Activation of dorsal rapheserotonin neurons is necessary for waiting for delayedrewards. J Neurosci 2012;32:10451e7.

[103] Doya K. Modulators of decision making. Nat Neurosci 2008;11:410e6.

[104] Challis C, et al. Raphe GABAergic neurons mediate the acquisi-tion of avoidance after social defeat. J Neurosci 2013;33:13978e88. 13988a.

[105] Moskowitz DS, Pinard G, Zuroff DC, Annable L, Young SN. Theeffect of tryptophan on social interaction in everyday life: aplacebo-controlled study. Neuropsychopharmacology 2001;25:277e89.

[106] Tse WS, Bond AJ. Serotonergic intervention affects both socialdominance and affiliative behaviour. Psychopharmacol (Berl)2002;161:324e30.

[107] Amat J, et al. Medial prefrontal cortex determines how stressorcontrollability affects behavior and dorsal raphe nucleus. NatNeurosci 2005;8:365e71.

[108] Freeman JB, Rule NO, Adams RB, Ambady N. Culture shapes amesolimbic response to signals of dominance and subordinationthat associates with behavior. Neuroimage 2009;47:353e9.

III. SOCIAL DECISION NEUROSCIENCE

17. SOCIAL DOMINANCE REPRESENTATIONS IN THE HUMAN BRAIN224